It was my official first day on the farm. We worked in the greenhouse and then the field planting scallions, lunch break, then more scallion planting until it started snowing.

Onion Trimming

In USDA zone 6b (-5 to 0 °F/-20.6 to -17.8 °C), the crops you start in the greenhouse this early in the season are often the hardy kind…brassicas like kale and allums like onions…and lettuces…things that can withstand the occasional light frost under row cover.

Onions in particular love to grow their greens tall, but what we really want once we’re planting them are healthy roots and shoots that make it easy to remove from cell trays. So we “trim” them in the greenhouse to about 2-3 inches when they get cumbersome to handle…kind of like giving them a haircut when they start to get a bit shaggy. This also encourages the plant to put its energy into its root system.

However, doing hundreds of trays with scissors by hand gave me time to think about other options such as electric shears or paired with a vacuum system like a Flowbee for veg. Some things simply need to be done as told though and there’s almost always a good reason for the way it works already.

The Greenhouse

The 20x80-foot greenhouse is already full of seedlings, at least a month’s work already by others. It has a 4-foot walkway down the middle, but otherwise each growing surface side is about 8 feet deep to the walls. The trays sit on “ladders” made of 2x4-foot beams and laid on sawhorses. Getting to the far reaches means taking some trays out to create standing space in the middle of a space, then moving around various trays while standing in the space.

Lunch

Lunch was a fresh fried egg sandwich made by Jamie. It’s really nice that we all sit together and that he feeds us. I plan to bring some home made things like breads and pickled veg throughout the season.

Planting and Physical Fitness (or Not)

After finishing up planting in the snow, my knees were covered in mud. I haven’t figured out how to bend and squat instead of kneeling across four foot rows.

Though I did a lot of farm work last season as a volunteer, these were only 3-4 hour shifts and I wasn’t too worried about looking out of shape. Keeping up with the crew is going to be an uphill struggle this month, despite how understanding and polite about it. One of them even goes to the gym after working an 8am-5pm day on the farm.

It will get easier and I am out of shape but not for long.

Side-note: a change of each article of base layer clothing is always a good thing to bring with you.

Official Part-time Employee

After that, due to some email issues, I went home to change and came back with my passport which was needed to get my required paperwork into payroll.

I’m a verbal communicator. In many contexts, I have to hold this back to work with folks effectively. I’ve always wondered where I could put this to better use rather than feel like it was a personal character flaw. My current hypothesis is that if on my way home from the farm, I record personal audio notes, I don’t have to hold back or be word perfect.

When I’m done recording, I just upload them to a specific Google Drive folder where some of my custom code looks to pick up and process new recordings. I’ve already got my own personal transcription service in my homelab as well as an LLM-based summarization service using simple context templates to generate mostly bullet points.

Here’s an example from today’s audio notes:

Activities Done:

1. Planting scallions
2. Potted up different varieties of tomatoes and eggplant
3. Trimming onions
4. Loading and unloading trucks

New Things Not Yet Encountered:

1. Sawhorses with ladders and a four-foot walking way in between
2. Specific type of ladder (H or pie-shaped)
3. How to move trays around the least, preserving efficiency

Questions and Future Areas of Research:

1. Names of different varieties
2. Management over effects of repetitive work on back and knees
3. Other ways to trim onions and chives (possibly electric shears)
4. Specific conditions required for irrigation systems

Suggested Actions:

1. Review and remember names of tomato varieties
2. Research mechanisms that might repetitive speed greenhouse trimming
3. Invest in a personal temperature gauge or indoor temperature sensor
4. Research specifics of irrigation systems and their differences from other fields

Today it was just the three of us. We started in the greenhouse seeding lettuce trays. Then I went up to the barn to learn how to clean and pack eggs, 64 dozen cartons. Then we seeded and potted up more trays in the greenhouse.

Egg Economics

When the national average price per dozen eggs is $6.27, the organic cage-free kind at $8.49…this is around $540 worth of eggs and 3-4 hours (around $60 of labor) about twice a week. I’m not yet sure how many chickens are laying now, how many are needed for the summer CSA, and how many need to be replaced/ordered per season.

In a future post, I’ll do research and a detailed breakdown of the variables involved in calculating the total cost and projected revenue from the chickens. They also are moved around in a mobile chicken coop across various parts of the field as high-nitrogen fertilizer throughout the year.

During summer CSA, the chickens should be laying at least enough for a dozen in a carton per full member-share equivalency bag. We should always have more to sell extra and ad-hoc at farmers markets. In general, farm-fresh eggs have deeper yellow yolks and while not exactly the same size or color like the boring ones in the store, the uniqueness is also a selling point.

The eggs get washed not simply because it is the law in Massachusetts (General Law Part 1 Title XV Chapter 94 Section 90A), but because customers deserve high quality eggs that look clean and are inspected for quality. Careful attention to eggs is important because come cracks are hard to identify if working too quickly.

Seeding and Potting Up

Back at the greenhouse, we seeded multiple varieties of lettuce and some basil, then potted up pepper starts from trench trays to 2 inch cell trays. For the lettuces, we used 98 cell (7x14) plastic trays with mesh flats underneath. On a large table at the front of the greenhouse (8x4 foot plywood sheets on saw horses), potting mix is mounded up. Empty trays get handfuls of the mix on them and brushed with the hands until all cells are full but not compacted. With all fingers in a horizontal row, divots are made in the cells from far to near a few times to make space for seeds. Many of these trays can be done in preparation for the seed stage. The basil seeds went in small 2 inch “paper pots” and will grow that way in the greenhouse until large enough to put in CSA member shares.

The hardest part of this work is on the neck…when looking down at one’s hands for a few hours, between egg washing and seedling trays…the muscles in my back between my shoulder blades are sore today. At least my lower back got a break from field planting which we’ll likely do tomorrow.

Tech Note: Context is Everything with LLMs

I realized that the outcomes I need to speed these blog posts are A) activities performed, B) new things learned or encountered, C) questions I might ask in the future about these areas of work, and C) what to remember tomorrow.

It’s a hobby system and there’s plenty to tweak on the weekends.

Third day of my first week and all the areas of my back are hurting. Not complaining, just realizing that doing this full-time for decades does things to a body. No wonder some farmers say that chiropractic care is so important.

Kale Fields Forever

Today we started in the field that was used for peppers at the end of last season, hilled rows covered with black weed fabric. We set up low tunnel hoops, ~3mm thick row cover support wire, first by taking bunches of them over shoulder and dropping them over the rows about 10-15 feet apart. This was the easy part. Then down 6 rows about 300 ft long, straddling the hilled row and pushing one end of the hoop at a time into each side, we distributed about 180 of them throughout.

This field will be used for kale and doing setup early like this helps us know how much floating row cover survived from last year intact enough to use and how much more needs to be bought. This was from 8-10:30am, then we transitioned to cultivating the planting holes in the fabric by sticking a spear trowel down in and turning a bit, which will loosen the soil for the kale to take root well.

Potato Planting

Because the weather was so nice, we got pulled to help plant a row of potatoes in the “dry” field near the garlic we planted last fall. In the past, I’ve bought potatoes in a supermarket and never thought about how they were planted, maintained, harvested and distributed. In prior years, I’ve received some potatoes in CSA shares, often buying them from farm stands as a “specialty commodity item”, never realizing how much work it takes to bring them to my table. Never again will I take for granted a potato, especially the freshly harvested kind.

Monty Don has a great episode of “Fork to Fork” on potatoes (can’t find it on youtube), where the gist is: though versatile, potatoes are the best when you harvest them the hour (not day) you plan to consume them. Boil them in salty water, drain, then let sit with olive oil and fresh mint covered with a towel until cool enough to handle. Serve with salt. That is all.

Dry Farming by the Ocean

This field is far into the property and looks out at the Crane estate across the Ipswich river and a snake-looking estuary which keeps the surrounding areas from drying out too much. “Dry farming” is where there is no irrigation other than what nature provides. The soil in this area is not sandy, not quite clay. Because it is a sort of flowing, undulating field with wetland around it, it doesn’t need much cultivation by a tractor to become workable with the hands.

Lunch and Chicken Coop Construction

Then we had lunch. Tuna salad on potato bread. I added thin slices of apple and chips for some extra texture. The property handyman who was working on a new mobile chicken coop joined us for a bite, then we helped him get plywood sheets up on the top of the framed roof.

The ‘mobile’ part of the chicken coop is a lean-to style roof, about 6-8 foot high, sat on a small dairy trailer suspension system. It has a door on the front and spaces for windows that are screened and can be closed up with a hinged flap as needed. This season, the 300 chickens now will be doubled to 600. Two of us helped to hoist the plywood to the builder on the roof.

Back Down to the Ground

Then we went back to potatoes, finished up 4 more rows 300 feet long, the first two fingerlings and the last three red varieties. The process of hand-planting potatoes goes as such:

Make a row, usually with the tractor and an attachment specifically to cultivate 4 inches down, leaving a 2-in trench Amend the trench with whatever is needed based on soil tests and intended crop needs; in this case, it was rock phosphate and Pro Gro 5-2-3 fertilizer (chicken bedding product) Fill 5-gallon buckets up with seed potatoes, chunks specifically grown late in the prior season to chop and store for planting, then “drop” the chunks in the trench about 8-10 inches apart By hand, go down the row and push the seed potatoes down into the soil, then cover over with soil…”scratch it up” in this field to exfoliate excess grass roots and clumps to the top and deposit them in the walking area of the row Repeat…thousands of times…400-450 times per row, 5 rows today.

I’ll be honest, after doing the first fingerlings and knowing what the rest of the day held, I took lunch to mentally prepare for the back-pain that the rest of the afternoon would mean this weekend. Fortunately, the crew already knew what I was only realizing then, and offered ample opportunity to take the “drop” option until there was nothing else to do but “plant”.

Planting in even a pre-cultivated field is not easy. Not all parts of the row are the exact same soil type and compaction towards the end of the rows is especially problematic. Also the tractor cultivation attachment depth really needs to be dialed in, otherwise you then have to “cultivate” by hand each potato you push down.

Tech Note: Context Templates and Changes

Main Themes:

1. Field work and manual labor in farm settings
2. Equipment usage and operation in farming
3. Crop management and cultivation techniques
4. Organic farming practices and certifications
5. Business operations and customer relationships
6. Challenges of working with grants and specific conditions

Activities:

• Setting up row covers, planting crops, and cultivating soil
• Assisting with building a chicken coop, planting potatoes, and tussling weeds
• Converse with colleagues about various topics related to farming and business operations
• Reviewing and updating soil samples, setting up garden plots, and navigating grant requirements

New Things Not Yet Encountered:

• Using metal hoops for row cover
• Working with tractors and attachments
• Planting potatoes in a field setting
• Rock phosphate fertilizer
• ProGro fertilizer

Questions and Future Areas of Research:

1. Why are there different types of fabrics used for row covers?
2. How do you measure the size of rows on a farm?
3. What is the optimal spacing between fabric spikes to ensure proper row cover coverage?
4. How does one deal with running out of rocks or other materials needed for row cover maintenance?
5. How to estimate the number of potatoes dropped during planting
6. The process of applying for grants, specifically federal grants from the USDA
7. How to balance the cost of organic certifications with business operations
8. How to navigate matching specific grant requirements with a farm business
9. What are some best practices for grant writing, particularly in relation to agricultural projects?

Suggested Actions:

1. Measure rows using Google Maps to improve accuracy
2. Invest in more fabric spikes and spare parts for equipment to avoid shortages
3. Consider creating a system for tracking crop growth, soil health, and weather patterns to inform future farming decisions
4. Develop strategies for dealing with physical labor fatigue and back pain in farm work
5. Review and update soil samples to ensure they are receiving adequate phosphate
6. Consider alternative fertilizers or methods that may be more cost-effective
7. Explore grant opportunities and research ways to apply for federal funding
8. Seek guidance from experienced farmers or grant writers on how to navigate complex grant requirements
9. Research and explore non-profit organizations like The Carrot Project that provide resources and support for grant writing and agricultural development

Second week of farm work. Imposter syndrome kicked in. The work is getting a little easier, but all it takes is a little rain to complicate things.

Overalls vs. Pants

I have a hard time releasing old jeans to their next life. By that time, they’re comfortable but ragged, stained and ripped…let’s just say, not suitable to wear on a date. By this time, jeans like this are perfect for work on the farm. You can also find great work clothes at your local thrift store.

Pants usually need a belt or suspenders to stay up while you’re bending over, squatting, carrying things, etc. I tend to tuck a long t-shirt in to my pants and then wear a hoodie or something over the shirt in the colder season. This keeps most dirt out of my unders.

Today I tried a pair of Carhartt Bib Overalls. They were $50.00 at Nelson’s a while back because they were the lesser desired Navy Blue color. In the summer time, dark clothes might be a problem as they absorb sunlight and heat up more than light clothes, but we’ll cross that bridget when we come to it.

Rain’s Blessing and Curse

In New England, you can never really predict what the moment-to-moment weather will be very accurately. The plethora of micro-climates means that you may get wildly different conditions and readings than your local weather station, even if it’s less than a mile away.

This morning, almost exactly at 8am on the dot, it downpoured for 12 minutes (where I was). Back where I live, the power went out for an hour. In the greenhouse, we took the time to seed scallions which get planted and harvested whole to include in the CSA shares each week (which is why there are so many all the time).

[CSA member note: how-to’s on easy processing of your share when you get it home…scallions go in a small container with a little water at the roots. They keep growing this way rather than rotting in the fridge.]

The minor downpour resulted in patches of deep muck in the cultivated fields when we finally got out there to plant. Sometimes even just walking through this mud risks losing a boot in it, not to mention being down on your hands and knees planting the dropped seedlings into the soil.

Farming Crews Are Hard Workers

Still, it’s the early/rainy season and there are ways to work around most of the usual surprises nature throws our way. And the crew is hard-working, especially the women, especially the mom. Moms are superheros…they deserve every good thing that comes their way. I’m grateful that these folks understand how I’m on an uphill hike already to get in better shape. I’ve already lost a few pounds, but so many more to go.

The work is repetitive, often requiring lots of physical endurance, but many chances to switch tasks of varying demand on different muscle groups. Others on the crew go to the gym regularly, especially on a day off, to maintain endurance so it’s not as hard when they get back to work.

As a gardner, I never planted dozens of trays of things. At most, I’d plant a few raised beds for others and call it a day. Dropping, planting and row-covering 6 rows that are 300 feet long is a whole different scale…and this is ‘small’ farming. Having a crew, even a few other people, to help makes a huge difference. While doing the fennel today, one member started us all on a “if we had a TV show, what would go into it”, which was great to pass the time. Sometimes, music from a pocket device is okay, but so far I prefer conversation and quiet times listening to nature.

As usual, I took audio notes of what I did during the day on my way back home. These then get put in a Google Drive folder, picked up by my homegrown transcription and LLM services, and summarized into the below:

Main Themes:

• Farm work: The intern’s day was filled with planting various crops in the greenhouse and field, including scallions, bok choy, fennel, and pak choy.
• Challenges: The intern faced challenges such as heavy trays, difficult seed germination, handling pests like flea beetles, and adapting to unexpected situations.
• Improvisation: The intern had to adapt to unexpected situations, such as a sudden rain shower and running out of row cover material.

Activities:

• Planting various crops in the greenhouse and field (scallions, bok choy, fennel, pak choy)
• Practicing planting efficiency and consistency
• Patching up torn areas of row cover and applying new material to hoops

Specific Tasks:

1. Seeded 12 trays of scallions by hand into 98 cell trays in the greenhouse.
2. Took bok choy plants to the field and planted 24 trays of them in two rows with spacing of one foot between each plant and three feet across the row.
3. Played with row cover, patching up torn areas and applying new material to hoops.
4. Planted a couple dozen trays of fennel and pak choy in adjacent rows.
5. Seeded 40-48 trays of pak choy, scallions, and fennel using the same methods as before.

New Things Not Yet Encountered:

• Flea beetles: The intern mentioned that row cover is essential to prevent flea beetle infestations at this time of season.
• Seed germination issues: Some bok choy seeds did not germinate in the trays, and the intern is looking for ways to improve seed germination rates.

Questions and Future Areas of Research:

• What are some effective methods to improve seed germination rates?
• How can the intern optimize planting efficiency while maintaining consistency in spacing and placement?
• Are there any other pests or challenges specific to bok choy and pak choy that the intern should be aware of?

Suggested Actions:

• Practice dropping plants consistently to develop muscle memory and improve efficiency.
• Experiment with different row cover materials and designs to optimize their effectiveness.
• Research methods to improve seed germination rates, such as adjusting soil moisture or using specific seed treatments.
• Consult with experienced farmers or researchers to gain insights on flea beetle management and other pest control strategies.

Beautiful weather for planting today. Six rows of dinosaur kale, red kale, and swiss chard into black fabric covered field.

Keeping the Greenhouse Watered

The greenhouse is full of well-off scallions, onions, lettuces, kales, chard, chois and early starts of peppers, tomatoes, and various herbs. Even in April, this healthy situation doesn’t last long without plenty of watering. This morning I spent about 45 minutes carefully watering everything…the more mature starts with the shower setting and smaller or not yet germinated seeds with the mist setting. There are tray types for just about every stage of growing.

Even with the overcast day we had, I was asked to go back and spot water anything that looked dry. This included the row channel trays on the heat mats as well as anything in a paper pot (peppers) for inclusion in a future CSA share. The greenhouse stays very humid and between 70-80 Farenheit.

Testing Out the Trencher in the New Greenhouse

After I was done watering, we rode to gather some rebar to use as line-measuring poles in the new greenhouse setup for upcoming tomatoes. I pounded in the rebar in predetermined locations on both (door) ends) inside the greenhouse so that we could tie twine between them all the length down as guidelines for trenches. Then we sprinkled cheap baking flower over the lines and moved the twine from pole to pole and sprinkled until three pairs of two trench markings were down.

The trencher (TRX-250 Walk-Behind) looks like a chainsaw on steroids. You stand on the back of the engine part, lower the chainsaw digging part to the desired depth, and drive backward while it chews up the soil to make a 2-3 inch trench, in our case about 4-6 inches deep. This is where the tomatoes will be planted all the way down, trained up twine hanging from the cross-bars as most greenhouse tomatoes are grown.

This machine is also used to dig trenches for the electrical conduit being laid to support the greenhouse power needs, such as the air compressor to create a bubble of air between two poly sheet layers over the greenhouse…making it act like insulation and keeping it just tight enough to the form to resist wind gusts.

Planting, Take One: Dinosaur Kale

Once the greenhouse assistance was done, we loaded up the big trick with about 25 trays of various kales and chard. Before lunch break, we got about two rows of dinosaur kale planted into fabric holes which were already cultivated on Monday. As we were doing that, some other helping hands cultivated rows 3 and 4 in prep for red kale.

Lunch

Lunch was sauté of chicken breast, onions, and cheese. Had it on rolls with mayo and pickles. Easy, brilliant. Helps with the look-ahead to an afternoon of field work.

Planting, Take Two: Red Kale and Chard

After lunch we planted the red kale, then we cultivated the last two rows before dropping and planting the chard into them. This all took the rest of the afternoon, a little less than 4 hours.

The Summary from an LLM’s Perspective

Spent equal time today between seeding in the greenhouse and planting in the field. We planted napa cabbage and more scallions. We moved the onions outdoors to harden off. We seeded lettuces. We sang in the field.

Dirt is not Soil is not Compost is not Potting Mix

I’ve been gardening for 15 years, but I still occasionally transpose the words ‘soil’ and ‘dirt’, but they are not the same. Compost is not potting mix, and dirt is not soil. Getting the words right isn’t pedantic, it’s specific, and as an intern I am specifically here to learn the right things, not exercise innacuracy.

Usually, people say “soil” when referring to the life/liveliness/ecology in it, when they have cultivated it by hand or mechanical means in preparation for planting, or when they plan to adjust “dirt” to become suitable for growing things in.

Compost is human-made organic matter that has broken down to the point where it is no longer in it’s original raw inputs form. In large commercial setups, much of the input is leaves and grass clippings, but unfortunately as such come with unknown quantities of pesticides, herbicides, and fungicides that are perstistent even through hot (thermophilic) composting processes. You really have to investigate what your local composting options, producers, and their processes entail to be sure that this isn’t happening by default.

At home, I ‘vermicompost’ whatever can go to my small stack of composting worms (only veg) because we don’t have the space for a compost pile. I’ll do a bunch of posts later about why every gardener should have a worm bin (or 3), but sufficed to say that their poop is garden gold and especially so when brewed in compost tea. For the rest of our kitchen scraps, we subscribe to the Black Earth Compost service and chuck anything that would rot if with the worms but is still compostable (meat, bones, prepared foods, eggs) into the Green Bucket that they haul away at 6am on Monday mornings.

Finally, potting mix is often a combination of ingredients…finished compost, a well-draining moisture retaining component such as peat moss or coconut coir, and possibly other elements like perlite or vermiculite to provide moisture and drainage control. The commercial potting mix we use at the farm is from Vermont and comes in a huge cubic yard bag on a pallet. We shovel buckets of this from the big bag near the greenhouse and deposit them onto the potting table to use with the seed trays. When dry, the mix is light and fluffy and easy to work with. When wet, it is ‘friable’ (it clumps together when you squeeze it) but still drains very well.

Idle Hands and Standing Around

When I was a kid, I worked hard and there were lots of chores to keep my hands busy. With a job, ‘idle hands’ generally is not what people pay for. Since I grew up in the 80s, the first half of my life did not include a mobile phone, so I also consider non-job usage of devices ‘idle hands’.

Standing around isn’t always a bad thing so long as it short-term and you’re taking a moment to observe and think about what’s next. But standing around waiting for the next instruction is just not good. As my kids start to think about getting jobs of their own, I hope that some of this work ethic has transferred to them.

On the farm, there’s always too much to do and not enough time. Though the crew that I’m with this year is kind and it’s not rush season yet, I wouldn’t call any of the work we do ‘laid back’. My mode right now is to do the task aassigned, then fill gaps with things I know have to be done and are small enough to put down at a moments notice.

So when I get there a few minutes early in the morning and I see that the potting table has no mix mound on it, I won’t want to be told this has to be done. I know it needs to be done because we seed every day. Same with cleaning out the used trays…we’ll need them for something soon so better to use the small gaps in between big jobs for things like this.

Preparing Trays for Seeding

We’re always starting things in trays, every day in this early season. Today it was four different kinds of lettuces, some very small and bare seeds, some pelleted.

The process goes:

1. Get a clean cell tray (should be plenty ready to go) with a web/mesh tray underneath for support
2. Lump handfuls of potting mix onto the tray and kind of rub over the whole tray with your hands until mix fills (but not tightly) each cell
3. Lightly bang the tray on the table a few times and add a pinch of mix to any cells that need it. Preferrably, the mix should come up to about 1mm below the cell wall top.
4. With the fingers from both hands lined up horizontally, press small divots into each cell, starting at the top and working downward. For a 96-cell tray and four fingers on both hands, this means 8 divots at a time down 6 vertical rows, then repeat with the rest of the 8 columns left.
5. Repeat as many times as needed, stacking atop each other to about 7-8 trays high max. Stack gently so as not to erase the divots. Over-filling the cells will also make it easier to erase the divots because there will be an excess of mix rubbing around between tray stacks.

It’s not complicated, but a little attention to detail makes the next step easier/faster, as most things on the farm go.

Singing in the Field

One of the crew, kind of a right hand to Jamie, is a singer-songwriter and guitarist. We all talk about a lot of things while working, about our kids and family, about past experiences, recipes, favorite entertainment and local restaurants…and sometimes sharing my background leads to details about how I was raised singing in church and lots of music education.

I’ve played the guitar since I was 12 and mostly in high school and college, nothing serious or professional, but its nice to pick up and play along with my kids when there’s an opportunity. I’ve never been more than okay at guitar but I have a natural ear for harmony and a brain for music theory. I mentioned that I prefer to sing alternative harmonic parts that include things people sometime think sounds weird, such as adding a 9th to a typical 1-3-5 or 1-4-6 chord (very Ben Folds though).

So she says “let me hear, let’s sing something”. Of all dumb things I can think of, I offer ‘Go tell it on the mountain’ as something we both would know. We sang through a couple of passes, some with a traditional harmony and then with alternatives. It was nice to share something unexpected and things like this make the work go by faster.

Main Themes:

1. Safety and Prevention: The intern highlights the importance of wearing gloves when handling wires to prevent blisters.
2. Personal Preparation and Self-Care: The intern realizes the need for a toiletries bag with essential items, such as soap, lotion, and a nail brush, to maintain personal hygiene and comfort during work hours.
3. Teamwork and collaboration: The intern highlights the importance of working together with other hands, particularly one who is experienced and skilled in farming.
4. Adaptation to new tasks: The intern is learning new skills and adapting to different activities, such as planting lettuce and cabbage.
5. Attention to detail: The intern emphasizes their attention to detail and precision in planting and maintaining the garden.

Activities Done by the Intern:

1. Filled buckets of compost mix
2. Prepared seeds for transplanting (lettuce)
3. Transplanted scallions from greenhouse to dry field
4. Hand-planted scallions with help from experienced hand
5. Loaded Napa cabbage onto truck and planted it in farthest field towards the ocean
6. Prepared trays in greenhouse
7. Cleaned up and repaired row cover
8. Driving wire hoops into the soil using bare hands
9. Seeding lettuce in the greenhouse
10. Potting up kales in little paper pots

New Things Not Yet Encountered:

1. Handling wires without gloves, which led to a blister on their middle finger and palm
2. The need for a toiletries bag with specific items (fingernail scrub brush, sponge, soap, lotion)
3. The importance of having a spare pair of glasses in case of an emergency
4. Seeding lettuce (although the intern had previously done so, they were emphasizing their experience with it)
5. Planting Napa cabbage using a staggered approach (diagonal planting pattern)

Questions and Future Areas of Research:

1. How to prevent blisters when handling wires without gloves?
2. What are the best types of soap and lotion for maintaining hand health between work days?

Suggested Actions:

1. The intern could benefit from practicing faster planting techniques to improve efficiency.
2. Further training on specific crops, such as lettuce and Napa cabbage, might be helpful for the intern.
3. Consider exploring different soil mix compositions or composting methods to optimize plant growth.
4. Consider investing in a toiletries bag with essential items for future workdays.
5. Look into alternative methods or tools that can reduce the risk of blisters when handling wires.
6. Research recommended products (soap, lotion, fingernail scrub brush) for maintaining hand health and comfort during work hours.

We planted 6 rows of kohlrabi, broccoli, and leeks. I watered the greenhouse mid-day. Then we hoed 3 rows of scallions in the dry field.

Tomatoes Grow Up Quickly

To start, we potted up a bunch of tomato plants that were getting a bit leggy for 2-inch cells. These were started a few weeks ago and already need 6-inch pots. Eventually they’ll be planted into the trenches in the new greenhouse…which were worked on a bit today with the trencher.

Great Planting Conditions

The soil in the lower field had been freshly cultivated with very little compaction, so the job of planting two rows of kohlrabi and two rows of baby broccoli was easy on the hands. Still, planting by hand takes a toll on the back and knees, but for me it’s getting easier to go faster with less unnecessary strain.

Leeks Are Delicate Yet Hardy

Leeks are allums, so even if grown from seed they are dainty looking sprigs of green, they can easily be extracted from seed tray cells after watering and pulled apart during planting. Unlike other veg rows, the leeks are spread far apart…one on each side of a 4-foot bed about 3 feet apart and 8-10 inches between the rows. This gives them plenty of space for their greens to branch out, for the stalks to get big and beautiful, unblanched by the sun and still provide enough space to run a small manual cultivator between them during the season.

Don’t Begrudge the Kohlrabi

I don’t particularly care for kohlrabi, as neither do many consumers. It’s a turnip cabbage with all the pros and cons of both combined. So when we planted rows of it, all I wanted to do was go to my happy place and get it done quickly. But it is easy to grow, hardy for early season, and I’m sure there are ways to cook (with) it such that it isn’t just a culinary chore in your CSA. At least it comes with both stem and leaves, two textures in one, so maybe that’s positive.

Hoeing Around

After lunch, we planted the leeks then I watered the greenhouse because it was a relatively warm and sunny morning. Then we went out to the dry fields and used stirrup hoes around the tiny (but still living) scallion starts to terminate any grass weeds that were re-establishing. Weeding always seems like a boring chore to me, but it has to be done and there’s plenty to chat about while carefully working around the plants.

Ollama, My LLMa Summary

We used the rest of a hay roll and then collected salt marsh hay to finish mulching the leeks. I had my first hand at washing all the eggs by myself.

Mulching with ‘Hay’ (straw)

After watering the greenhouse and some other morning tasks, we headed over to the dry field and spot-mulched the garlic where there were little patches of grass peaking through. Hopefully this will keep the grass from spreading but the garlic is established enough to push through. Then the rest (about a half) of the big roll was loaded onto the Ford by a John Deere 1050 wheel loader tractor which maintains permanent residence near those fields.

The ~hay roll (really straw, since hay often has seeds) was brought over to the lowest main field to mulch the leeks, but we ran out and had to use some old left over salt marsh hay near that row. Then we hooked up a small wagon to the Gator and went to collect dead grass from the salt marshes nearby for the rest of the row.

Salt marsh hay usually costs about $12-15 per square bale unless you live close to a producer and pick it up yourself (~$10). The four of us rode out and collected about 5 times that in 30mins, scuffing it up with our boots and picking it up by hand. Doing the math, this was about $34 of labor but in a pinch and as-needed to finish up a row and protect crops, is a fair trade-off. I think it’s easier to use pre-bundled straw, but when it’s on back-order and you still have to get things done, necessity is the mother of invention.

Mulching baby leeks is a delicate thing. No prevent against unnecessary weeding later in the season, we were aiming for about 4 inches thick cover. Six inches from each edge which is the same inset depth as where we planted the leeks and then the middle of the row, approximately 3 feet wide and all the way down…basically the whole row at 4 feet x 200 feet x 4 inches, or 1/3rd of 800 feet, we used approximately 275 square feet of hay on one row. Generally a square bale covers 70 square feet at a 4 inch thickness, so one row takes about 4-6 bales and some loss is expected when transfering and applying.

Cleanup

Then we took about an hour to clean up various piles of used tarp, drip line, and row cover no longer suitable for use. We are still waiting on an order which includes new semi-permiable floating row cover since there was high wind last week and some of the kale and lettuces were affected. With the black tarp, we unfurled each piece until we could assess if it was worth keeping, and also to pull any leftover garden staples out. We reused the big, pallet-sized yard-cubic bags that our potting mix comes from Vermont in for stuffing the trash items into. We consolidated bags of rocks into piles close to various beds we plan to plant into soon. Then we broke for lunch.

Egg Washing and Packaging

I was put on the task of processing the eggs today. I had carefully watched and assisted the process two weeks ago. Jamie gave us a refresher, but the process is pretty straight forward with some very specific food safety rules. Namely no cross-contamination, and at the end of the process, all packaged eggs should be flaw-free.

We use air pump bubblers in 5-gallon buckets with hot water (not scalding, we don’t want to cook them) and detergent in the ‘wash’ cycle for 15 minutes. The wash water is dumped then refilled with fresh water for a ‘rinse’ cycle for 5 minutes. At least for now, I am using a timer and learning to keep my eye on a clock. Between waiting for eggs in baskets to each go through these cycles, hand finishing the eggs that have been washed is a process of making sure that all eggs are blemish-free (no remaining dirt or surface matter, no smudges, smooth and clean to the eye). This is tedious but not hard and takes a long time, but results in rows of cartons of beautiful, farm-fresh eggs waiting to make customers happy.

I processed sixty (60) dozen, with a flaw count of about four dozen. On a good day, there’s always flawed eggs, say about two dozen or so. I think the only difference in my handling is that I’m still getting used to how full the baskets should be and how slowly they need to be lowered into the hot water. Other than that, and maybe upstream handling process, there’s few other factors I can see which results in higher number of cracked eggs at the end. All told, this took about three and a half hours just me, including cleanup and barn shutdown again.

We sell the eggs for $8 per dozen. It cost about $60 of labor and produces about 60-64 dozen twice per week with the 300 chickens producing now, and there’s more to come in order to produce enough to fill CSA and upcoming store orders. So it’s about an 7-8x value task. And people really love their eggs.

Below is the summary of my transcribed audio notes which includes a fair bit of detail about the salt marsh hay collection and egg washing. Also note that it completely skipped the leek mulching and hay collection. Next time someone says “I think AI will help farms become far more efficient”, resist the urge to slap their face, and show them this blog where even a well-tuned and micro-managed LLM has a very hard time getting details about real field work right.

Main Themes:

• Farming and agricultural work (including egg collection, washing, and polishing)
• Internship experience and responsibilities
• Daily tasks and activities involved in farm management (e.g. weeding, watering, cleaning)

Activities:

• Egg collection and washing
• Hand polishing and hand washing eggs with a damp towel
• Counting and loading eggs into totes
• Trimming egg cartons
• Weeding and preparing soil for planting
• Watering greenhouse plants
• Cleaning trash and debris from the field

New Things Not Yet Encountered:

• Special detergent used in the washing cycle
• Bubblers at the bottom of the five-gallon buckets for hot water
• Egg polishing technique using a damp towel
• Rotating eggs between wash, rinse, and dry cycles
• Using large totes to store egg cartons

Questions and Future Areas of Research:

• Improving egg washing and polishing techniques without overfilling baskets
• Preventing nail problems caused by manual labor in farming
• Optimal conditions for greenhouse plants (temperature, humidity, light exposure)
• Alternative methods for egg cleaning and polishing that reduce waste or improve results

Suggested Actions:

• Develop a more efficient system for rotating eggs between wash, rinse, and dry cycles
• Improve hand care techniques to prevent nail problems from manual labor
• Research alternative methods for egg cleaning and polishing
• Investigate optimal conditions for greenhouse plants
• Track egg quality and freshness to ensure consistent production

We did a lot of shorter tasks while waiting for a shipment that included row cover and irrigation components.

There’s Always Something to Do

Even, maybe especially, while you’re ‘waiting for a shipment to arrive’, there’s always at least a dozen shorter tasks around a farm that can take up the time and prepare for whatever is coming next. For instance, today I:

• Mulched more leeks, like yesterday
• Collected salt marsh ‘hay’ (straw), like yesterday
• Filled greenhouse trenches with compost
• Migrated fertilizer bags to the greenhouse
• Cut and laid out drip irrigation lines in the greenhouse
• Watered and collected eggs from the chickens
• Baited the electric fences with peanut butter (trains the deer)
• Seeded radicchio and other lettuces (about a dozen)
• Helped load the shipment off the truck and on to the Gator trailer
• Brought various items from shipment to parts of the property
• Learned about using a “Gro-Trak” attachment to lay down biodegradeable mulch

More Progress on the New Greenhouse

The trenches had been dug since Tuesday and now it was time to fill them with compost and a sprinkle of fertilizer. Lugging dozens of buckets of soil was quite the exercise, but fortunately it was 60°F and even in the greenhouse it wasn’t too hot.

Then we cut and laid drip irrigation hose by having the spool at one end, holding the end and walking it down to the other end, then dropping it there and walking back to the spool end to cut and start again. This type of 1-inch tube came pre-perforated, just enough to emit water every 6 inches or so at the right flow rate. We needed 16 lines because each of the 8 trenches needed two lines on either side of a trench. At the far end, each line was tied with three knots in order to terminate the flow. Others on the crew then connected these lines to a 2-inch tube used as a header to equally distribute the water down the 1-inch lines.

Chicken Assistance

Then a few of us picked up 5-gallon containers of water, some hay, the egg buckets, and some feed to bring down to the chickens. The mobile chicken huts are currently at the very end of the field that the leeks are in, so we had plenty of hen chatter to listen to for the past week.

We collected about 20 dozen from ~300 birds in one day. This is why after about 3 days, we get around 60-65 dozen for washing. Then we ran the eggs up to the barn to store in the walk-in fridge.

Seeding Radicchio

Then we seeded a dozen or so trays up. The variety of Radicchio seeds were small (not tiny) and look like little fluted cones, bright yellow, and easy to see. Some seeds are tiny, round, and dark…making the seeding process for a fat-fingered guy like me rather difficult. There are ways to handle tiny seeds, such as putting them on a plastic cover with a small cut to allow only a few seeds to escape slowly at a time, but it’s best just to get used to ‘pinching’ them between your fingers, then carefully rolling two fingers to let one or two go at a time.

In a later blog post, I’ll share research on devices used to help seeding, but sufficed to say that this farm does many things manually on purpose, leaving only the most onerous tasks to tractor and automation. I respect that the farm GM Jamie has worked with the owners of the property in stages, each year further improving the approach to the labor and its outcomes for the local community. I have many years of observation to go and it’s nice to get some exercise anyway.

Deer Training and Fences

After lunch, I was put on the leisurely task of ‘baiting’ the electric fence that goes all the way around the main plots with peanut butter. For each section of the fence between poles, I smudged a small bit of peanut butter with my fingers onto the wire. This attracts deer who are not yet trained to lick the fence and get a minor shock. I have gotten shocked a handful of times myself and, while it’s not pleasant, it’s completely harmless. This effectively trains the deer to rightly think that the vegetables are off limits. For about 2 acres, this took about 30 minutes.

The Shipment Finally Came

Ever since the old row cover blew off a few of the rows over the weekend, we were waiting for a shipment with new row cover, not to mention an assortment of other necessary things. This included two new water pumps (for the greenhouse and replacement of the old one), a bunch of row cover, biodegradable black mulch sheets, drip irrigation, garden fabric staples, and a few other things. About a cubic pallet’s worth, we loaded off the A. Duie Pyle truck and on to the trailer hooked up to the Gator. Then we drove it around to various drop-off locations around the property. This took about an hour or so.

Gro-Track Mulch Sheeting

The last task of the day wasn’t even really a ‘task’, but more an educational session. Now with shipment arrived, Jamie drove two of us over to the dry field next to the garlic where we will plant onions next week. He went over the basics of using a tractor attachment called the Gro-Trak system which effectively ‘paints’ black biodegradable sheet over rows for planting. This sheet is so thin that his little terrier dog could easily perforate holes in it with a paw.

The Gro-Trak attachment connects to a standard 3-point hitch (hydraulics) with an additional wire connection for auto-correction and better row alignment. The sheet roll of mulch sits on rollers and feeds down through to the back wheels of the Gro-Trak which also has discs to throw dirt onto the edges to keep the sheet pinned to the ground. Doing something like this manually would take a ton more effort, so you can see why tractors are vital to farming, small-scale or large.

End of Week, My Body Hurts

As usual with an out-of-shape ex-tech worker, even just three full days in a row makes me hurt in places I didn’t even know I had on my body. Mostly my sciatic nerve and lower back, but also my legs and shoulders, could use a little break. Too bad for them, my own personal garden needs love and attention this weekend. Fortunately I have some young helpers of my own too and family time comes in all forms.

For now, this is what my LLM thought I did today:

Main Themes:

• Farming activities (farming, greenhouse maintenance, chicken care)
• Field work and farm maintenance (equipment usage and maintenance, animal training)
• Equipment and technology integration (drip irrigation system, electric fence)

Activities:

• Loaded compost into trailer
• Filled 5-gallon buckets with compost
• Shovelled compost from back of trailer into buckets
• Poured compost into trenches in greenhouse
• Applied fertilizer to trenches
• Rolled out drip tape and irrigation system in greenhouse
• Connected drip tape hoses to header
• Tied ends of drip tape hoses with knots
• Visited chickens and collected eggs
• Applied peanut butter to electric fence to train deer
• Seeded radicchio and other vegetables in greenhouse
• Laid down biodegradable black plastic mulch

New Things Not Yet Encountered:

• Greenhouse trenches
• Low fertilizer used in greenhouse (5-3-2 pro grow)
• Drip tape irrigation system
• Biodegradable black plastic mulch
• Vegetable plastic
• Rain Flow attachment for tractor
• New pump for forced water irrigation

Questions and Future Areas of Research:

• Optimal pressure settings for new pump
• Deer behavior and training methods using peanut butter-based methods
• Alternative biodegradable mulch options
• Effectiveness of drip tape irrigation system

Suggested Actions:

• Provide additional training or guidance on greenhouse maintenance and irrigation system setup
• Review fertilizer and amendment schedule to ensure proper application rates and timing
• Conduct thorough audit of drip tape irrigation system to optimize performance and reduce waste
• Encourage intern to ask questions and seek clarification on unclear or unfamiliar aspects of farm operations

We planted 3000 onion seedlings in one of the dry fields covered in biodegradable mulch film and then covered 4 field rows with floating row cover, shoveled in on the sides to protect from flea beetles.

New Gear

The days are hotter and sunnier now, today was 75°F by noon though the ocean breezes were helpful. Also helpful, after getting a bit sunburnt especially on the back of the neck, I got wise and ordered a few long-sleeve breathable hooded fishing shirts which worked out great today. I also ordered the makings of a toiletries kit such as a nail brush and two small sealable travel bottles to hold orange soap and hand lotion (half lotion, half aloe vera). Notwithstanding the shirts, I wasn’t too sweaty but need to dig out my sports-fit undies because boxer shorts are chaif-inducing in the long term.

Onion Planting

After watering the greenhouse from 8-9am, I joined up with the other 2 on the crew to drop and plant 30 98-cell trays of onions that have been growing and being trimmed for a a few weeks. We planted them in to the two (and a half) dry fields that were covered with the black biodegradable mulch film last week. This was next to last year’s garlic which is coming up nicely and that we spot-mulched with hay last week too.

This took about 3 hours and 3 people to plant around 3000 onion starts. The dropping process is more like placing since the onions are easy to lose their plug potting soil and the film has grid lines in 5 inch squares to make spacing easier. The spacing for these onions were 6-8 inches between rows of 3 also similarly spaced.

After this, my knees and leg muscles as well as my lower back were ready for a break. Lots of bending and stretching to get it done.

Lunch

Ice cubes in my water bottle are a winner. Also I plan to add a box of Crystal Light powder packets because I went through about 4 liters of water today just sipping between tasks and plain water seems to just leave me sloshy in the belly unlike flavored waters.

Tomorrow I bring a loaf of hearth bread for tuna salad sandwiches so I was planning to get to making that as soon as I got home. But when I get back from soccer chaufer duties at 6:30pm, the dough was already on its first rise thanks to my lovely partner who did me a kindness by starting it because I had taken the soccer drop-off and pick-up. So long as I remember to actually bring it this time…which I forgot last week.

Side Note: Remembering Things to Take

Today I was keeping a mental list of things that I could add to the things that travel with my car to the farm. I already have:

• a complete change of all clothing items (socks, undies, long pants, short sleeve shirt, bandana)
• field boots
• a thin rain coat
• toiletries kit

What I still need to bring:

• A wide-brim hat
• A box of Crystal Light packets
• Some small caribeners
• My DJI drone and charging options (stays in the car now that the weather is good)

What I need to buy more of or find in my house:

• Some extra sports-fit undies
• Multi-purpose utility knive (the kind that butterflies out to become a set of pliers)

Okay, back to the afternoon tasks…

Floating Row Cover

After lunch, we took 20 foot rolls of floating row cover and spread them out over 4 rows in fields on the main part of the property. Some of the crops we planted the past few weeks were seeing flea beetle damage and high wind gusts familiar to April in New England had taken some of the leftover row cover we had spread the past few weeks.

The process goes as such:

1. Position the row cover row at one end of the row
2. Grab some over your shoulder and start walking the length of the row (in this case, mostly 200-300 feet)
3. Depending on the direction of the wind, pin down a corner on the end of the row, then the other
4. Pin one edge the long direction with garden staples, fabric spikes, field rocks, or rock bags
5. Pin down the other long edge so that the row cover is mostly tight across the soil or metal hoops
6. For flat soil rows, shovel dirt from the 2 foot walk-space between planting rows onto row cover edges
7. For flat soil rows sibling to other field space not yet planted, use tractor and discs to spare shoveling
8. For hooped rows, ensure that gaps between staples and/or spikes are weighed down by rocks or rock bags
9. Similarly pin down the short open ends so that gusts don’t take up what you did the whole row down

I had to leave at 4pm to handle kid sports transit things on a Tuesday, but the rest of the crew also closed out on row cover duties then too.

AI Takes The Easy (Information Worker) Jobs

I think much of the current tech hiring freezes and layoffs have many determining factors. Like every farm, every tech company makes decisions in their own…unique…way. Unfortunately, many of the hard tasks are still left and information workers aren’t used to really hard intellectual or physical work…except for those that are like my friend Muyiwa.

Olumuyiwa Kolayemi used to be a farmer in Nigeria and now is now a DevOps and AWS Engineer in Boston. He presented a lightning talk in 2023 on the parallels between farming and DevOps. I sat and listened, later I produced the video, and then spent time talking to him about how he went through the Per Scholas Boston IT program to get where he is today. I am doing a bit of the reverse…re-balancing too much tech career with farming.

“AI” can’t make this kind of story up because it doesn’t live in the real world. It doesn’t actually produce the food we eat. It doesn’t go to school to gain skills for you. It doesn’t write from your heart. All it really does is skim the pond of obvious automatable tasks and acts as a buzzword to attract moron funding.

Much like yesterday, we planted more onions in the dry field and then put new row cover on the lowest field where the Napa cabbage and Kohlrabi planted a few weeks ago were getting hit by flea beetle hardest.

More New Gear

The new white, long-sleeve breathable hooded shirts have been working out well over the past few days of 70°F plus weather. My only issue with them is that they don’t have pockets, but all field work really needs is a pare of gloves (for working with equiptment), a spear trowel, maybe a water bottle…all optional for someone who’s been doing this a long time. The more water you sip on a hot day, the more likely you’ll have to take a bathroom break, which I prefer to do either before or after going out to a field to work.

I tried my sports-fit boxer briefs yesterday and that was a win. Also, planting on hands and knees really hurts the knees more than it does the back as your back muscles get used to the exercise and stretching. Not so much with my knees. So I ordered a cheap pair of Fiskar workman’s knee pads to try out next week.

Onion Planting

Today, we planted one row of red onions. This took about 2 hours because the person who’s super-fast at planting is working at the Tulip farm for the next month so her time is split between here and there. Her father was a farmer in China and she been planting since she was young. Strong woman, a mama of 3, not a superhero but pretty darn close IMO. Then we went back to the greenhouse and did a little general cleanup before lunch.

Lunch

Tuna salad with cucumbers and celery on my homemade bread. It’s really nice to share a meal with these people.

More Onion Planting

Then we loaded up the truck with yellow onions and went out to plant one more row. This one went a little faster because this time Jamie was able to join us, so there were 4 people on one row. I like to drop which is generally faster than planting. Even though it’s a lot of bending, it’s not down on the knees. But it all has to get done, so after dropping the whole row, I got to planting.

When working together, it’s okay to work on opposite ends and meet in the middle, for instance on a row…but then you don’t have the option of talking with the rest of the crew, which is something this one likes to do. In this case, we played a “which would you rather” game. Things like “would you rather have to eat everything as pills or as blended smoothies” and other goofy stuff to pass the time. As a psychological distraction, it works too…feels more like a bunch of friends doing something together than the monotinous, laborious and lengthy task it could be otherwise. Helps people get to know each other too.

Row Cover and Flea Beetles

After the onions were planted, we went back to the main lower field where the Napa cabbage had been planted a few weeks ago. It showed pretty consistent damage from flea beetles having not had the benefit of row cover which was late to be delivered.

I worry about how much more complicated agriculture is going to get under the current and future policies and disruptions to the supply chain being imposed from above. It’s almost like people in Washington and who seem to be retained by our political system have never worked a physical job day in their adult life, but most of the voters are that and yet we continue to choose from the best of two worst options when we go to the polls. One day isn’t enough. I think a prerequisite to holding any office in our country at every level should be to have to work one year every seven in a minimum wage job that involves manual labor. At least this would keep people from forgetting the context that they should have when in positions of responsibility and power.

On the very opposite end of the size and food chain perspective, the flea beetle doesn’t give a crap about any of that. It’s everywhere, many varieties, in just about every climate, and over-winters in New England. It’s early to come out and attacks early crops. Of course, pesticides aren’t an option on this farm, as they well shouldn’t be anywhere else. So row cover is our primary deterent, but getting that was late to the planting schedule. Even with the partial damage, the starts will most likely recover quickly since the soil fertility in these fields are high thanks to proper land management over the years by Jamie.

Controlling Flea Beetles

I plan to research this (rapidly) more, but for this season it’s doing what is known to work and taking a small hit.

Google says: “Flea beetle control involves a combination of prevention and targeted treatments. Preventative measures include removing weeds, practicing good garden hygiene, and using row covers. Targeted control methods include using sticky traps, applying insecticides (organic or synthetic), and employing biological control agents like nematodes.”

• Remove weeds
• Good garden hygiene
• Row covers
• Starting seeds indoors
• Trap crops

We’re doing 4 of the above 5 things. I haven’t seen trap crops employed yet and will ask about them, and I don’t see too many ‘targeted control methods’ such as sticky traps, Diatomaceous earth, beneficial nematodes, or Surround WP (kaolin clay). But again, each of these things are micro-management responses to a common challenge.

On that note, the other day I was asking about ‘living pathways’ (growing green cover in the walkways between crop rows) and Jamie mentioned that he had tried it in the past, but for a scale like this becomes yet another thing to micro-manage and that it really has to be dialed in otherwise it competes with crop space and resources (water, sun) especially if not mowed regularly. I think for smaller scale than this (such as a market garden), it might make sense, where most of the cultivation and field work is done without machinery, in farms that employ no-till or in other conditions than here. It gave me something to seriously think about, that there are few ‘best practices’ without first really understanding both the land (ecosystem) and your intention (business or footprint) about it.

Well, sufficed to say, we don’t intend for flea beetles to take more than we can stand, and we now have fresh row cover.

I use ChatGPT every so often to compare certain uses with that of my own LLM and RAG homelab setup, particularly in new areas of my implementation when I need ideas and a not-person to bang against. For now, it’s as easy as having a back-and-forth conversation with the app while I’m driving to work (in 5 minutes). Later I’ll go back over and cherry-pick out (no pun intended) what feels like progress or useful further areas to deep-dive into.

For now, this is what my LLM thinks I did today:

Main Themes:

1. Farm work and daily activities
2. Onion planting and preparation for harvest
3. Maintenance of greenhouse and dry field crops
4. Pest management (flea beetles)

Activities:

1. Watering the greenhouse
2. Loading onions onto the truck and continuing onion planting in the dry field
3. Seeding summer squash and lettuces in the greenhouse
4. Seeding row cover over Napa cabbage to prevent flea beetle damage
5. Digging in row cover along edges of the Napa cabbage patch
6. Replanting kohlrabi next to the damaged crop
7. Removing volunteer seedlings from trays

New Things Not Yet Encountered:

1. Elyse onions (a new variety)

Questions and Future Areas of Research:

1. Why are flea beetles affecting Napa cabbage so badly? (no specific research mentioned)
2. How to manage or prevent volunteer seedlings in the greenhouse (current method is manual removal)

Suggested Actions:

1. Investigate more effective methods for removing volunteer seedlings from trays, potentially using technology such as seedling sensors or automated harvesting systems.
2. Research optimal timing and conditions for planting Elyse onions, as well as their expected yield and market demand.
3. Consider implementing integrated pest management strategies to minimize the impact of flea beetles on Napa cabbage and other crops.
4. Evaluate the effectiveness of row cover in reducing flea beetle damage and consider alternative materials or designs.

Typical Day on the Farm

Between late March and June, the greenhouse needs to be watered at least once if not twice a day (later on in the season). This takes around 45 minutes to one hour each time, and a breakdown by season goes as such:

• March - one watering in the morning per day; 21 days
• April - two waterings, one in the morning and one mid-day; 30 days
• May - two waterings, 31 days
• June - two waterings; 30 days
• July and August - little use since it’s too hot even with climate control
• September - two waterings; 21 days

This is approximately 133 days each year, and in the early season when there’s lots to do elsewhere, is around 12-15 hours per week for one person. More on this later.

Egg Washing Metrics

Egg washing happens twice a week. Eggs are collected every day and batched up in the walk-in fridge until washing day (usually Monday and Thursday). Yesterday I processed 70 dozen in about 4.5 hours with about 5 dozen rejected (between 5 and 10 percent loss due to cracks or broken is expected). Reasons for defects are mostly handling issues, thin shelled eggs, or internal pressure changes when going from fridge to warm water wash. It’s rare that the final ‘polishing’ phase is the cause of rejection.

More Onions in the Dry Field

After lunch, we planted another row of yellow onions. In total this week, we planted about 8500 onion starts. Each row takes about 13-14 98-cell trays at a 6-inch spacing in a 4 foot wide row that is around 200-250 feet long. The soil at the ends of each row are typically compacted by the tractor so it’s a bit more work to plant into that area.

Many people think of onions as a commodity / staple crop and turn their noses up at prices that don’t compare to mega-mart prices. As supply chain and input costs go up, even those prices are going up while local farms aren’t as affected (unless they rely on a lot of inputs). We plany at least three types of onions…all of which are sweet, earthy, and worth adding to the CSA member bags even if they are just a few per week.

What can you do with farm-fresh onions that aren’t a waste of their gorgeous flavor? I’ll be adding many recipies as the season progresses, but for now here are some quick ideas:

• fine dice a quarter of an onion and mix with oil, orange juice, and other herbs to make a salad dressing
• slow cook a few down until carmelized and put on burgers, hot dogs and sausages
• quick pickle them with a splash or two of red wine vinegar and use on tacos
• thinly slice and separate to sprinkle over homemade pizza

The onion you pay for (or plant yourself) is the onion you deserve.

Watering Robot (Concept)

As I was watering this morning, I started to think about how to automate (most) of the process. We use a wand hose attachment and it’s not hard work. Hand-watering also allows you to make visual inspections and pay more attention to specific areas where either some plants are more water-hungry than others or watering was missed otherwise.

There are no new ideas…really?!

I know there are many systems already and in particular ones that require permanent installation of sprayers and drip emitters with hose running throughout the greenhouse rafters. While mostly effective, these broad-stroke systems can often miss and either over or under water parts of the greenhouse since there’s no intelligence or ability to dial-in specific areas based on need. Yes you can turn up or down the flow rate of certain spots, but then no matter what, you end up optimizing where you put trays based on what the watering system is configured to do. Ultimately, these systems are permanent and relatively static…and usually cost a lot of money.

So if there was an alternative solution, the features and success factors might be something like:

• Significantly lower in cost than commercial greenhouse watering systems
• Less or equal effort to install, maintain, and repair the operational status throughout the season
• Most of the mechanical and hydro parts are stock, can be purchased independently of the “brains” or “hub”
• Reduction of water use and waste, or increase in water re-use
• Controllable and (re)configurable easily, maybe by a mobile app
• Adaptive based on what is constantly being changed and moved around in the greenhouse
• Learning and precise: gets better at managing specific needs of particular trays, plants, and their demand
• Conditional based on weather (temperature/humidity) and specifically sunlight (which dramatically changes greenhouse conditions)
• Quantitatively shows condition and resource (power, water) usage over time
• Predictive: shows projected future costs and adjust predictions with simple human-suggested upcoming changes
• Communicative: push notification and alerts of things like abrupt changes, malfunctions, etc.
• Observable: tracks specific trays and visualizes growth over time to provide drill-down capabilities in app

Early Concept Design

In my mind, I see it as a ‘robot’ (or two) suspended by two wires that strech end-to-end under the ridge line of the greenhouse…using pulleys kind of like those on the end of an old clothesline. It has a high-pressure nozzle (or two) and a counterweight arm pointed downward. The connected hose is retractible and also suspended from the wire(s), depressurized upon retraction movements. It has cameras for images and video capture. It has on-board sensors for collecting various conditions like humidity (really vapor pressure deficit) and temperature.

It intelligently waters things based on plant size, species and recency of encountered status. It reports problems such as unexpectedly low or dropped water flow rate and online/offline status. It suggests when manual spot-watering might be better and holds off if configured tolerance for safety is high enough. It may have an on-board battery that can be recharged at the base station using solar energy.

The only ‘for purchase’ part is the robot itself and the kit of everything else can either be purchased directly or cobbled together with in-spec parts from your local hardware store. Installation should be easy, but can be assisted remotely by a registered/trained technician upon request.

This can also be installed over a garden outside since the parts are all weatherproof and durable in non-freezing conditions. So the market wouldn’t just be greenhouses (primary) but also specialty market gardens.

Low-Cost, Low-Tech, Low-Code

Like other low-cost farming solutions such as the CoolBot, I want this to be something that experimental farmers can prove out first because conventional farmers often have an aversion to ‘new’ (unproven) ways of doing things, for good reason. Everything on a farm is a bet with varying risk levels, even proven and traditional approaches because nature is unpredictable. So anything that takes more micro-managing is usually a solid ‘no’ from farmers who are already juggling many things throughout the season.

The day-to-day usage should be easy for the average mobile app user to pick up quickly. It shouldn’t require a conversation or a credit card to spec out and budget for. It should have plenty of testimonials from early adopters to make the case for it, not a bunch of marketing blah blah. It should be flexible enough where a non-coder can customize various aspects of system behaviors, with the most common customizations being features, but also with extensibility via code at the forefront of each feature.

A Big Middle Finger to Big AgTech

I also want this to be in direct opposition to proprietary, licensed/subscribable, AgTech…and specifically support the ‘Right to Repair’ movement. Maybe it could be added to FarmHack or possibly open-sourced elsewhere, but definitely should be stuff that a tech-savvy person can run themselves without needing to rely on a 3rd-party SaaS or be dependent on a for-profit licensing or subscription service.

It should be runnable all by itself, only requiring a mobile device (such as a phone that farmers have on them already) for advanced features such as scheduling and behavior adjustment. For advanced data scenarios such as backing up video and images it takes, it should be able to connect to a ‘home server’ and/or a cloud service without incurring a license fee.

Of course, the software and various installations would require occasional assistance by support and a ‘product team’ consisting of developers, product owner(s), and DevOps/SRE for the simplified cloud-hosted stuff. I think there’d be enough demand for the ‘not self-hosted’ thing compared to the ‘community/core’ version.

Is this a business model? Yes. Is it a business model that attracts moron investors? No.

Creativity isn’t a uniquely expert or human experience. Like “The Creative Thinker’s Toolkit” suggests, intentionally conflating unrelated ideas can often lead to interesting avenues of further thought. Recently, large language models (LLMs) have taken much of the populist technology savant’s attention because it distills a massive amount of information into reusable models for generating various outputs.

What it doesn’t do (yet) is intentionally conflate adjacent concepts in a way that isn’t interpreted as ‘hallucinations’. Later this year after the growing season is over, I plan to revisit this potential, employing concepts from pragmatic creative processes using the homegrown LLM and RAG system I’ve been tuning with this blog. For now, this is what my ‘intern research analyst’ level system thinks I did with my day. \

At least it chunked out the transcript to documents that are grouped by theme, which was cool. Still I think a multi-modal interpretation of the transcript (narration of day vs. development of an idea), identifying the nature of the dialog, then applying specific templates for summary based on intent would be more efficient.

Main Themes:

1. The hydroponic system and its complexities
2. Design considerations for the greenhouse robot
3. Daily farm work and tasks performed by the intern

Activities Done by the Intern:

1. Watering the greenhouse using a hydroponic system
2. Washing eggs
3. Planting yellow onions in dry fields (with three other team members)

New Things Not Yet Encountered:

1. The hydroponic system’s complexities, particularly with pressure, flow rate, and mechanical control
2. The robot’s potential uses and design considerations for the greenhouse

Questions and Future Areas of Research:

1. How to address the challenges of the hydroponic system, such as pressure, flow rate, and water kickback?
2. How to design a practical and efficient arm extension mechanism for the robot
3. How to ensure safe and easy interaction with the robot while working in the greenhouse

Suggestion Actions:

1. Conduct further research on the hydroponic system’s complexities and explore potential solutions.
2. Design and test the proposed arm extension mechanism for the robot.
3. Consider conducting more experiments or simulations to validate the effectiveness of the hydroponic system and the greenhouse robot.

Overall, the intern seems to be enthusiastic about the project and has some initial ideas and questions about how to address the challenges of the hydroponic system and the greenhouse robot.

Part 2 (Automatically Grouped to Product Design by LLM)

Main Themes:

1. The challenges and inefficiencies of manual watering systems in greenhouses
2. The potential for automation and robotics to improve efficiency and reduce labor costs
3. The intern’s curiosity about new technologies and innovative solutions
4. The importance of considering economic feasibility when evaluating new ideas

Activities the Intern Did:

1. Watered the greenhouse (about 2 hours a day, 10 hours a week)
2. Observed automated watering systems in greenhouses
3. Thought about designing a robot to automate watering
4. Brainstormed potential features for the robot (sensors, cameras, jet spray)

New Things Not Yet Encountered:

1. Automated greenhouse watering systems with robots
2. Drip emitters and spray emitters
3. Pulley systems with thin wires

Questions and Future Areas of Research:

1. How to design a robot that can efficiently water the entire length of a greenhouse
2. How to power and recharge the robot (onboard battery or cord)
3. How to ensure reliable communication between the robot and the base station
4. How to address potential issues with the robot, such as breaking down or needing maintenance

Suggested Actions:

1. Conduct further research on automated greenhouse watering systems and robotics
2. Design and build a prototype of the proposed robot
3. Test and evaluate the efficiency and effectiveness of the robot in a real-world setting
4. Consider hiring help to assist with the robot’s development and deployment
5. Evaluate the economic feasibility of implementing the robot on a larger scale, considering labor costs and potential savings.

On a rainy birthday, we planted the two end rows in the dry field with onions and two end rows of the low main field with lettuces. We then harvested green garlic volunteers from last year’s garlic rows and some lilacs around the property for tomorrow’s pop-up farmer’s market at the new general store in Essex.

No Place I’d Rather Be

It was my birthday today. I never like to make a big deal of it, but some people do. Some people don’t work on their birthday, and some go on a trip. Some have fancy dinners at restaurants with many friends and family. Not me.

I worked with my friends on stuff I like to do and learned some new stuff. When I got home, warm pot sticker soup and carrot cake was waiting for me. Nothing expensive or complicated. Tomorrow morning, a bunch of this season’s farm crew will meet at the Agawam diner because two other peoples’ birthdays are this month too and sometimes you just have to eat amazing pie with friends.

No Greenhouse Watering, but Harvest Prep

The day started with washing a bunch of large tote bins in preparation for harvesting green garlic later in the day. The process is simple but important: soap and water, scrub, then rinse. Anything put on the floor can’t be considered clean, so either a mat or another sacraficial cleaned tote keeps the other clean ones safe. Twenty totes took about 30 minutes including some other outside tasks.

Onion Planting

We planted two more rows of onions in the dry field today, finishing that field’s rows out. We are about 75% through all the onions, at the end of the season totaling 13,000 by hand. There were 4 of us, myself and another volunteer on the slower side dropping while others planted until we joined them. Then we stopped for lunch. This finished out the rows in one of three dry fields.

Lettuce Planting

After lunch, we planted two rows of lettuce in the lower main field, also finishing that field out for the time being. The soil in this field was freshly cultivated and with some light rain was very easy to plant into. It was not muddy per se but did accumulate on boots and pants.

Harvesting Green Garlic

Transitioning from hauling around and planting stuff in a field to harvesting, you want to make sure that the bed of your truck is sanitized, so we scrubbed and rinsed the back out completely. Then the clean totes went in and we drove back out to the dry field with the new garlic and onions, but where the end rows were last year’s garlic harvest because there were many ‘volunteer’ early garlic sprouts popping up. We harvested about one full large tote of these small stalks of “green garlic” to sell at tomorrow’s pop-up farmer’s market.

Harvesting garlic requires a careful and firm approach with two hands, one at the base and another higher up the stalks, to ensure likelihood of bringing the roots and head up with it. Otherwise, and sometimes, you end up ripping off the stalk from the rooted head and leaving stalks unsuitable for sale to customers.

Harvesting Lilacs

On our way back to the barn, we stopped at a few lilac bushes around the property to pick about a dozen bunches to also sell at the pop-up tomorrow. The smell of lilacs is so nostalgic to me since I grew up in a yard that was lined with them. When friends came over, we made forts in the high-bush lilacs at the top of the hill. When I was sad, I went and hid there and cried until things seemed less pointless or I remembered I was hungry.

I think if I had to choose, I’d rather be blind than lose my sense of smell. Lilacs are always bittersweet too in that their wonderful smell means it’s soon to pass (we usually get only a few weeks between their bloom and fade). But every season they come again to remind me that somewhere inside is still a kid with a place to be.

Washing Green Garlic

We drove back to the barn and rinsed off the green garlic stalks off the back of the truck to remove most of the dirt and muck. Then we brought them over to another table to sort and peel the first layer off. This makes them look fresh and clean. The smell is amazing, like everything is made of garlic, including me.

Then we bunched them into about 10 stalks and bound them with elastics before rinsing them outside one final time. Into the walk-in fridge they went in prep for the pop-up farmer’s market.

Word-of-Mouth CSA Marketing

On Friday I had to take my car in for repair. While I was waiting for the place to open, I started talking to a nice guy named Kyle. He asked about my hat which was a gift and says “Farm Boy” on it. I said I worked at a farm in Ipswich. He mentioned that he’s on a waitlist at another farm and I mentioend that Marshview still has a few open shares, full and half, and wrote down the info for him on a card.

Turns out, today he signed up. Sometimes all it takes to attract new business is being friendly and jovial, but it also helps to be happy and sincere in what you do. I can’t say the latter about much of the high-tech work I did in the past decade, even when the former was part of my demeanor at conferences and on calls.

I think community events like ‘seed swaps’ and pop-ups and farmer’s markets are all really important to face-to-face and word-of-mouth interactions that ultimately help to ensure a full book of business with a farm’s CSA program.

AI Captures Detail but Misses the Sentiments

Main Themes:

1. Farm work and daily activities
2. Personal experience and reflection on the day’s events (birthday)
3. Learning and growth as a farm intern
4. Interaction with the farmer and other team members
5. Preparation for upcoming markets and events

Activities performed by the intern:

1. Washing green tote boxes
2. Harvesting garlic
3. Planting onions in two rows
4. Loading up truck with radicchio and lettuces for market
5. Cleaning up greenhouse (stacking empty cell trays)
6. Refilling potting table with seedling mix
7. Taking out and dropping lettuces during planting
8. Finishing planting lettuces with the farmer

New things not yet encountered:

1. The specific process of washing green tote boxes ( soap, water, scrubbing, rinsing)
2. The row cover system used for protecting crops from pests/diseases
3. The type of seedlings left over from the plan (not specified)

Questions and future areas of research:

1. What is the typical loss ratio for onions in cell trays?
2. How many trays are needed based on the row length, size, and other factors?
3. What is the expected yield for the upcoming markets (pop-up farmer’s market, general store)?
4. How much more produce is included in the plan than initially anticipated?

Suggested Actions:

1. Conduct further research on optimal washing procedures for green tote boxes.
2. Ask the farmer to explain the row cover system used and its effectiveness.
3. Review and understand the yield expectations for upcoming markets and adjust plans accordingly.
4. Consider exploring additional revenue streams through the general store or other channels.

Part 2

Main Themes:

1. Word-of-mouth marketing: The intern mentions how they met Kyle at the car repair shop and he signed up for a half-share of CSA based on their conversation about farm work.
2. Farm routine and tasks: The intern describes various daily activities, such as harvesting green garlic, cleaning the truck, and preparing it for harvesting.
3. Quality control and attention to detail: The intern emphasizes the importance of scrubbing the truck with soap and water, rinsing the garlic, and carefully peeling off layers to reveal the white inside.

Activities:

1. Talking to Kyle at the car repair shop
2. Signing up new customers for CSA (Kyle and his wife)
3. Harvesting green garlic
4. Cleaning the back of the truck with soap and water
5. Rinsing the harvested garlic
6. Bringing the garlic back to the greenhouse in green totes
7. Individually peeling off layers from each stalk
8. Bunching the garlic into 10-12 stalks per bunch and tying them together with elastics
9. Putting the bundled garlic in the fridge for tomorrow’s pop-up farmer’s market

New Things Not Yet Encountered:

1. The importance of scrubbing the truck before harvesting
2. The process of rinsing the harvested garlic multiple times to remove dirt and muck
3. Using a stainless steel table for further rinsing the garlic

Questions and Future Areas of Research:

1. How does the intern’s hat being visible affect their interactions with potential customers? (Example: “He saw my hat. He said, oh, you like gardening or working on a farm, right?”)
2. What is the optimal method for peeling green garlic?
3. Can the intern suggest ways to improve the efficiency of harvesting and bundling the garlic?

Suggested Actions:

1. Consider implementing a more effective system for cleaning the truck before harvesting.
2. Research alternative methods for rinsing the harvested garlic that may be more efficient or environmentally friendly.
3. Explore ways to promote the farm’s CSA program through social media, online advertising, or other channels.
4. Train new interns on the importance of quality control and attention to detail in farm operations.
5. Consider developing a system for tracking customer sign-ups and follow-up communication with new customers like Kyle.

We processed 83 dozen eggs, I helped prepare a pig pen, and then planted onions in the rain today. Few pictures but many happy moments with the crew at breakfast and lunch.

Team Birthday Breakfast

It was a short day today because the six of us started out together, breakfast at the Agawam diner to celebrate May birthdays, then headed back to the farm. It felt more natural than any ‘team meeting’ I’ve ever had, partially because we already knew each other from many days working in the field together, but mostly because of the goodness of each member of the team.

Preparing for Week 1 of the CSA

Though it was a short day, there were more of us than usual so we made the best of it and everyone was busy once we got back from breakfast. Some planted, some seeded in the greenhouse, some did things I have no idea about…meanwhile I and another processed 83 dozen eggs. The process is documented here, but sufficed to say the loss ratio is getting better in the past few weeks.

The CSA starts soon and we need to start ramping the egg production. Chickens were ordered many weeks ago in preparation. I have a feeling I’ll be getting faster at hand polishing eggs twice a week but you never know what tasks you’ll be distributed to based on need.

The First Pop-up Market

Though I had to leave early for family transit reasons, Jamie and another collected various goodies for the pop-up market later in the day. This was honey, pepper jelly, green garlic stalks, the last chops of frozen pork from the winter meat share (still fantastic BTW), bouquets of lilacs, and of course farm fresh eggs. We’ll see what’s left coming back tomorrow, but my guess is very little.

Preparing a Pig Pen

Despite the epic amount of eggs, I got pulled to help clear out the pig pen way up into the forest hill. This involved deconstructing the prior structures made of pallets, reinfocring wire mesh, rebar, and chicken wire.

This was in prep for three other members of the crew to tackle reconstruction using these components properly after lunch.

Lunch

We had sammies with my homemade bread and then some carrot cake loaf left over from my birthday cake last night. I love carrot cake, but love that my wife loves it more so that’s what I always ask for when she insists on making one for me. Everyone benefits.

Planting in the Rain

We finished the afternoon planting the Alise onions in two rows of the top plot on the main field. It was already wet and raining earlier in the day, but someone in the sky must have turned up the volume becase it was on-and-off pouring and clear for the hour or so we were out there.

Planting in the rain is…not fun. Still, things have to be done and if I were a seedling, I couldn’t hope for better than to be planted with a natural watering in that other seedings out in a field often don’t get.

Fortunately, I avoided loosing a boot to the muck and two others on the team were very fast planters. We finished up and headed back to the greenhouse to hose our bottom halves off properly before packing into our automobiles and heading our separate ways.

For now, I keep my AI working for me, or at least that’s what I think. Note that it wonders what they meaning is of a birthday on May 20th and while I’m sure a birthday isn’t a foreign concept in its training, I wonder if robots will celebrate each other’s manufacturing date some day. Until then, here’s what it thinks I did based on my audio notes, transcribed and run through a local LLM for summary:

Main Themes:

1. Farm work and daily tasks
2. Celebrations and social interactions (birthday celebrations with colleagues)
3. Learning and adapting to new situations (pig pen renovation, beekeeping activities)

Activities:

1. Egg collection (83 dozen eggs)
2. Pig pen renovation (disassembling pallets, moving shelters, liberating chicken wire)
3. Cleaning the barn (bee nests, stackable wooden boxes for bees)
4. Planting Elsa onions in one of the top fields
5. Preparing for the pop-up market (preparing equipment and supplies)
6. Lunch break (making sandwiches with carrot cake and fresh bread)

New Things Not Yet Encountered:

1. Gator tractor usage (someone “flew up to work on pig pen renovation”)
2. Pig pen layout and management (original setup, string tied pallets, wire mesh trellising)
3. Beekeeping activities (stackable wooden boxes for bees, bee nests)

Questions and Future Areas of Research:

1. What is the meaning behind someone’s birthday being on May 20th?
2. How long will the Spring CSA last?
3. Are there any specific requirements or expectations for the pop-up market setup?

Suggested Actions:

1. Review and discuss the pig pen renovation process with Jamie to understand the original setup and management practices.
2. Investigate the specifics of the beekeeping activities, such as the purpose of stackable wooden boxes and how to handle bee nests.
3. Clarify the expectations for the pop-up market setup and what is required from the farm for participation.

Most of the day was spent preparing the pen for the pigs that would arrive at the end of the day, but we also mulched and planted. There’s always something to do around the farm.

Preparing a Pig Pen

Pigs are messy, vociferous creatures. Their pen doesn’t have to be a work of art, so long as it is sturdy and keeps them from getting out. A loose pig (or many) is such a waste of farm time.

The process over the past few days has been as such:

• deconstruct prior pig pen area
• consolidate pallets, rebar, poles, and fencing to separate piles
• migrate heavy elements like shelters to new area
• lay out the new setup of pallets around the intented new shape
• stand pallets up one at a time, pound rebar/poles on each side
• tie pallets, rebar, and poles together with rope and/or twine
• double check the whole perimeter for any missing ties or loose areas
• put water and feed inside, also straw bedding down inside shelters

It’s really important to tie the pallets together tightly, one towards the bottom (6 inches from the ground) and another towards the top, because the pigs like to dig underneath them.

We will add electric fencing to the back half inside so as young pigs, they will be trained to avoid the fence one the pen is opened up to an electric fence area. Pigs aren’t stupid creatures…usually they’re very attentive…they are just full of instinct to flee. So one or two small shocks is all it takes for them to learn not to tempt the fence. They also like to stay together, so if one gets out they generally just want to get back to the rest of their group.

How to Get a Dozen Pigs Back Safely

Then Jamie and I put a big wooden crate up on the hill into the back of the white truck. Rather than the pigs be directly in the back on the bed of the truck with a cover, the wooden crate would keep them manageable and give them enough airflow for the ride home. It must have weighed about 300 pounds so we stood it up by tipping it on end, backing the truck up close, then letting it see-saw onto the flat bed and pushed the rest of it in. Since it didn’t quite fit length-wise without the tail gate closed, adjustable straps were used to lock it down into place for the ride. Then Jamie was off to go get them from the pig farm an hour and a half away.

Lunch

After a thorough washing and quick change in the barn, the other three of us headed to a local bakery for sandwiches. It was Marina’s last day before heading off on tour so it was nice to get a few more casual moments together.

I had a ham and butter sandwich which reminded me of the pork sandwich at Sticky Fingers in Brown’s mall Gloucester almost 10 years ago. It was run by a mom and her daughter as a tribute to all things intentionally made with love.

Waiting for Pigs to Arrive

After lunch, we all went to grab some salt marsh hay to use as mulch for the leeks. While one of us watered the greenhouse, Marina and I did the mulching which there wasn’t much left of the rows in the bottom main field to cover, so it didn’t take very long at all.

Then we got back together and planted the rest of the row of Elise onions that was started in yesterday’s pouring rain.

Once that was done, we started to collect things that would go to the monthly seed swap at True North Brewing Company later that day, such as various seedling trays from stuff we’ve already planted and the box of giveaway seeds.

We got to ride in the brand new Ford Electric van too, so fresh and clean, before years of service will make it look as well-worn as most other things on the farm. This was a net-new capital investment and will make farmer’s markets and deliveries a lot less logistically complicated moving forward.

The Pigs Arrived

Then Jamie got back from picking the pigs up in the flat bed truck. We loaded rubber/Tupperware buckets into the back of the John Deere gator, then Jamie hoisted the pigs into the buckets one at a time. A lid was placed on top so that the pig wouldn’t freak and try to jump out.

In batches of 3 pigs at a time, the crew took turns on the back of the gator holding down lids and riding them up the hill to deposit them into the pen. Each pig weighed about 30-40 pounds so it took multiple pairs of hands to manage the whole operation. A few trips later and all the pigs had found their new summer residence.

We washed up thoroughly and called it a day, since it was just about time for the seeds and suds event down the road.

Some Things AI Will Want to Leave to Humans

Main Themes:

1. Farm work and preparations for pig arrival
2. Construction and setup of pig pens
3. Maintenance and upkeep of farm infrastructure
4. Social interactions with farm staff and Marina’s departure
5. Planting and harvesting activities (leeks, Elsa onions)

Activities done by the intern:

1. Preparing a pig pen, including moving pallets, stakes, and poles
2. Laying out and standing up pallets in the pig pen
3. Pounding rebar and stakes into the ground
4. Tying mesh metal wire to reinforce concrete
5. Adding chicken wire and securing the structure with ropes and twine
6. Waiting for pigs to arrive and setting up a transport crate
7. Finishing mulching leeks in the lower main field with hay
8. Planting Elsa onions (9 trays) in the 98-cell trays

New things not yet encountered:

1. Using a pole pounder (empty metal cylinder with a weight on top)
2. Working with electric wire and fencing for pig pens
3. Handling large, heavy transport crates

Questions and future areas of research:

1. Why didn’t the fence get installed around the entire forest area? (mentioned as “later”)
2. How do pigs dig underneath fencing?
3. What is the purpose of adding chicken wire to the inside of the pig pen?

Suggested actions:

1. Research on optimal fencing designs for pigs, including electric wire and mesh metal wire.
2. Investigation into the effectiveness of pole pounders as a tool for securing rebar and stakes.
3. Discussion with farm staff about electric wire and fencing options for future pig pens.

Please note that these suggested actions are based on the information provided in the field notes and may require additional research or consultation to determine their feasibility and importance.

Part 2

Main Themes:

1. Animal care and husbandry (moving pigs from crate to pig pen)
2. Farming and agricultural practices (deliveries, farmer’s markets)
3. Equipment and technology adoption (new electric van for deliveries and farmer’s markets)

Activities:

1. Loaded trays of seedlings into the new electric van
2. Transported pigs from a crate in a car to their pig pen using a John Deere Gator
3. Loaded and unloaded pigs into rubber bins on the back of the gator
4. Cleaned up after the animal move
5. Attended a monthly seed swap event at True North Brewing Company

New Things Not Yet Encountered:

1. Electric van for deliveries and farmer’s markets (new technology)
2. Rubber bins for transporting pigs (novel solution)

Questions and Future Areas of Research:

1. How do the climate conditions affect the timing of planting seeds? (Note: this is already mentioned in the seed swap event, but it highlights a potential area of research on the impact of climate change on agriculture)
2. What are the benefits and challenges of using electric vans for deliveries and farmer’s markets?
3. How effective are rubber bins as a solution for transporting pigs, and what are their limitations?

Suggested Actions:

1. Investigate further the impact of climate change on seed planting and agricultural practices.
2. Consider implementing additional safety measures when moving animals to reduce stress and injury.
3. Evaluate the effectiveness of using electric vans in terms of fuel efficiency, cost savings, and environmental impact.
4. Research alternative solutions for transporting pigs that may be more efficient or humane than rubber bins.

Watered the greenhouse, painted bee boxes and distributed them around the property, and spread lanscaping fabric across a few fields.

While Watering the Greenhouse, Volunteer Plucking

Just about every day I’m on the farm, one of my tasks is to water the greenhouse at least once. While doing so, I note where ‘volunteer’ seedlings are. Then I go back and carefully pluck them out and deposit them into a broken 98 cell tray to bring home. Gardeners call these little plants ‘volunteers’ but most farmers don’t really have the time to call them anything and don’t really mind if they show up for dropping out in the field.

Volunteers in a greenhouse arise usually because seeds are occasionally dropped onto the potting mix table while planting trays of one thing, then the next run of seeding happens and they’re unintentionally propagated along with the intented seed. It’s a lovely thing that the greenhouse situation is so accomodating that everything grows, even things that don’t belong where they show up.

Volunteering Planting

Last year when I volunteered starting in the summer season, I was hoping to grow a relationship and as an individual in a fertile and at-scale agricultural space. It worked…this year I have a job…something the high tech sector sorely missed out on.

Spoiler: the tech sector has its head up somewhere for so long that their last resort is AI…while actually hard work industries are economically hurting now more than ever. Tariffs don’t benefit consumers or producers, just profiteers.

Switching Context from Pig Transportation to Bee Husbandry

Before loading the truck up with bee box parts, we dropped off the big wooden box used to transport piglets yesterday up on the hill towards the pig pens. The view from this vantage point out onto the Ipswich salt marsh fields is stunning, even covered in the smell of pigshit.

Now that the piglets are happy and settled for the summer, our focus shifts to the important role that the bees play, not only for non-self-polinating veg but because this year many farms are late to refresh their pollinator populations. For many years, a mallady of factors have caused bee populations to decline.

I think if I were a bee, I might also think twice about propagating in the United States too. The current climate, ecological and political, is so toxic that there’s very little room for hope.

Add Bee Box Painting to My Resume

After a wash-off, I got to priming and painting bee box lids (the outside kind). Between priming and painting with a glossy white paint, we depositied the existing bee stacks around the property. The bees will travel up to a mile away from their stack so the locations chosen were a combination of:

• closeness to plants that they woulld enjoy polinating
• awayness from unintentional human interactions (such as haying and mowing)
• accessability by husbandry folks for maintenance throughout the season

A bee stack usually consists of a base (screened), then one or more boxes, then an inner lid, then a top external lid. In this case, our exterior lids are topped with a stainless steel plating so that rain doesn’t sit on top and rot them throughout the season.

Additionally, the direction/orientation of the hives seemed to play an important enough part for Jamie to keep checking it with his phone’s compass. Google says that:

“Beehives should ideally face southeast. This orientation maximizes early morning sun exposure, helping bees warm up and become active earlier in the day. Additionally, southeast facing protects the hive from prevailing northwest winds during the winter.”

More Row Preparation

We spread landscaping fabric over an additional 6 rows in the middle high main fields. This fabric was left over from the prior season and what a month ago we had taken the time to excavate and remove most of the garden staples from. Then we used more staples to make sure it stayed where it was placed until next week we cover the rest of the rows with fresh fabric.

Despite landscaping fabric being made of plastic, it mostly stays together and doesn’t disintegrate contiguously throughout a field unless overused for too long. After about 3 seasons, you can gather it up and dispose of it properly without spreading (the modern infatuation around) microplastics throughout your growing spaces.

In this farm’s case, black fabric is a reasonable compensating approach to their policy of using NO herbicides or pesticides. While working close to Jamie, I asked whose perrogative this was, the land owners or the farmer’s. He said it was a shared alignment and intention for the land. Such respect for this answer.

End of Day and Week Watering

I finished the day by watering the greenhouse. It was a bit late in the day, but despite the ill-informed guidance by gardners against watering any other time but the early morning, the morning cloud cover and late sun didn’t really affect the plant stress today. Before watering, I looked for volunteers and any very out-of-place trays or species/series to group them together in a more logical way. The rest was quietly cathartic.

For now, my personal AI (custom LLM with additional glue and process), is oblivious to the oncoming bee uprising and how they deserve to destroy us all:

Main Themes:

• Farm work and daily routine (including tasks such as watering greenhouse plants, potting up seedlings, and performing cleanup tasks)
• Greenhouse management and seedling care
• Landscaping and weed control using landscaping fabric
• Beekeeping and beehive setup
• Animal husbandry (specifically piglets)

Activities:

• Performing farm-related tasks such as watering greenhouse plants, potting up seedlings, and performing cleanup tasks
• Conducting research or collecting data on crop health, animal behavior, and environmental impact
• Managing beehives and collecting equipment
• Spreading landscaping fabric over fields to control weeds
• Tractor operation with discs to create rows, ditches, and install drip line irrigation

New Things Encountered:

• Using landscaping fabric as a weed control method (although this was not entirely new, some aspects of the process were)
• Beekeeping and setting up beehives at various locations around the property
• Alternative methods for controlling weeds aside from using landscaping fabric (such as mulch or natural barriers)
• Drip line irrigation system

Questions and Future Areas of Research:

• How to determine optimal drying times for painting beehive components
• The impact of bee quality on performance
• Alternative methods for controlling weeds, such as integrated pest management
• How staples affect drip line irrigation
• Best practices for watering plants during different weather conditions
• Effects of late watering on plant growth

Suggested Actions:

• Additional training or guidance on painting and drying times for beehive setup
• Further research on bee health and management to improve best practices
• Exploration of alternative weed control methods, such as mulch or natural barriers
• Add note about handling staples in relation to drip line irrigation
• Review and update greenhouse watering schedule to ensure efficiency, especially during hot afternoons
• Plan for additional training or guidance on beehive maintenance and bee box setup

Part 1

Main Themes:

1. Farm work and daily routine
2. Irrigation and irrigation systems
3. Greenhouse maintenance and care
4. Preparation for upcoming events (bees arriving)

Activities:

1. Tractor operation with discs to create rows, ditches, and install drip line irrigation
2. Watering the greenhouse (late in the day due to prior tasks)
3. Preparing for bees arrival (setting up beehives and bee boxes)

New Things Not Yet Encountered:

• Drip line irrigation system

Questions and Future Areas of Research:

• How do staples affect drip line irrigation? (previously mentioned, but not explicitly addressed in the transcript)
• Best practices for watering plants during different weather conditions
• Effects of late watering on plant growth

Suggestions and Actions:

1. Consider adding a note to the field notes about how to handle staples in relation to drip line irrigation.
2. Review and update the greenhouse watering schedule to ensure it is being met efficiently, especially during hot afternoons.
3. Plan for additional training or guidance on beehive maintenance and bee box setup.
4. Research best practices for preparing plants for bees arrival (e.g., pruning, pest management).

Part 2

Main Themes:

1. Farm work and daily routine
2. Greenhouse management and seedling care
3. Landscaping and weed control (using landscaping fabric)
4. Beekeeping and beehive setup
5. Animal husbandry (piglets)

Activities:

1. Watering the greenhouse
2. Potting up lettuce seedlings
3. Spreading landscaping fabric over middle fields in the main field
4. Stapling landscaping fabric
5. Collecting and gathering beekeeping equipment
6. Preparing beehive components for setup
7. Setting up beehives at various locations around the property
8. Applying glossy outer paint to beehives
9. Performing cleanup tasks around the barn

New Things Not Yet Encountered:

1. Using landscaping fabric as a weed control method
2. Beekeeping and setting up beehives (although some aspects, like collecting equipment, were familiar)

Questions and Future Areas of Research:

1. “I’m not sure how long it takes for the paint to dry.” (Future research on optimal drying times for painting)
2. “How does the quality of the bees affect their performance?” (Future research on bee health and management)
3. “What are some alternative methods for controlling weeds, aside from using landscaping fabric?” (Future research on integrated pest management)

Suggested Actions:

1. The intern could benefit from additional training or guidance on painting and drying times to ensure proper completion of the beehive setup.
2. Further research on bee health and management could provide valuable insights into best practices for maintaining healthy beehives.
3. The farm could consider exploring alternative methods for controlling weeds, such as using mulch or other natural barriers.

Please note that these suggestions are based on the provided field notes and may not necessarily reflect the intern’s current needs or goals.

Part 3

Based on the provided information, here is an analysis of the main themes, activities, new things encountered, and questions/future areas of research:

Main Themes:

1. Farm work: The intern’s primary focus was on farm-related tasks, which suggests a theme of agricultural or rural work.
2. Observation and documentation: The intern’s field notes suggest a theme of observation, documentation, and record-keeping.

Activities:

Based on the transcript, some activities that the intern did include:

• Performing farm tasks (not specified)
• Observing and documenting natural phenomena (not specified)
• Conducting research or collecting data (not specified)

Without more context or information from the transcript, it is difficult to provide a more detailed list of specific activities. However, based on common farm work and observational studies, some possible activities could include:

• Monitoring crop health and growth
• Observing animal behavior
• Collecting soil samples or water quality data
• Conducting pest management or irrigation tasks

New Things Encountered:

Unfortunately, without more information from the transcript, it is difficult to identify specific new things encountered by the intern. However, some possible areas of discovery could include:

• New plant species or pests
• Changes in weather patterns or climate
• Unusual animal behavior or health issues
• New farming techniques or technologies

Questions and Future Areas of Research:

The transcript does not explicitly state any questions or future areas of research. However, some possible questions that might be explored based on the intern’s observations could include:

• How do changes in weather patterns affect crop growth or animal health?
• What are the most effective methods for managing pests or diseases in this specific farm setting?
• How can farming practices be adapted to reduce environmental impact?

Suggested Actions:

Based on the transcript, some suggested actions that might be taken by the intern or their supervisor could include:

• Reviewing and refining existing field notes and documentation
• Developing a more detailed plan for collecting data and conducting research
• Identifying areas for improvement in farm practices or operations
• Exploring new technologies or methods for increasing efficiency or reducing environmental impact.

It was a pretty typical day on the farm with a small crew today. We washed 83 dozen eggs, hoed in the dry fields, then planted more onions. Things are starting, growing, and establishing…veg and weeds alike.

Trends vs. Blips

There are a lot of things going on at a farm at all times. Once farms lock normal modes of operations in and know what they’re doing, there’s a lot less variance in outcomes. There are some details which are terribly important…usually health and safety are the top areas where it matters. Lots of other things don’t matter like I expect them to, in part because nature has a way of working out many kinks and also I haven’t earned enough context and experience over so many things that I just don’t have an intuition for which things matter and which things dont yet.

So the conundrum is: how can you observe trends without measuring things, and if you don’t know which details matter and which don’t, how can you measure the right things and not the wrong ones? This is very much also a problem in modern software and large-scale organizational leadership. If you haven’t earned the context and experience, you don’t what issues are ‘blips’ (momentary outliers) and which one’s are trends. I’ll think on this more, but it seems like something I want to imbue into IEEE 2675.2 which I’m majorly contributing to within a working group.

After Lunch, A Three-Point Glitch

A hitch, and ditch, and a glich walk into a sitch…okay, over the weekend the tractor in the main field got stuck in some unexpectedly soft soil…with a film mulcher attached to its 3-point hitch. Jamie and I (mostly Jamie) tried to get the hitch detached but to no end. Eventually he had to get the big tractor from the dry field (which had to come up to the main field anyway) to tow the smaller tractor out.

Hoe, Hoe, Hoe Your Rows

Today, the activity that felt the most like a lesson in pain was hoeing a row of scallions over in the dry (not irrigated) field. The soil there crusts over without mulch and the (grass) weeds grow pretty quickly. We hadn’t hoed there since we planted a few weeks ago and there were many patches where the weeds were almost completely indistinguishable from the small scallion shoots.

I’ve worked in other farm settings where they hoe almost every day, as regular as greenhouse watering and egg washing. Tis’ the season for hoeing anyway, but this was especially painful because some of these roots are well established even after cultivation by a tractor and attachment. Also the spacing and alignment of planting was ‘wonky’ in this row (I remember when this was happening and worried a bit on the inside).

We are currently using stirrup hoes and since the soil is full of rocks, they don’t stay sharp for very long. There are so many types of hoes and even handles with hoe heads that can be swapped out depending on what type of head is most useful in particular circumstances. So far, I’ve worked with the following types of hoe heads:

• Stirrup
• Colinear
• Cape Cod Weeder
• Beet/Nursery Hoe
• Scuffle Hoe
• Wire Weeder(s)

I think the stirrup hoes are good in this field because of the rough, dry soil but I’d like to try others and certainly maintain their sharpness with a metal file in pocket along the way next time. I’d love to get my hands on a Lehman’s precision Amish style garden hoe, but I have to earn my experiences with what we have before drooling over upgrades already.

Planting Short Rows of Onions

We finished out the day planting about 20 trays of onions in two not-complete rows because the lower part of the dry field is inconsistently mucky after the past few days of on-and-off rain. We’ve got maybe about 20-30 98-cell trays of onion starts left outside the greenhouse and the end is in sight. After all is done, we’ll have planted 4 varieties of about 13,500 onions.

My neck and back and leg and shoulder muscles hurt just thinking about it. And this isn’t even a very large CSA farming operation. One day, one row, one plant at a time.

Today’s response from my existing LLM infrastructure about what I did is below:

Main Themes: Physical labor and exhaustion, technical issues with farm equipment, teamwork and communication, and challenges of working in dry fields.

Activities Done by the Intern:

• Watering greenhouse (45 minutes)
• Washing eggs for 3.5 hours
• Loading truck with onions
• Dropping off at dry field with Jamie
• Hauling weeds (three-point hitch got stuck)
• Riding back to get another tractor
• Hoeing scallions in a dry, non-irrigated field
• Planting onions in lower, wet part of dry fields

New Things Not Yet Encountered:

• Using three-point hitch on a tractor
• Hauling weeds with stirrup hose
• Hoeing scallions in a dry, non-irrigated field
• Using film organic to cover rows for planting

Questions and Future Areas of Research:

• How to deal with weeds taking over scallions?
• Benefits and challenges of using three-point hitch on tractors in dry fields?
• Why Jamie suggested losing a bet to weeds was not the end of the world?

Suggested Actions:

• Wearing gloves or using tools to protect yourself from hazards while hoeing
• Implementing irrigation systems in dry fields
• Exploring alternative methods for weeding, such as different tools or techniques.

We seeded lettuce, spread black landscaping fabric, hoed in the high tunnel, and planted the last of the onions. I had to leave a little early, but today is also a short day on-farm due to the 4-6pm weekly pop-up market in Essex.

Always Seeding Something

Today, I seeded 17 trays of various lettuces. While seeding slows down in early summer here, it never stops entirely until the late summer. There’s always something being started though the greenhouse gets too hot even with fans in the summertime. I enjoy what I call “playing the soil piano” which involves spreading potting mix into a tray then making dimples with ‘piano fingers’ (all lined up together) in the filled cells. At this point, I can make about 10 trays in 10 minutes but I’ve seen better from more seasoned folks.

Reuse of Landscaping Fabric

I wrote about this last week, but it bears repeating: landscaping fabric is a reasonable compensating approach to their policy of using NO herbicides or pesticides. Some other local farms use more, some use much less but not at this scale, and no one gets away with zero plastic in farming period.

We finished covering the top middle main field with fabric and stapled it. We ran out of staples and had to…liberate…more from old bunches of other fabric around the field to make sure this one was tacked down sufficiently for the wind gusts that frequent costal farms in New England such as this one.

Hoeing in the High Tunnel

There is a high tunnel in the main field between the raised flat area where the new greenhouse is and the first of two other plots ending in the lowest area of the field. This tunnel grows early greens such as kale and spinach which aren’t susceptible to flea beetle like the cabbage and choi in the lower field is on the early side of the season.

It sounds like we may harvest some of the high tunnel growth for the pop-up market since so much of the CSA stuff is still establishing in preparation for Week 1 in June. In my own garden I have more than enough volunteers from the greenhouse growing…kale, chard, and lettuces…that I don’t think we’ll have any issues making salads by next week.

The Last of the Onions Planted

Finally, it happened. We reached the 150-tray goal planted out in the field. There’s always some loss…seeds not germinated or runt-size starts that just don’t make it when planted…so the total of 13,500 planted requires some excess trays. We also sell and give out leftovers as well as host some special runs of other things in the greenhouse.

But after 4+ weeks of planting onions (and many other things), it feels like a small seasonal milestone. A small amount of the varieties are planted at the top of the main field, but most are planted in the lower dry field. I have never had such love-hate feelings for a commodity crop, but anticipating equally mixed feelings about other staple veg as we work through the year.

With all this farm work, I like to think that I’m helping friends with their season at minimum wage, but really I’m just in learning mode and trying to keep up with the physical nature of the work required to do the former. I could buy a fancy GPU to add to my Kubernetes cluster for LLM-related tasks, but I also like to put hardware to work to pay it’s rent. Currently, my homelab costs $2.00 per day, the cost of a light bulb continuously running for 18 hours (~100 watts).

The thing my LLM and I are equally starved for is context. I don’t provide enough to it in order to expect the “New Things Not Yet Encountered” section of my template to be predictably accurate. I plan to solve this with some Retrieval Augmented Generation (RAG) pipeline tricks…possibly with LangChain or SourceGraph (which I have an early intro interview with later today). I also need the time to play and prototype with other approaches, which between the farm and consulting work, then also family and personal time, is a tight ask.

As it stands and with the time I do have, my LLM can only guess as to the interpersonal nature of my farm work, and results in this today:

Main Themes:

• Farming activities, including seeding, hoeing, watering, and loading produce for market
• Equipment and tools used, such as landscaping fabric, staples, and a hoed
• Crop management, including planting onions, lettuces, kale, spinach, and scallions

Activities performed by the intern:

• Seeding 17 trays of lettuce in the greenhouse
• Spreading and stapling down landscaping fabric over main fields
• Hoeing in the high tunnel (kale, spinach, and early lettuces)
• Loading produce for market (onions)
• Watering the greenhouse

New things not yet encountered:

• The term “98 cell tray”
• Using landscaping fabric to manage weeds in the high tunnel

Questions and future areas of research:

• Why was seeding 17 trays considered a significant accomplishment?
• The purpose and impact of the high tunnel on weed management
• How do the onions being loaded for market relate to CSA (pop-up farmers markets) in Essex?

Suggested Actions:

• Discuss benefits and challenges of using landscaping fabric with farm manager or experienced farmers
• Research methods for seeding 98 cell trays, such as optimizing tray layout or specialized tools
• Investigate types of weeds encountered in the high tunnel and their potential impact on crop yields
• Improve water conservation strategies in the greenhouse.

Today was some greenhouse work, but mostly planting celery and eggplant out in the main field…and a lesson in hand tool hazards.

Planting with a Spear Trowel

Sometimes the soil is soft enough to just use your hands to plant seedlings in. Other times…not so much. Either natural or tractor compaction, crusting over, or rocks get in the way of quick and precise planting into rows. In this case, a spear trowel is a great implement to have on hand. Especially for overgrown seedlings that should be planted deeper than their soil line, such as eggplant and tomatoes, a spear trowel makes it easy to create a hole and drop the seedling in it with good soil contact.

Times to Drop Seedlings First and Times to Not

Sometimes it makes sense to drop seedlings into position first, then go along and plant them in afterwards. For instance onions planted into the FilmOrganic mulch with 5-inch spacing guidelines printed on it. Without guidelines though, rows that will be hoed by hand or where landscaping fabric is used also don’t need extreme precision in spacing so dropping the seedlings from standing height is okay.

In situations that need precision but have no built-in guidelines (mulch film, tiller groves, string wires, etc.), it really must be done by hand carefully and down near the ground to ensure that the spacing meets the (typically mechanical) precision required later in the season for weeding and harvesting with a specially-made tractor attachment. Even if not using tractors (which contribute to soil compaction) and still hand hoeing and harvesting, irregular spacing and/or row misalignment (a.k.a. “wonkiness”) can sometimes slow the work down.

Eggplanting

After ensuring that the landscape fabric was in place in the middle areas of the main field earlier this week, the two of us were able to plant about 5 rows while Jamie prepared beds for other things later this week. This took about 16 50-cell trays of 3 different varieties of eggplant and close to 3 hours.

After forgetting to put on gloves before using the spear trowel for a little more than the first row, I developed a nasty blister on my right hand between my thumb and my index finger where the trowel handle sat in my grip. Once I started to feel it…and it didn’t feel good…I did what I could to clean it off, then put a plastic glove over it and put my gardening glove over that. I also adjusted my grip so that my thumb was at the top of the handle while the rest of my fingers braced the bottom half, less like a barbarian hold a spear and more like a ninja might hold a weapon.

Before moving to the next task, I went up to the barn where the first aid kit lives and did a proper cleanout and re-dress with gauze before covering again with plastic then also work glove on top.

Celery Babies Leave for the Field

What once were only inch-high celery seedlings a mere 3 weeks ago, were now at a bushy 4 inches tall and ready for planting. How quickly our celery babies grew up! It almost feels empty in the greenhouse without them near the potting table, but there will always be something to take the space they once occupied.

We planted two different plots in the top main field…one right behind the new greenhouse up on the flat hill and another on the top area because Jamie was concerned that any other prepared area would be too wet. Celery are susceptible to rot if they sit in soil that is prone to pooling or poorly draining fields. Quick to grow, easy to break, sweet to slugs, and finickey around water, celery needs to be planted in just the right spot before being harvested and the space reused for something different.

The first row was short, around ~150 feet and the second was twice as long. This took about 2.5 hours and we used about 20 trays in total, where some were single plants in 98-cell trays while others were two seedlings on opposite corners of each cell in 50-cell trays (at the time of seeding, we were out of 98-cell trays).

Electrifying the Pig Fence

Like the deer and other animals, the pigs need to be ‘trained’ about the electric fence. This means a short, hopefully once-only encounter with it where they get a shock and associate that with their curiosity around the area in question.

We rode up to the pig pen and helped Jamie install a small section of electric fencing inside the pig pen so that once we allow the pigs to roam around in the forest (which is nicer for them than being stuck in a pen for the summer), the fence we use around the large forest area is effective to keep them where they belong. In about a week or so, we’ll set that wider area up. For now, the pigs get to learn that some things are for understanding and other things are just to be left alone.

End of Day Thoughts about AI in AgTech

As I watered the greenhouse, I thought about how tech and AI has already and will further influence agriculture. I thought about the contextless developers and product managers at software companies that have never worked a hard day of farm labor in their life. I thought about the startup entrepreneurs and capital investors that have no idea why agriculture is as complicated as it seems.

I think if there is a path forward where agriculture can get better, not just more profitable but easier to anticipate growths in, it’s going to need people who have walked the walk and understand how variable nature is…how it MUST remain allowed to be despite our options to abuse it for personal gain. I personally wouldn’t buy software products or services from AgTech companies that don’t require farm experience in their employees, leadership as well as individual contributors.

Main Themes:

1. Learning and self-improvement: The intern aims to improve their skills and knowledge in areas such as identifying plant varieties and problem-solving.
2. Problem-solving and resourcefulness: The intern demonstrates creative solutions by fixing a broken hose using a repair kit.
3. Exploration and education: The intern sees an opportunity to educate others about different plant varieties and their uses.

Activities Done:

• Cleaning up the greenhouse
• Filling tables with potting mix
• Fixing a broken hose at the barn
• Creating a blog post about identified plant varieties

New Things Not Yet Encountered:

• Identifying specific plant varieties and their uses
• Publishing a blog post to share knowledge with others

Questions and Future Areas of Research:

• Prioritization and task management
• Anticipating priorities and planning ahead
• Integrating learning and self-improvement into daily routines

Suggested Actions:

• Continue to work on identifying plant varieties and sharing knowledge through blog posts or other educational materials.
• Practice prioritization and task management to become more efficient in daily operations.
• Explore ways to integrate learning and self-improvement into daily routines, such as setting goals for the day or week.
• Consider seeking feedback from supervisors or mentors on areas for improvement and progress.

Part 1

Main Themes:

1. Learning and self-improvement: The intern is actively seeking to improve their skills and knowledge, such as predicting priorities, identifying plant varieties, and providing solutions for others.
2. Problem-solving and resourcefulness: The intern demonstrates problem-solving skills by fixing a broken hose using a repair kit, showing resourcefulness in finding creative solutions.
3. Exploration and education: The intern sees an opportunity to educate others about the different plant varieties and their uses, highlighting the importance of exploration and sharing knowledge.

Activities Done:

1. Cleaning up the greenhouse
2. Filling tables with potting mix
3. Fixing a broken hose at the barn
4. Creating a blog post about identified plant varieties

New Things Not Yet Encountered:

1. Identifying specific plant varieties and their uses
2. Publishing a blog post to share knowledge with others

Questions and Future Areas of Research:

1. “I mean, I don’t spend time pricking out volunteers before I do all the other obvious things like filling up the table, the potting table with potting mix, you know, cleaning up the greenhouse.” (This suggests that the intern is still learning about prioritization and task management.)
2. “How to look ahead and anticipate what those priorities are” (This indicates a desire to improve their planning and anticipation skills.)

Suggested Actions:

1. Continue to work on identifying plant varieties and sharing knowledge through blog posts or other educational materials.
2. Practice prioritization and task management to become more efficient in daily operations.
3. Explore ways to integrate learning and self-improvement into daily routines, such as setting goals for the day or week.
4. Consider seeking feedback from supervisors or mentors on areas for improvement and progress.

Part 2

Main Themes:

1. Learning and growth as a farm intern
2. Preparing for field work with necessary equipment and planning
3. Animal husbandry and training (pigs)
4. Plant propagation and maintenance (seedlings, greenhouse management)
5. Business and marketing aspects of the farm (CSA program, volunteers)

Activities:

1. Visiting pigs in a pen to train them to avoid an electric fence
2. Setting up an electric fence inside the pig pen
3. Carrying a backpack with necessary items for field work
4. Watering seedlings and monitoring their progress
5. Pricking out volunteer seeds from trays of one variety to make way for new plants
6. Switching seedlings between vehicles (van, white truck) at the barn

New Things Not Yet Encountered:

1. Using an electric fence with pigs
2. Working with a different pig farm than usual
3. Dealing with sickly or “rough” pigs

Questions and Future Areas of Research:

1. How to better prepare for field work with a backpack?
2. More information on how to train pigs to avoid electric fences.
3. Why the pigs from this new farm are looking sickly compared to previous seasons.

Suggestion Actions:

1. Consider investing in a field backpack to improve organization and efficiency.
2. Research ways to prevent volunteer seeds from getting scattered or appearing in unwanted places.
3. Create a plan for handling the “orphaned” seedlings ( volunteers) that can be potted up and sold to customers.
4. Develop a more detailed system for tracking and managing greenhouse space.

Note: These suggestions are based on the information provided by the intern and may not be exhaustive or definitive solutions, but rather starting points for further exploration and research.

Part 3

Main Themes:

1. Farm work and daily activities
2. Planting and preparing soil for various crops (eggplant, celery, head cabbage, bok choy)
3. Equipment usage (Rotary Tiller, seeder, spear trowel)
4. Safety concerns and preventive measures (wearing gloves)

Activities:

1. Potting up plants from 98 cell size to 50 cell trays
2. Watering the greenhouse
3. Greenhouse clean up
4. Filling potting mix table with potting mix
5. Bringing out and planting cosmos and zinnias
6. Planting eggplant in a field with landscaping fabric
7. Planting celery using a Rotary Tiller
8. Using a seeder to plant arugula
9. Loading up the truck with more celery and planting it in the high main field

New Things Encountered:

1. The use of a Rotary Tiller, which is an older style tractor attachment that cultivates the ground and leaves behind cultivated but slightly piled up dirt lines.
2. A seeder that looks like a bicycle walk-along seeder for planting seeds
3. The concept of using different attachments to weed and cut weeds in between plants

Questions and Future Areas of Research:

1. “I’ll have to think about it.” (referring to the name of an Asian eggplant variety) - needs further research or identification.
2. How to avoid developing blisters while planting with a spear trowel without gloves.

Suggestion Actions:

1. Wear proper gardening gloves when using sharp tools like spear trowels to prevent injuries.
2. Research and identify the name of the Asian eggplant variety mentioned in the field notes.
3. Consider investing in protective gear, such as gloves or sleeves, for future farm work to prevent similar injuries.

I washed eggs while the other four team members covered the pepper field, then we planted. Multiple varieties of hot peppers and some cherry tomatoes were sunk, then we plucked suckers and low branches from the potted tomatoes to be ready to plant them in the new greenhouse next week.

Egg Washing is a Chore

There’s no getting around it…washing eggs is a chore. The value of local farm fresh eggs in the CSA is in large part the demand for them, not the profit. Profit on a dozen of these eggs is measured in dollars or less. Anyone complaining about the prices of these eggs at $8.00 while also being willing to purchase non-local, not-so-fresh eggs at a place like Whole Foods for more than that is…disconnected. The input costs of producing eggs at small-scale include:

• space for chickens…in this case mobile coops
• the chickens themselves (this year being doubled for growth reasons)
• safe, organic chicken feed
• labor which includes
  • collecting the eggs seven days a week
  • water and food delivery, also 7 days a week
  • moving the coops to the appropriate fields, infrequent but part of yearly planning
  • processing eggs about twice a week (results in sellable product)

There are other aspects of producing-chicken lifecycle I’ve not yet been exposed to. Maybe the above is enough to provide a deeper glance into why a dozen costs what it does. Economically, eggs are not just a recent national conundrum…they’re a commoditized staple of the American diet that, like many other supermarket items, has been largely divorced of context into how it gets from farm to table.

In short, you eat what you pay for eggs in terms of quality, locality, and safety. It’s not always the case that ‘the eggs you get are the eggs you deserve’…a derivative monacum statement ‘the [fill in the blank] is the [fill in the blank] you deserve’. Meanwhile this simply does not apply for most peoples not fortunate enough to affort eggs at more than one-half a minimum wage hourly rate (after taxes of course).

It’s more like, the eggs you can affort are the eggs you’re going to eat…duh…until the floor price of a staple goes beyond affordable for most. Today we see that many people will stop purchasing eggs when they’re too expensive, either due to scarcity or reality of input cost and profit margins being toyed with. Sharp spikes in lack of demand have huge ripple effects on any industry so we’ll have to see how current events play out.

Different Team Every Day

There are a handful of people that work on the farm, each with their own schedules of which days they can/can’t work. I am currently doing Monday through Thursday 8-9 hours and only don’t know one other person. Some days, we work all together on big projects and sometimes separate individual tasks here and there. It’s all in until the job is done.

Every person brings a different kind of ‘energy’ to work. I don’t mean effort level, that’s always assumed at your best; by energy I mean vibe, attitude, approach, etc. Some days that energy contribution is different because of other circumstances, other work fatigue, or something good happened recently. When combined, it’s always a slightly different group energy.

Contributing good energy while you work together is important to farm teams because the task can be a grind. Repetitive motions, often labor-intensive or simply monotonous, over larger scales than seem doable is the hike to prepare for…going the trail with others means being ready to listen, ask questions, provide your own thoughts. There is a social aspect to field work together and the aggregate ‘team energy’ makes it either easier or harder.

Planting Hot Peppers and Cherry Tomatoes

We first planted about 6 short rows of hot pepper varieties including Jalapeño, Shishito, and Habanero adjacent to the high tunnel we weeded on Tuesday. Then we switched to planting another pepper variety into the landscape fabric already on it’s 2nd season in the middle plot with the kale on the other side of the high tunnel greenhouse in the main field. The process involves:

• weeding the hole first (easy at this time of season, not so much later)
• adding a small handful of ProGrow 5-3-4 organic fertilizer
• cultivating the planting area with a spear trowel
• dropping seedlings near the hole
• planting the seedlings in the hole

Some of these things you can combine, such as weeding and cultivating. Others take both hands such as fertilizing and planting. So it’s nice to have had 5 people working on this and we got 2 300-foot rows planted in about an hour or so.

Pruning, Plucking, and Prepping the Tomatoes

Next week we’ll be planting tomatoes in the new greenhouse and likely also starting them on string to grow upright. To prepare for this, a few of us took the last hour to pluck the first few stems and also suckers from the paper-potted tomatoes. Doing this is for a few good reasons:

• they will be planted deeply, and you generally want to avoid planting foliage in soil
• anywhere the stem touches soil, it will eventually create roots, and more roots are better in this case
• foliage contact with soil tends to attract disease and pests, and there’s no landscaping fabric in this space
• airflow is better at the ground level with less foliage, discouraging molds and wilts

Right now, I feed my LLM transcriptions of audio notes for the day I worked and plan to write about. Along with some fine-tuned templates, for the most part this works well to produce a summary in bullet point sections that I use during my human writing hour to keep to the most important things I want to capture about this work. What I write is not always the dirty (no pun intended) details of farming work, though often that’s what makes for good writing. It’s also the human meta, the essence from my perspective, of what I was thinking about and what I learned.

My solve for this will be to implement a RAG (Retrieval Augmented Generation) model that searches for the most relevant blog posts I’ve written here and that are tagged with primary keywords in the transcript to build its vector database from, cataloged in stores. Then based on the most relevant documents in the store(s), it will fill the rest of the context window with as much of the prior posts’ relevant text. Then the prompt templates will have to change to reflect that the summary of the day should focus on what happened in the transcript based on activities that day, but be informed by prior experiences in the additional context section.

Main Themes:

1. Farm work and tasks
2. Planting and preparing for growth
3. Maintenance and preparation for future tasks
4. Teamwork and collaboration

Activities:

1. Preparing hands and gloves for work with gauze and a band-aid
2. Processing eggs (3.5 hours)
3. Laying landscape fabric over the field near the high tunnel
4. Planting hot peppers in the field next to the high tunnel
5. Planting cosmos, zinnias, and other flowers in the small field
6. Planting kale and cherry tomatoes in the midfield field
7. Stake-staking cherry tomato plants with string
8. Plucking tomato suckers and branches from plants

New Things Not Yet Encountered:

1. The specific tasks of processing eggs (e.g., how to identify faulty eggs)
2. The Florida weave method for supporting cherry tomatoes
3. The new greenhouse setup, including trenches, trencher, and composting process
4. Possibly the specifics of the landscape fabric placement and soil conditions

Questions and Future Areas of Research:

1. What is the purpose of the Florida weave method for supporting cherry tomatoes?
2. How do the property owners plan to use the cosmos and zinnias planted in the small field?
3. Are there any specific challenges or concerns related to planting hot peppers in a high tunnel environment?

Suggested Actions:

1. Research the specifics of the Florida weave method for supporting cherry tomatoes.
2. Consider consulting with experienced farmers or gardening experts on the best practices for planting and maintaining hot peppers in a high tunnel environment.
3. Investigate potential uses for cosmos and zinnias beyond their aesthetic value, such as attracting pollinators or providing habitat for beneficial insects.

It was a small crew today, just three of us. We cultivated then planted zuchini and summer squash; then we processed eggs. This gave me a lot of time to think. The following is a list of the things I was involved in:

• watering and feeding the chickens, picking up eggs
• watering and feeding the pigs, checking the electric fence
• cultivating soil holes in landscaping fabric
• planting zuchini and summer squash (delicate seedlings)
• hoeing the lower field around lettuces
• processing eggs (washing, sorting, labeling)
• cleaning the greenhouse and consolidating trays
• planting peppers

A Lot of Time to Think

While working, we often talk to each other…play word games or tell stories. I try to think about what’s coming up and what to anticipate in activities on the farm, to exercise the knowledge I’ve learned and stretch my intuitions then verify them later. In between all this, I also have time to think.

I think that agriculture is in a period of transition. There are demands and pressures without proper supporting systems behind them. There’s a lack of consumer competence and interest in what really makes for high quality agriculture product vs. convenience and price. There’s an extreme deficit in the number of farmers and the number of people who are willing to work on farms. There are still new farmers coming into the industry and a lot of old farmers leaving.

There is a lot of work to be done.

Misconceptions in Organic Farming

There are misconceptions that organic farming is easy, that it’s a hobby, that it’s a lifestyle. There are misconceptions that farming is way to hard. There are misconceptions that technology has no place in natural and traditional farming, though in every century of farming, there has been ‘technology’.

The grain silo (or at least consolidating it into a space that it doesn’t rot) is a technology. The hoe is technology…every tool is. The tractor is a technology. Greenhouses are technology. Irrigation systems are technology. The repeatable and profitable process of seed saving (such as Johnny’s Selected Seeds) to sell them to organic farmers is technology.

It’s not just bookeeping and CSA member emails and social media that is the technology of farming…so the next time someone suggests that organic farming and technology are juxtaposed to each other, maybe they’re just not thinking about it creatively.

The Value of Keeping Organic Farming Low-Tech

Often the way technology influences things is by making them more efficient or affordable. This leads to their value being measured usually only in speed and cost, similar to much of the world of business. It rarely is measured in the importance of safety, physical or psychological, or how it achieves other goals.

So too when people apply technology to agriculture, they often overlook the value of keeping it low-tech. The farm I’m on has a tractor that was made in 1953 which rarely breaks down. It’s not big and can’t do the job of a huge tractor, but it does the job it needs to: sitting in the fields that need cultivation and basket weeding from time to time. It’s low fuss, low variability, low overhead.

There are hundreds of things that need to be looked after in a farm season, if not thousands. Farmers are constantly mentally reprioritizing based on time and timing, what needs to happen when. There is no computer that can do this yet. In organic farming, the point is that computers never should be in charge of natural systems. The computer is a tool, not a replacement for human decision process.

Improving the Farm

It’s not my place to suggest improvements, not what I am paid for. I am in learning mode and show up on time and work as hard as I can. I don’t have enough experience or intuition to feel comfortable suggestiong things, but I have 20 years of working with gardens so I’m not a complete novice in soil. I listen to many books about soil science and organic farming. I research things that come up in the day…I don’t just do stuff and forget about it.

At some point, I’ll have enough experience to make suggestions, but I’m not there yet. I think there are at least a dozen places where many small farms would significantly benefit from just the right amount of technology and small financial investments.

Main Themes:

1. Farming practices and techniques
2. Soil conservation and management
3. Crop care and maintenance
4. Experimentation and innovation in farming methods

Activities done by the intern:

1. Watered the greenhouse in the morning
2. Assisted with chores until 9:30 AM
3. Planted two rows of eggplants, including cultivating the ground
4. Weeded persistent grasses and a running vine through the field
5. Cleaned out holes to prepare for planting

New things not yet encountered by the intern:

1. Landscaping fabric system
2. ProGro 534 fertilizer
3. Compaction of soil due to tilling

Questions and future areas of research:

1. How to reduce compaction of soil in fields
2. Methods for amending soil with carbon-rich materials (e.g., compost, straw)
3. Effects of planting sweet potatoes directly in hay on root systems
4. Comparison of root systems between same row, same level, same everything else

Suggested actions:

1. Conduct further research on methods to reduce compaction of soil
2. Experiment with amending soil with compost or other carbon-rich materials
3. Consider planting sweet potatoes directly in hay and monitoring its effects on root systems
4. Discuss and plan for future planting experiments, including the possibility of mixing straw with the dirt

Part 2

Main Themes:

1. Farm work and daily tasks
2. Research and experimentation (limited to small-scale)
3. Harvesting and processing eggs
4. Managing farm operations and logistics
5. Weather conditions and their impact on farm activities

Activities done by the intern:

1. Fixed holes in landscaping fabric with straw or compost
2. Loaded truck with peppers
3. Fixed additional landscaping fabric for eggplant planting
4. Planted short row of peppers after lunch
5. Washed and processed 60 dozen eggs (approximately 3.30 hours)
6. Labeled green buckets/bins with week numbers for CSA
7. Found four more trays of peppers for the final row
8. Ran back to barn to set traps and drop off garbage/recycling bins

New things not yet encountered:

1. Amending landscaping fabric with slightly loamy material to loosen soil
2. Processing 60 dozen eggs in a short time frame (approximately 3.30 hours)
3. Specific challenges with egg carton closure due to large egg sizes

Questions and future areas of research:

1. How effective is amending landscaping fabric with loamy material?
2. How does this method impact plant establishment and growth?

Suggestion actions:

1. Consider conducting a larger-scale experiment on amending landscaping fabric for more data.
2. Continue to monitor egg production and processing times to optimize efficiency.
3. Investigate ways to improve egg carton closure due to large egg sizes.

We planted peppers, spread the last of the landscaping fabric, staked tomatoes, and celebrated our farmer’s birthday. It was a good day.

Preparing a Field for Peppers and Tomatoes

Before planting, we still had to spread the last of the landscape fabric for the season over some cultivated and hilled rows. On a windy day this is a bit of a challenge, but with proper communication and a little teamwork, we were able to get it done. Most of this was new fabric, but on many of the fields we have used fabric from previous seasons.

After making holes in the fabric we planted three rows of Carmen peppers which take about two to three months to mature, so the first of these will likely show up for CSA members in late July but more likely early August. Even though the soil had been recently cultivated, it took an honest amount of hand troweling to cultivate the hole and get the seelings planted.

Staking Tomatoes

The tomatoes we planted on Monday in the other field needed stakes so two of us went to where the composting shed is located on the other side of the pond and loaded about 100 stakes into the truck. Then we drove back over to the main field and staked the tomatoes already planted. The rest we dropped off at the newly landscaped rows. Three of these rows will be for the peppers and three for field cherry tomatoes.

We have yet to do the Florida weave with twine on the staked tomatoes but likely this will happen tomorrow or Friday since we don’t want floppy, thin young seedlings being whipped around by the wind gusts without support for long. There were only one or two of the taller seedlings that got kinked a bit and we always have extras to fill in gaps or failed seedlings so its not a big deal.

Celebrating Farmer’s Birthday

At lunch, the admin staff including Glen the owner and the groudskeeping crew joined us on Jamie’s deck for pizza and an ice cream cake. Another team member also brought homemade sushi so it was a really nice lunch. Despite knowing that Jamie is humble and it might be a bit awkward, I chanced it to say something nice about his approach to making space for learning and improvement in his field team and how he is a servant leader. It wasn’t awkward and I think he appreciated it.

Planting Peppers and Hand Fatigue

We finished the field day by planting three rows of peppers. I did one of them which wasn’t hard except that we’ve been planting daily for about a month now, and last week I was still learning the cons of not having the right gloves to protect my hands. I have a few blisters and my hands are sore, but there are ways around that too. Mostly I just need to have a variety of gloves for different tasks and be prepared to use them when necessary.

Short Day for Me

A Roadmap for My Personal LLM Work

I also want to standardize how these various workflows and services notify me…likely via my private Slack workspace in channels there; but for this, I’d like some sort of passive listener to various container logs to extract events rather than having to cobble together an events pub/sub backplane. Though logs-listening is an older two-hop hack, it a known and well traversed industry path. Modern observability practices encourage directly emitting specific events so that there’s no mapping, interpretation, or translation required…all of which incur overhead either of delay or cost. Maybe I’ll try both approaches and learn my own lessons, none of which an AI-driven answer can provide the confidence I’d like.

Part 1

We planted tomatoes, lettuce, and fennel. We trellised tomatoes using the Florida weave method, and had a coyote encounter.

The Last of the Field Tomatoes Planted

We planted in various cherry tomatoes in the field prepped with the new landscape fabric earlier this week. Tomatoes also got planted in one other row higher up on the hill in the middle of the field. After planting, we trellised the tomatoes using the Florida weave method.

Also known as the ‘basket weave’ or ‘stake and weave’ method, this is a way to support the tomatoes throughout their lifecycle. You basically pound stakes about 5 feet apart into the ground, then tie twine to them down the row. You start by knotting the first stake at around 4 inches from the soil, maintaining tension on the twine as you loop it around the next, then repeat all the way down one side of the row. Once you get to the end, you go back and do the same thing on the other side. Ideally, this should result in each plant between the twine, but sometimes plants still need to grow up a bit.

While we were doign this, we notice that the older of the two little farm dogs was barking up a storm in the lower field where the chicken coop is. We went to check it out and saw a coyote running away from the dog. We also saw it later up on the hill towards the barn house. Aside from coyote deterence, I’ll also have to look up mice trapping methods because there a few mice in the barn house that traditional mouse traps aren’t catching and Jamie asked me to look into it today…something like the bucket and bottle trap.

After cultivating and planting, pounding in stakes, bending hundreds of times to weave the twine…my body was ready for a break at lunch.

Lettuce and Fennel

After lunch, we planted two long rows of lettuce…I and one other dropped and two others planted the most of the rows because at the top of the main field are new beds that are prepared with the Allis-Chalmers tractor so the planting is fast and easy.

We also planted a row of fennel. Fennel is a great addition to salads, meat and fish, even sandwiches. It’s a bit of a delicate plant, grown tall in the greenhouse with whispy tops that weigh down when watered, but they always spring back upright once dry. Planting them in another prepared bed took barely 20 minutes with the four of us today.

Things from AI You Can’t Make Up

Today’s AI-generated summary of my field notes includes a few…aberrations. Incorrect spellings of proper nouns I get and are easily remeded, but the following are some examples of things that are just plain weird:

So clearly, there’s room for semantical improvements to my summarization process. This whole notes-summary thing is simply my muse for learning by combining hands-on context with its digital process-based counterpart to produce a more complete, more accurate, and more useful account of my time on the farm.

In particular to improve the “New Things Not Yet Encountered” section, it’s likely that I’ll have to extract and normalize ‘things already encountered’ from the notes in some way that is easy for new summaries to quickly reference. I’m hoping that the RAG by tags approach discussed in prior post will be sufficient, but possibly I’ll have to use a proper and more powerful search engine or method to do this.

Main Themes:

1. Farm work and tasks
2. Internship experience and learning
3. Interactions with animals (Jenkins the dog and chickens)
4. Equipment usage (Alice Chalmers tractor and basket weeder)

Activities done by the intern:

1. Harvesting from certain areas of the farm
2. Helping to take apart Florida weave and stake trellises
3. Inspecting the chicken coop
4. Loading the truck with lettuces
5. Planting in newly cultivated soil (lettuce and fennel)
6. Taking trash up to the barn
7. Loading the van for the farmer’s market

New things not yet encountered:

1. Florida weave and stake trellises
2. The Alice Chalmers tractor from 1953
3. Chicken coop inspection
4. Lettuce planting in newly cultivated soil

Questions and future areas of research:

1. “What if we had known about the coyote earlier?” ( suggesting a potential future study on wildlife detection or alert systems)
2. “How would it have been different if Jenkins the dog had alerted us sooner?” (suggesting a potential future study on animal behavior or training)

Suggested actions and suggestions:

1. Consider implementing a more effective coyote detection system to prevent future encounters.
2. Research ways to improve the efficiency of lettuce planting in newly cultivated soil, such as optimizing crop rotation or soil preparation methods.
3. Document Jenkins’ dog behavior and training methods to understand how he was able to detect the coyote so effectively.
4. Consider adding a section on animal welfare and safety protocols to future field notes or research papers.

Part 2

Main Themes:

1. Plant care and cultivation
2. Environmental considerations (e.g., watering, soil conditions, pests, disease)
3. Farming techniques (e.g., stake installation, Florida weave)
4. Crop management (e.g., alternating rows of tomatoes and peppers)

Activities done by the intern:

1. Watered the greenhouse, including the insides and outsides
2. Transplanted seedlings from the tray to the field
3. Cultivated the holes in landscaping fabric
4. Planted three rows of tomatoes near peppers, with alternating rows for airflow and pest control
5. Installed stakes in the ground, spaced 5 feet apart, and pounded them into place using a pole pounder
6. Set up the Florida weave by tying plastic string around each stake to keep tomato plants upright

New things not yet encountered:

1. The concept of transplant shock, where seedlings experience stress when moved from one environment to another
2. The use of landscaping fabric as a medium for planting
3. The technique of using stakes with the Florida weave to support tomato plants
4. The idea of alternating rows of tomatoes and peppers to control pests and diseases

Questions and future areas of research:

1. The intern asks to confirm with Jamie that they did alternate rows of peppers and tomatoes due to airflow considerations.
2. The intern mentions the need for further research on how to deal with pests and diseases in the field.

Suggestion actions:

1. Follow up with Jamie to confirm the reasoning behind the alternating row pattern.
2. Conduct further research on optimal spacing and layout for tomato and pepper plants to minimize pest and disease issues.
3. Continue monitoring and adjusting stake installation and Florida weave techniques as needed to ensure the best support for the tomato plants.

Additional notes:

• The intern’s detailed observations and explanations demonstrate a strong understanding of plant care and environmental considerations.
• The use of specific terminology (e.g., “Florida weave,” “transplant shock”) suggests that the intern has some prior knowledge or training in horticulture or agriculture.
• The intern’s enthusiasm for learning and their willingness to ask questions suggest a strong potential for future growth and development as an agricultural professional.

Today we prepared for 50mph gusts, I processed 40 dozen eggs, and we cleaned up the barn in prep for power washing next week. Very little time for photos, but there are plenty from other days to appease the visually inspired. It was also a short day and I was able to go home early to work on tech stuff.

High Winds in New England Are No Joke

We had a 50mph gust today, so we had to prepare for it before the rain came later. We put up the row cover on the high tunnel, and went through the field to finish shoveling in and weighing down row cover and landscape fabric.

All it takes is a few areas and uncharacteristically strong winds to rip the row cover off a bed the team spent hours planting and covering. Worse than a row or two, high tunnels and greenhouses that aren’t properly secured can be destroyed by a gusty day. So Jamie and two of us assisted him tying up the sides of the high tunnel 3mil poly to the ridgeline so that the wind would just blow through the tunnel instead of ripping it apart.

In my own greenhouse on my deck at home, the winter winds the first year ripped a few panels off before I got wise and belted them down with some climbing gear webbing, gromets and caribeners. I haven’t had a problem since besides a little rattling on the extremely windy nights.

Eggs Are What an Intern Deserves

Processing eggs is boring. It’s not hard, but it’s not fun either. It’s just a chore that has to be done. I’m not complaining…there are many much harder, dirtier tasks on the farm for sure. And it’s not economically high-value either, but there are many other reasons than profit which make eggs in the CSA a necessary thing to do.

I like to think that people who buy and eat these eggs value that each one is hand-washed and inspected for imperfections. They are locally produced every day by a bunch of ‘ladies’ (laying hens) who are treated well and have a good life, aside from the occasional visit from a coyote. Each one is different in size, shape, color and freckling. They are worth every penny that goes into making them and every other penny we don’t make on them by keeping them at market price despite their vastly superior quality.

They say that it takes 10,000 hours to master a skill. I’m about 40 hours in and there are still things I could improve…certainly the speed of my hands, but also minor process tweaks and improvements.

Cleaning Up the Barn

About a week before the first CSA bags go out, there’s a lot being done and still to do. We’ve been cleaning up the barn and getting it ready for power washing, and a lot of cleaning and organizing. Before any of the harvested greens that are growing nicely in the field right now are brought up to the barn, bins have to be thoroughly cleaned, all surfaces have to be sanitized, and some new adjustments to ‘standard operating procedure’ have to be made. Then the team needs to be briefed on the new procedures and process in general so we’re all on the same page.

After doing a bit of this cleanup today, we went back to the greenhouse to check in with the others and see how the seeding is going. At this point, there wasn’t much more for a team of 5 to do and I had the opportunity to go home early.

This AI Summary Desperately Needs Improvement

I won’t go into why or what next as I’ve done in every other post so far, but every day I do this, I’m more annoyed with the AI summary. Even ‘mostly accurate’ is not good enough with auto-summarization. Details matter. Blatently wrong things like “…implied that the intern is reflecting on their first day of work…” are remedied by including context (via a RAG processor). It’s all a matter of iteration and time to do so, which I have very little of due to all the other things in my life right now. So this is what I get for now:

Summary

Two of us washed 43 dozen eggs in about 2 hours, then we planted dahlias and (finally) some tomatoes in the new greenhouse! Due to some interview prep work I want to do this week, I only was able to be on the farm this morning, not the whole day.

Enough Already about Eggs

In prior posts, I’ve already said enough about the process and frequency of how we wash our eggs. The only thing I’d add today though is that it goes SO much faster with another person. It reminds me of ‘optimal batch size’ from the agile software days…or more like ‘optimal worker count’ in lean manufacturing. Having three people help with eggs is not as productive as just two, but only because our ‘queue length’ (i.e. number of active washing buckets at one time) is two and we only set up one polishing and packing station. Anyway, eggs are done until Thursday.

Planting Dahlias

As part of the produce for pop-up markets and the new general store in Essex later this year, we planted about 10 dozen Dahlia bulbs in a smaller plot on the property near the owner’s farmhouse. The four of us farm team members and four more from the groundskeepers crew made quick work of cultivating pre-made holes in landscaping fabric and planting the bulbs about 6 inches deep with 2 inches of soil back over the top of them. I can’t wait to see them shoot up tall in the coming months.

Flight of the Tomatoes

Finally, this week we started in on planting the rapidly overgrown Hot Spike tomatoes in the new greenhouse. A month ago, we used a trencher to create 8 trenches and then put a truck load of compost in them with some ProGro 5-3-4 fertilizer. It was so easy to plunk the 3 foot plants in these beds and the plan is to wind them around twine hanging from wires stretched the long way from the ceiling structure.

These tomatoes were one of the first things started early in the year, taking over much of the growing greenhouse space and making watering other things around them somewhat cumbersome. Not so much anymore…they’re quickly taking flight from the nursery to their final seasonal destinations.

And they’re not the only things growing up…

My Personal AI Infrastructure Got a Trim and a Shave

If this is your first time reading this blog, I regret to inform you that I’ve been coding since I was 9. So while I love farming and agriculture, I also like to improve life with technology. There are so many places where modern life is harmed by added technology, but I’m on a mission to find truly useful and non-toxic places where technology and agriculture overlap. Blending these two things in appropriate ways is a challenge, but at least to accelerate this blog and my learning journey, I’ve been using AI to help me.

This is still not informed by prior blog posts, but is getting better based on another AHA moment I had using a coding assistant over the past week:

Added Reasoning to LLM Summaries

During the containerization process, I also parameterized which model and server to use for the Langchain summarization process. When running as a pod in Kubernetes, this process now uses an Ollama instance running on the same cluster. Since this cluster doesn’t have any fancy GPU nodes, it runs inference on CPU cores instead…which take much longer than with a GPU, but I can live with that for now and add a specialized node to the cluster later.

Using a Coding Assistant to Pipeline Builds

At first and based on some shoddy OSS codebase in openai-whisper that pulled in a ton of unnecessary dependencies, the image sizes were around 4GBs which wasn’t ideal. I tried a bunch of things out to no avail:

• reducing my multi-platform buildx commands to just AMD64 to match that of my Kubernetes cluster
• compiled ‘whisper.cpp’ from source in my base image only to find the same bloat…due to compiling on a mac using QEMU and aarch64 under the sheets
• compiled ‘whisper.cpp’ from source on the target cluster architecture directly…smaller, but layers need lots of cleanup

In this process, rather than incurring a complete soup-to-nuts build of everything all in one Dockerfile, I decided to break things out into separate images for each major stage of the build. That way, I can rebuild just the top layer when the app code changes and leave all the dependencies that didn’t change in the base layer.

Finally, I decided to use SourceGraph Amp (that I’m doing an interview demo about on Friday) to help write a .Gitlab Pipeline (gitlab-ci.yml file) to run a ‘docker buildx…’ for each major stage of the build. It got almost everything right, and when I asked it to correct the script to use ‘docker buildx…’ instead of classic ‘docker build’ commands, it updated it flawlessly.

And voila! The image sizes are around ~400MB (mostly due to Python and dependencies) and the build time is around 1 minute (once the build cache is populated first go around). This was where I wanted to be because all I want is to make rapid changes to the autoscribe codebase and just bump the deployment manually when I want to test a new version. Later I’ll probably use GoCD or some GitOps overlay to automate the deployment on updates to the main branch.

What My New (Still Not RAG) AI Assistant Thinks

Consolidated Summary:

Main Themes

1. Farm Labor: The intern engaged in hands-on tasks such as washing eggs, planting tomatoes, and Dahlias, emphasizing essential labor in farm operations.
2. Teamwork: Collaboration was evident as multiple individuals worked together on various gardening tasks.
3. Attention to Detail: Specific practices like trellising and deep planting were highlighted to enhance plant growth.

Activities

• Washed 53 dozen eggs.
• Assisted with watering the tomato greenhouse and checking its new water pump.
• Planted hot spike tomatoes with a group, noting that they weren’t trellised yet.
• Collaborated to plant Dahlias through landscaping fabric, focusing on correct planting depth.

New Things Not Yet Encountered

• Greenhouse Projects: The intern encountered a new greenhouse and its water pump setup.
• Dahlias: A specific plant species requiring particular attention for optimal growth.
• Gardening Techniques: Deep planting methods for vegetables like tomatoes and Dahlias, focusing on soil preparation.

Questions and Future Research Areas

• Improved trellising techniques to prevent damage and enhance plant structure.
• Optimal methods for planting and growing Dahlias.
• Understanding irrigation systems’ impact on plant growth.
• How different plants respond to soil conditions and planting depth.

Suggested Actions

1. Trellising: Implement carefully to avoid damage and improve plant structure.
2. Planting Depth: Ensure tomatoes and Dahlias are planted deep enough for healthy root extension.
3. Collaboration: Maintain teamwork to efficiently manage farm tasks.

This summary encapsulates the intern’s experience, highlighting both labor-intensive tasks and the nuances of gardening, while suggesting areas for further exploration in farming practices.

Part 1

It was just two of us today, so after greenhouse watering and chores we focused on setting up trellis twine for the tomatoes in the new greenhouse.

Watering and Chores and Pigs First

Since it was a small team of two and every day there are basic chores to do, we got to those first. While I deep watered the greenhouse, Jamie did some other daily chores like collecting eggs from the hens, then we fed and brought water to the pigs together.

Trellis Twine for the Tomatoes

After that we focused on setting up trellis twine for the tomatoes in the new greenhouse. We put up poles, used ladders, drilled holes, installed twine loops every couple of feet, and attached clips for the tomato plants. We managed to do one row before running out of time, then took a time-lapse video of the setup.

As we worked, we talked about things…the nature of long-term farming, the importance of planning and preparation, and the challenges of adapting to changing circumstances. Specifically “time and timing” is what Jamie says is one of the biggest challenges in managing all the aspects of the farm.

The timing of getting these tomatoes trellised couldn’t be better, despite all of us who water the greenhouse constantly asking Jamie “when will we move these tomatoes to the new greenhouse?” for the past month. The fact is, these lovely plants are doing just fine and will grow to produce enormous amounts of delicious tomatoes throughout the season. This isn’t the farm or farmer’s first rodeo and anything could happen thanks to nature, but I predict a very happy CSA season for everyone involved.

AI Summary from Field Notes

Today with the help of SourceGraph Amp, I adjusted last week’s ‘new_post.sh’ script to adjust the date format, add an appropriate slug to the frontmatter, and add interactive prompt defaults for other inputs. It didn’t make a mistake syntactically, but procedurally I forgot to add context to make the slug from the user-inputted title and not the prefixed state. Shame on me, but it was super-easy to fix.

Summary

We harvested spinach and radishes for Week 1 of the CSA, then trellised tomatoes in the new greenhouse.

Washing Produce Is Wet

I started the day by scrubbing totes in preparation for storage of freshly washed produce later on. When done without wasting time, this is a great way to get drenched. However, with my galoshes and overalls on, this wasn’t too much of an issue. Everything must be scrubbed with soap and water, then rinsed with clean water. What we bring out to the field must be sprayed with a light acidic, organic safe water solution to prevent bacterial growth.

Harvesting Spinach and Radishes

Then we harvested spinach from the high tunnel. Spinach is best harvested when the plants are young and tender. All in all, we got about two totes from one row.

Then we moved on to harvesting radishes. After about 20 minutes, we had enough to fill one big round field tote and including the spinach it was enough for me to go back up to the barn to start washing and refrigerating. We got about four totes of beautiful radishes to include in this week’s CSA shares.

Trellising Tomatoes

After lunch, we moved on to trellising tomatoes. We used twine and poles to create a trellis system for the tomatoes to grow on. There are six rows with approximately 35 plants per row, making a total of ~ 200 plants.

I cut and tied loops at the top of twine lengths while Jamie installed the rest of the poles to tie them to. Last week, the two of us did this for the first row, but with a full crew today we finished the other five rows. Many hands really make quicker work of something like this.

There were a few gaps and some of the tomato plants needed to be replaced for one reason or another. Each was secured with one biodegradable tomato clip and carefully wound around the twine (or vice versa) to further grow upright.

For the three-ish hours we worked on this, the new water pump was sending water through the drip irrigation lines that we had installed last month. Jamie said that a few hours every few days is what the tomato plants need…particularly since tomato plants love deep waterings and not every day which make their roots go deeper and farther to find water.

After that, I moved to watering the nursery greenhouse which was watered after lunch, but since the sun was out today it was needed to ‘top off’ the plants until tomorrow morning.

My AI-driven Process Is Getting Better

After a bunch of personal tech work last week to get familiar with SourceGraph Amp, I was able to break various workflow components out into separate microservices that handle the following concerns separately:

1. ‘gauth’: Google Authentication Service Token management
2. ‘gdrivenotes’: Google Drive searches
3. ‘notesum’: Consolidation of docs and summaries into single, simple JSON result set
4. Updates to utils/new_blog.sh script to use new microservices

Since I’m always in prototype mode to some degree, it’s better that I break out these services into small and reusable parts of the overall architecture so that I can more reliably test and run them.

I’ve already started to further encapsulate the functionality of providing an LLM template and supporting content into a single service which can be used by other services. I call this ‘llmbo’ and it will eventually replace some of the functionality in ‘autoscribe’.

I also need to do this because I want to decouple the GPU-intensive LLM inference from any infrastructure that runs workflow orchestration steps. My hope is to find a cost-effective way to run LLM inference on a GPU without having to pay for a fancy (and demand-driven overly expensive) card or external GPU device. I recently came across this article which describes methods for exposing embedded GPUs on Intel and AMD machines using kubernetes node feature discovery.

After Wednesday’s follow-up interview, maybe I’ll get some time to do that this weekend. For now, transcription and summarization, not to mention any additional LLM-powered improvements like auto-generating default titles, description, and tags…sigh…will have to wait.

Main Themes:

1. Harvesting and Produce Management: The intern focused on harvesting spinach and radishes, emphasizing their quality and proper storage methods.
2. Greenhouse Operations: Tasks included setting up plant supports using twine and poles, maintaining greenhouse structures, and assisting with ventilation panel fixes.

Activities Performed:

• Harvested and washed spinach and radishes.
• Set up trellises for tomato plants in the greenhouse using twine and poles.
• Assisted in fixing ventilation panels and monitored greenhouse conditions.
• Washed produce and stored it in green totes with specific stacking techniques to maintain air circulation.

New Experiences:

1. Trellis Setup: Utilizing twine for plant support in a greenhouse.
2. Produce Storage Method: Stacking radishes in green totes with air holes for optimal storage.

Questions and Future Research Areas:

• How effective is the current washing method for produce? Are there more efficient techniques?
• What impact does stacking have on radish storage life and freshness?
• Are there better alternatives to twine for supporting plants, such as wire or stakes?

Suggested Actions:

1. Explore Alternative Plant Supports: Consider materials like wire or stakes for more durable and adjustable supports.
2. Test Washing Techniques: Implement trials to evaluate different washing methods for produce efficiency.
3. Optimize Radish Storage: Investigate if stacking is the best approach or if other methods could be more effective.

Part 1

Based on the field notes provided, here’s a breakdown of the main themes, activities, new experiences, questions, research areas, and suggestions:

Main Themes:

1. Farm Maintenance Tasks: The intern performed routine tasks such as checking irrigation lines and cleaning gates.
2. Animal Care: Monitoring chicken eggs for quality and health.

Activities Performed:

• Checking Irrigation Lines: Inspected for proper functioning to ensure optimal water distribution.
• Moving Compost Piles: Assisted in organizing compost materials for future use.
• Cleaning Gates: Maintained fences and entry points by cleaning gates.
• Monitoring Chicken Eggs: Collected eggs and observed their quality and health.

New Experiences:

1. The intern encountered composting as a new experience, learning about its benefits and application in farm management.

Questions Raised:

1. How long do chickens continue to lay eggs at peak production?

Future Research Areas:

• Investigate the duration for which chickens lay eggs at their most productive stage.
• Explore advanced techniques for efficient composting and nutrient cycling in farming systems.

Suggestions:

1. The intern suggests seeking guidance from experienced farmers or agricultural experts to refine tasks like compost management and chicken egg monitoring.

If you have further questions or need additional details, feel free to ask!

Part 2

Main Themes:

1. Harvesting and Produce Management: The intern discusses harvesting spinach and radishes, focusing on their quality and proper storage methods.
2. Greenhouse Operations: Includes tasks like setting up trellises for tomato plants and maintaining greenhouse structures and ventilation systems.

Activities:

• Harvesting produce (spinach and radishes).
• Washing and storing produce in green totes with specific stacking techniques to maintain air circulation.
• Setting up plant supports using twine and poles in a greenhouse.
• Watering seedlings and monitoring greenhouse conditions.
• Assisting with greenhouse maintenance, such as fixing ventilation panels.

New Things Not Yet Encountered:

1. Trellis Setup: Using twine to support tomato plants in a greenhouse.
2. Produce Storage Method: Stacking radishes in green totes with air holes for optimal storage conditions.

Questions and Future Research Areas:

1. Washing Efficiency: How effective is the current method of washing produce? Are there more efficient techniques?
2. Radish Storage Impact: How does stacking radishes affect their storage life and freshness?
3. Plant Support Alternatives: What are better alternatives to twine for supporting plants, such as wire or stakes?

Suggestion Actions:

1. Explore Alternative Plant Supports: Consider materials like wire or stakes for more durable and adjustable plant supports.
2. Test Washing Techniques: Implement trials to evaluate the effectiveness of different washing methods for produce.
3. Optimize Radish Storage: Further investigate if stacking in green totes is the best approach or if other methods (e.g., using specific containers) could be more effective.

This analysis provides insights into the intern’s experience and highlights areas that could benefit from further exploration and improvement.

I washed eggs, totes, lettuce, bok choy, broccoli raab, kale, and just about everything I could see today. It’s the first week of CSA so totes are at a premium, but the produce is also premium.

Washing is (Almost) Everything

Sanitization is a key part of what keeps your produce worth buying. While the fields themselves are free from pesticides, bugs and debris still make their way into harvested produce. Pollen and dirt also are abundant in vegetables that come from the field, so washing before storing is important to make sure the CSA and market goodies get to people in good condition.

Everyone should wash their produce, whether it comes from a local farm or from a big supermarket. You never know what something from hundreds of miles away has been through, despite the chemicals used to keep it fresh over long journeys. Better to keep it local, recently harvested, and washed a little under your own control.

So while I was washing eggs, others were hand-scrubbing totes used with each CSA share member’s pickup. After I was done, I joined in. Corners, edges, crevices, holes, nooks, and crannies. Jet spray first, then soap and water with a scrub brush, sometimes a soft sponge, rinse, spray-down with sanitizer, rinse again, and then air dry. It is a lot of work, but an important part of running a farm that produces such high-quality organic produce.

For the most part, today was up at the barn doing most of the washing while three others were in the fields harvesting lettuce, bok choy, broccoli raab, and kale for this first week of our CSA.

Adjusting the Harvest to Upcoming Heat Wave

The temperatures for the past month have mostly been what one would expect for May in New England, despite an overall early and warm start to the season. Except for one or two super-early heat spikes, the next few days will be the first over 80°F days of the season.

While Jamie and another loaded up the truck and went to the pop-up farmer’s market in Essex, the other two of us harvested more bok choy for the last hour or so of the day. This was because it had already reached maturity and the likelihood of the choy bolting the next few days was high, so we took 200 of the biggest heads from almost one whole row. It’s better to harvest and store for a few days, maybe sell to local green grocers than for the whole row to bolt and become bitter.

After hauling them to the barn, we washed them and secured their space in the walk-in refrigerator. You can’t get fresher than this unless you live or work on a farm, certainly not at Whole Foods.

It’s a bit weird to think that I’ve made it so easy to turn voice notes taken on a drive home into a process-oriented capture of my day, augmented by AI transcription (private) and summarization (below). By the time I get to writing the human part of this blog, ‘new_post.sh’ script backed by a small battalion of microservices I’ve built over the past month, is already armed with enough information to create the skeleton of the post and then I can focus on my own flavor to add.

At some point, I’ll probably play around with comparing my own writing to that of what a properly-trained LLM could generate, but that won’t be for a while I think. I write because I’m human…I don’t think of it as a cost center, but one of value, even if that’s just for myself.

For today, no significant changes to the LLM templates or accuracy, so this is what it produced:

Consolidated Summary:

Main Themes:

1. Post-Harvest Handling: Emphasis on providing shade and immediate care to prevent stress after harvest.
2. Logistics Management: Utilization of blue buckets for efficient transport to shaded areas.
3. Market Timeliness: Ensuring fresh produce is delivered timely to meet market demands.

Activities Performed:

• Harvesting lettuces from the field.
• Loading into blue buckets using a truck.
• Transferring to the barn for washing or shading.
• Keeping lettuce in shade to prevent heat stress.

New Experiences:

• Specific use of blue buckets and post-harvest handling processes, which may be unfamiliar to those without farming experience.

Questions and Future Research Areas:

1. How long can lettuce remain fresh after harvest?
2. What are the most efficient transport methods to maintain quality?

Suggestions for Actions:

• Implement faster harvesting techniques.
• Invest in shade structures or covers for transport vehicles.
• Explore alternative storage solutions to enhance freshness retention.

This summary captures the essential aspects of the field note, highlighting key processes and considerations for managing produce post-harvest.

Part 1

Main Themes:

1. Adaptability in Farming Tools: The use of versatile containers like flat totes demonstrates how farming tools can be repurposed for different tasks based on seasonal needs.

2. Efficiency in Produce Handling: The emphasis on quick washing and harvesting highlights the importance of efficiency to maintain produce quality and meet market demands.

3. Seasonal Adjustments: The adjustment in tool usage from tomato harvesting to CSA pickups reflects the dynamic nature of farming practices across seasons.

Activities:

• Washing eggs and cleaning buckets.
• Harvesting and washing various produce (lettuce, bok choy, broccoli rabe).
• Collaborative harvesting of bok choy with another person.
• Managing tasks for CSA members.

New Information:

• The use of flat totes for both harvesting tomatoes and bulk pickups by CSA members.
• The impact of heat on bok choy flavor when not harvested promptly.

Questions and Future Research:

1. How do the flat totes compare in effectiveness to other containers?
2. What methods can be employed to wash produce quickly without affecting quality?
3. How does the timing of harvest affect the taste or marketability of crops like bok choy?

Suggestions for Action:

1. Invest in efficient washing systems or tools.
2. Train staff on efficient harvesting and washing techniques.
3. Explore alternative container types that might offer better utility.

This analysis captures the key aspects of the farm note, focusing on practical tasks, adaptability, efficiency, and the need for future research to enhance farming operations.

Part 2

Answer:

Main Themes:

1. Post-Harvest Care of Lettuce: The intern emphasizes the importance of providing shade and immediate handling to prevent lettuce from becoming stressed after harvest.
2. Logistics and Transport Methods: Efficient transport using blue buckets and ensuring produce is moved to shaded areas quickly is highlighted.
3. Meeting Market Demand: The ability to deliver fresh produce to market efficiently is noted as crucial.

Activities Performed:

1. Harvesting lettuces from the field.
2. Loading harvested lettuce into blue buckets and a truck for transport.
3. Moving produce to the barn for shading or washing.
4. Ensuring lettuce is kept in shade after transport to prevent heat stress.

New Experiences:

• The specific use of blue buckets and the process of transporting and storing produce post-harvest, which may be unfamiliar to those without farming experience.

Questions and Future Research Areas:

1. How long can lettuce be stored after harvest before it becomes unsuitable for market?
2. What are the most efficient methods for transporting harvested lettuce to ensure quality retention?

Suggestions for Actions:

1. Implement faster harvesting techniques to reduce stress on produce.
2. Invest in shade structures or coverings for transport vehicles to protect produce during transit.
3. Explore alternative storage solutions that maintain produce freshness and reduce stress.

This structured approach ensures that the key points are addressed, providing a comprehensive understanding of the intern’s activities and implications.

It was a very hot day, so we weeded in the morning, harvested before noon, and seeded under the shade in the afternoon.

Weeding by Hand vs. Tractor

The carrot beds had been prepared using the Allis Chalmers tractor, then seeded with a roller three rows across. This tractor can also weed but only if the weeds are young enough. These weeds quickly grew so the only thing to do was hand weed the row. Sometimes even if there is a better way to handle something on a farm, you don’t always have the luxury of time or timing for it to work optimally.

Washing Lettuce and Bok Choy

Just about every day of the season, something is being harvested…which means it has to be washed and packed and refrigerated. This also means that we have to set up a washing station and today Jamie had a good idea to use the tent for farmer’s markets over the space so that in the extreme heat, it wasn’t also sunning down on the station and veggies.

More Weeding, Onions and FilmOrganic

Thenll six of us went to weed the onions in the top main field where FilmOrganic mulch is used similar to the dry fields. Their stalks were looking particularly fresh and tall, and only in the sections where the biodegradable plastic had been lost to damage were the weeds any trouble. We got done with two 300-foot rows in about an hour.

Packing for Market and Deliveries

After lunch, we all went back up to the barn to help pack and load up the truck for deliveries as well as the Ipswitch farmer’s market. Some of the CSA shares have special instructions, such as an extra dozen eggs per share, so we had to make sure to include these special requests.

Seeding Lettuces, Pumpkins, and Squash

Two of the team then proceeded to the farmers market while another did deliveries, which left the other three of us to seed about 30 trays of lettuces, pumpkins, and squash. Since the greenhouse can get very hot on a day like this, we decided to set up a table outside in the shade and do the seeding there.

Squash and pumpkin seeds are much easier for me to handle than lettuce seeds, so I was able to get through those 14 trays pretty quickly.

AI Isn’t Really the Worst Part of the Hiring Process

Unfortunately, I did not get the job at SourceGraph. Even though I did a pretty good job of the demo interview, I think they had other candidates already further down the process and with characteristics that met their criteria. This is disappointing for many reasons, notwithstanding had I got the job it would have meant free usage of their AI-assisted coding agent.

I really wish it had gone different and I’m still waiting for detailed feedback, but few and far between are people who will spend any time providing this to candidates who did not make it. So I guess I’ll be contracting, searching for work, and farming for a while still. It could be worse.

Consolidated Summary of Farm Notes

• Main Themes:
  • Manual labor tasks, including weeding and seeding.
  • Teamwork through collaborative group activities.
  • Organic farming practices using biodegradable mulch (Film Organic Mulch).
  • Crop management involving diverse plants like pumpkins, squash, and lettuce.
• Activities:
  • Hand weeding of carrots and onions.
  • Assisting with loading produce for the market.
  • Cleaning work areas post-lunch.
  • Seeding in the greenhouse using specific crops and tray sizes (96-cell and 50-cell trays).
• New Encounters:
  • Use of biodegradable mulch.
  • Seeding methods using different tray sizes.
• Questions for Future Research:
  • Understanding why biodegradable mulch is ripped up and its impact on growth.
  • Determining optimal practices for seeding in different tray sizes and suitable crops.
• Suggested Actions:
  • Explore tools to improve hand weeding efficiency.
  • Investigate alternative materials or planting arrangements if mulch issues persist.

Part 1

With just three of us, on a stormy morning, we processed 48 dozen eggs, harvested and washed more bok choy, kale, and lettuce for the CSA, then planted peppers in the new greenhouse.

Washing All the Time

We wash, and wash, and wash. We washed 48 dozen eggs, washed bok choy, kale, and lettuce, and washed peppers. We also wash the tote bins that anything goes in…constantly. Every day, there are dozens of things to wash because sanitization is so critical in a CSA. Washing was all of this morning.

Harvesting Under Attack

In organic farming, there are a lot of insects. Because this farm prioritizes no spraying of insecticides, herbicides, or fungicides, there’s a lot of weeding and bugs…just like nature intended it. Some of them go after seedlings, some later grown plants. Other bugs such as spiders attack the nuisance insects and their eggs, which is nice. Today, small gnats and mosquitos we abundant while harvesting kale and swiss chard, especially because of the epic downpour from 8:30-9:30am. This is one of reasons why we wash our produce thoroughly.

Planting Peppers

We planted the two edge rows in the new greenhouse with carmine peppers. This involved weeding and sinking pepper plants into trenched areas that were already were amended with compost and ProGro 5-3-4. Just two of us planted about 4x 98-cell trays or roughly 350 plants in about 1.5 hours.

It was a short day, but we got a lot done.

AI Summary from Field Notes

My next curiosity will be to use my Mac M1 to be the preferred node (or service) which does this work, since its GPU is compatible with faster processing libraries in both of these components. I’m not sure yet if I want to use my M1 as a Kubernetes node that joins on-and-off as necessary or if I want to build a process that spins up a cloud GPU instance joined to the cluster just for the duration of the processing cycle.

This is all because I can’t afford a $2,000+ modern GPU as they are in high demand and this would be a major undertaking in it’s own right. For now, I stick with CPU-based inference and patience to get simple summaries like this:

Activities Performed:

1. Egg Washing: 48 dozen eggs were washed as the initial task of the day.
2. Lettuce and Bok Choy Cleaning: Produce from the fields was cleaned to prepare it for storage or further processing.
3. Mid-Morning Break: A break was taken, essential for maintaining energy levels during physical work.
4. Harvesting Kale and Swiss Chard: These greens were harvested post-lunch and washed again.
5. Barn Preparation: The barn was readied for the following day’s use, involving organizational steps.
6. Pepper Planting: Specifically planting carmine peppers in two of eight rows, utilizing interstitial planting methods. This involved weeding, composting with Pro Grow, and placing pepper plants into trenched areas.

New Terms Encountered:

• Interstitial Planting: A farming method where specific areas are left for spacing or other purposes.
• Carmine Peppers: A specific variety mentioned; further research could explore its characteristics and uses.

Questions and Future Research Areas:

1. Understanding interstitial planting practices and their benefits in agriculture.
2. Clarifying the process involved in closing out the barn for the next day’s use.
3. Investigating the impact of composting with Pro Grow on plant growth.
4. Exploring specifics about carmine peppers, including cultivation and usage.

Suggestions for Actions:

1. Optimize task order to enhance efficiency, possibly incorporating equipment or shade structures for effective produce washing.
2. Develop tools tailored for weeding light soil to streamline field tasks efficiently.

This summary encapsulates the key activities of the workday, introduces new farming terms, identifies areas needing further exploration, and offers actionable suggestions for improvement.

Part 1

Field Notes Analysis: June 6th Workday Summary

Activities Performed:

1. Egg Washing: 48 dozen eggs were washed as the first task of the day.
2. Lettuce and Bok Choy Washing: Produce from the fields was cleaned, likely to prepare it for storage or further processing.
3. Lunch Break: A mid-morning break was taken, which is essential for maintaining energy levels during physical work.
4. Harvesting Kale and Swiss Chard: These greens were harvested after lunch and then washed again.
5. Barn Closure: The barn was prepared for the next day’s use, possibly involving organization or preparation steps.
6. Pepper Planting: Specifically planting carmine peppers in two rows out of eight, following interstitial planting methods. This involved weeding, composting with Pro Grow, and dropping pepper plants into trenched areas.

New Terms Encountered:

• Interstitial Planting: A farming method where specific planting areas are left for spacing or other purposes.
• Carmine Peppers: A specific variety mentioned; further research could explore its characteristics and uses.

Questions and Future Research Areas:

1. Understanding the process of interstitial planting and its benefits in farming.
2. Clarifying the exact steps involved in closing out the barn.
3. Investigating the impact of composting with Pro Grow on plant growth.
4. Exploring the specifics of carmine peppers, including their uses and care requirements.

Suggestions for Actions:

1. Optimize task order to enhance efficiency, possibly using equipment or shade structures for washing produce more effectively.
2. Develop tools tailored for weeding light soil to streamline tasks.

This analysis provides a comprehensive overview of the workday activities, highlighting areas that could benefit from further exploration and optimization.

It was the perfect weather for just about everything on a farm, so we harvested, weeded, washed, and chored. The mosquitoes only came out in the afternoon and I had some help with the egg washing so that went faster than expected. A pig had to be returned due to a hernia.

Hoes Don’t Work Well in Wet Soil

The thing about the farmable soil on this property is that, while well taken care of for many decades, much of it is loosened clay. It’s not compacted, but regularly worked with a tractor, amended with OMRI-certified ProGro fertilizer and it’s weedy veg matter is tilled in to act as a little green compost.

Stirrup hoes work best in somewhat loose, somewhat dry soil. In extremely dry or caked over soil, effort required often leads to mistakes and chopping the plants you want to weed around. In wet or damp soil, it tends to cake on the hoe and not cut cleanly. Then all you’re doing is dragging a clump of weeds and dirt over other weeds which doesn’t work at all. There is a time and a place for every tool.

On a positive note, rain does tend to loosen up soil and after waiting a while, is perfect for planting and weeding by hand. So that’s what we did around the beets for about an hour. One 200-foot row with three people…and plenty of bug spray.

Meanwhile, a Pig Exchange

A month or so ago when we got a dozen pigs, Jamie noted that they didn’t seem to be in the best shape, but that they’d been through a lot and would recover quickly in their fresh new environment. As it turns out, what was identified then…an abnormality on one of them…developed into a visible hernia. So today, that one was returned for another one.

After a little research, hernias seem to be a relatively common occurrence in pigs from pig farms. The best article I could find during my time-boxed search suggested that there are two main reasons: umbilical/scrotal (mostly stocksmanship) and familial predisposition (genetics). Either way, it’s a business problem for the pig farm when there is an uptick in returns, so I hope it was just us (but likely not).

Scallion Harvesting

We started to harvest the scallions that we planted in the dry field all the way back in early April. They are almost indistinguishable from the weeds, but you get good at doing so quickly. Once this row is harvested, Jamie will go over it again with the tractor and till them in as green compost. Just in time too because much of the grass (rye, buckwheat) sprouts are beginning to form seed heads, and if they come to maturity and get added to the soil, well, there’s your next weed problem again right there.

Waiting for Chickens

Our existing 300 hens lay about 100 dozen a week. We are on week 2 of a 20 week CSA. Our target is around 160-175 dozen a week for the summer. We’re still waiting for 300 more chickens. This will become a problem in about another month if we can’t get them by then. Why is this even a problem? Well…

With the avian flu having wiped out many other big farm’s populations this past year, the supply of high-quality chicks is low. This has been creeping up on the American supply chain since 2022, but only since February of this year has the USDA put together a sizable strategy to combat the problem.

It’s only $1B dollars, and I personally can think of at least one or two tech companies we could sacrifice with a guillotine or at the stake to make that same amount up…but money alone wont solve this problem. It’s a matter of demand for chicken products and the monoculture macro-farms this pushes our industries and governing bodies to further invest in.

I can’t do much about that macro-problem, but I can put my efforts into being part of a local farm (i.e. food system resilience). As boring as egg washing is, I can learn how to produce farm-fresh eggs for a community even though it makes very little money, because it is safe and not simply because there’s demand for it. I can learn how much or little I can contribute to the local food system, and I can learn how to be a better steward of the land and the animals.

AI Summary from Field Notes

For some reason, today’s note summarization ‘think’ effort seems to be higher than usual. Maybe when I do a more diverse set of things, there’s more rabbit holes it potentially goes down. Bounding the think effort is something I want to look into, particularly with Deepseek over Ollama-based reasoning models such as Phi 4. Though I have the curiosity for it, I don’t have the time to take apart the reasoning process and do things via command chains, but I think I may have to play with that to make deeper progress on this front.

Summary

I organized the walk-in, washed produce, planted lettuce, weeded, prepared for the CSA pickup, and harvested scallions. With a big farm crew of six people, we got a lot done.

Defragmenting the Walk-in Fridge

Like everything on the farm, you don’t always have the luxury of putting things back in the most optimal way when you’re in a rush. The walk-in refrigerator is no exception. As produce is harvested, washed, and packed away, it’s easy to let things get a little disorganized. After a while, the refrigerator starts to look like a giant game of tetris, with produce piled on top of each other, and it becomes difficult to find what you’re looking for.

So today before a lot of washing and while others started in on onion bunching, I ask if I could organize the walk-in then I got to it. This involved lots of core strength, lifting totes here and there, getting things into stacks that made sense and were easily accessible. Because I had to be in there for a prolonged period, I used the light on my phone to illuminate the interior and keep the door shut. After a while, I got all the weeks of eggs we have stored up on one side in week-by-week stacks, and all the produce on the other side.

Then there were some CSA share bags that weren’t picked up the day before, and were put in the walk-in as well. After confirming with Jamie, I then disassembled the contents of the bags and refactored them back in to the walk-in veg totes. I think if people signed up for a specific day to pick up, and if they miss the day after as well, their share gets donated to a food bank…it might simplify the logistics on our end. But then paying customers wouldn’t get their veg and that’s the whole point of a CSA. I’m going to ask Jamie about how we handle the variability of pick-up without scheduled participation.

Scallions and Radishes Are a Messy Business

Scallions and radishes are two of the most common vegetables we grow at the farm, and they’re also two of the most messy to work with. They’re both leafy greens, but scallions have a long, thin stem, while radishes have a short, stubby stem. In both cases, when they’re harvested from the field they carry with them lots of dirt (as compared to cut things like head lettuce, spinach, field-bunched greens, and picked fruits).

The process is to de-clump them, bunch them up into appropriate amounts for distribution, and then wash them. Washing them usually involves a first rinse to get most of the dirt off in a ‘dirty’ bucket, then a thorough rinse in a ‘clean’ bucket…like other veg in the big basin. Often between crops like this and other produce, we will empty the basin of all the water (~30 gal) and re-fill it with fresh water. It all depends on how much we wash, what kind of dirt or pollen is on the produce, and how dirty the water gets. Most days, we can get away with one basin if we stage the washing in the right sequence (leafy greens, then heads, then rooty stuff).

Because the bunching work was done inside the barn instead of outside, for fear of rain, the floor and table needed serious washing. So after the rest of the crew went off to do something and I finished washing the scallions and radishes, I soap-water washed the floor and then completely sanitized the table outside before bringing it back in to use again. Yes, the floor will get muddy tomorrow, and yes I will take the 10 minutes to keep things sane in there again.

Planting and Hoeing in the (Light) Rain

The six of us then planted a 300-foot row of lettuce. Then four of us stayed to weed a sibling row next to that of beets. While the amount and maturity of the weeds weren’t bad at all, the going was slow because even in the light rain, this soil quickly gets sticky. It sticks to your boots, your pants, your hands, your tools…anything it comes in contact with, it will clump on and make less efficient.

Jamie asked me to try hoeing at the top of the field to see if slightly drier soil would make the hoeing easier. It kind of did, but it was still slow going. We got through most of the row before having to move on to CSA pack-up of the EV van up at the barn. Good thing the floor and table was clean.

In many of these posts, I point out flaws in my homegrown AI’s analysis. Many of them I ascribe to myself…information I don’t share or things it couldn’t possibly get because it’s just a really big statistical model. LLMs are a very amazing technology, but they are just that…a tool, not a magic box. Even the ‘reasoning’ models are just more process built on top of those statistics, so if the statistics are skewed, so will be the reasoning and outcomes.

But often I am left thinking that even if I did share all the information, it still wouldn’t be able to produce flawless summaries. There’s a lot of things that happen because of the inputs…the transcription, the prompt templates, the hardware on which it runs inference…and the inputs it doesn’t have like a curated list of things already encountered and additional articles I’ve already written on topics mentioned in the transcript.

Main Themes:

1. Produce Handling: The intern focused on washing, bunching, and cleaning various vegetables such as scallions, radishes, and lettuce.
2. Sanitation: Proper sanitation of equipment and surfaces was a critical component of the work.
3. Preparation for Distribution: Preparing CSA shares and organizing them for distribution was another significant task.

Activities:

• Washing and cleaning produce (scallions, radishes).
• Bunching vegetables efficiently.
• Planting new crops (bok choy).
• Sanitizing and preparing equipment for use.
• Organizing and distributing CSA shares.
• Weeding around planted rows.
• Loading produce for market distribution.

New Things Not Yet Encountered:

1. Processes: Efficient bunching techniques, extensive sanitization methods.
2. Vegetables Handled: Dealing with scallions, radishes, bok choy, and specific lettuce arrangements.
3. Tasks: Detailed cleaning of surfaces and equipment, CSA organization.

Questions and Future Research Areas:

1. Produce Management: How are quantities of vegetables tracked and managed?
2. Planting Processes: What is the specific method for planting bok choy and other crops?
3. Lettuce Arrangement: How are lettuces arranged in cell trays and maintained?
4. CSA Operations: What happens when CSA members don’t pick up their shares?
5. Scallion Farming: What is the farming system used for scallions?

Suggestions for Improvement:

1. Documentation: Develop detailed procedures for each task to aid future interns.
2. Tracking Systems: Implement better systems for monitoring and managing produce quantities.
3. Automation: Explore ways to automate repetitive tasks like sanitizing and planting.
4. Training: Create structured training programs for new staff, especially in specific tasks like CSA organization and sanitation.

This summary captures the key aspects of the farm intern’s work, identifies areas needing improvement, and provides actionable steps for enhancing efficiency and consistency in operations.

Part 1

Main Themes:

1. Produce Handling: The intern focused on washing, bunching, and cleaning various vegetables such as scallions, radishes, and lettuce.
2. Sanitation: Proper sanitation of equipment and surfaces was a key part of the day’s work.
3. Preparation for Distribution: Preparing CSA shares and organizing them for distribution was another significant task.

Activities:

• Washing and cleaning produce (scallions, radishes).
• Bunching vegetables.
• Planting new crops (bok choy).
• Sanitizing and preparing equipment for use.
• Organizing and distributing CSA shares.
• Weeding around planted rows.
• Loading for market.

New Things Not Yet Encountered:

• Processes: Efficient bunching, extensive sanitization methods.
• Vegetables Handled: Scallions, radishes, bok choy, specific lettuce arrangements.
• Tasks: Detailed cleaning of surfaces and equipment, CSA organization.

Questions and Future Research Areas:

1. Produce Management: How are quantities of vegetables tracked and managed?
2. Planting Processes: What is the specific method for planting bok choy and other crops?
3. Lettuce Arrangement: How are lettuces arranged in cell trays and maintained?
4. CSA Operations: What happens when members don’t pick up their shares?
5. Scallion Farming: What’s the farming system for scallions?

Suggestions for Improvement:

1. Documentation: Create detailed procedures for each task to aid future interns.
2. Tracking Systems: Implement better systems for monitoring and managing produce quantities.
3. Automation: Explore ways to automate repetitive tasks like sanitizing and planting.
4. Training: Develop structured training programs for new staff, especially in specific tasks like CSA organization and sanitation.

This analysis provides a comprehensive overview of the intern’s activities, highlighting key areas that could benefit from improvement through better documentation, tracking systems, automation, and training.

Part 2

Analysis of Farm Intern’s Field Notes

Date: June 10th (beginning of summer)

Main Activities:

1. Organizing the Walk-In Fridge: The intern started by organizing the refrigerator, referred to as “defrag,” which involves arranging produce in a way that allows older stock to be rotated out efficiently, especially for CSA distributions.
2. Processing Scallions: After cooling down, scallions were bunched and prepared for storage or further use.
3. Washing Lettuce: Harvested lettuce was washed, likely using containers called “totes,” which are crucial for organization and sanitation.

New Terms and Concepts:

• Defrag: A new term used to describe the fragmented organization of a walk-in fridge, emphasizing efficient rotation of produce.
• CSA (Community Supported Agriculture): Highlighting the importance of timely produce distribution.
• Totes: Plastic containers used for storage, essential for managing farm operations.

Questions and Future Research Areas:

1. Defrag Method: Understanding how this method works and its effectiveness in farm settings.
2. Organizational Strategies: Exploring methods to keep the fridge organized for multiple users.
3. Scallion Storage: Investigating optimal storage conditions post-harvest.
4. Lettuce Washing Procedures: Looking into effective methods to prevent waste and contamination.

Suggestions for Improvement:

1. Labeling System: Implementing clear labeling to enhance organization and reduce confusion among workers.
2. Training Programs: Providing training on fridge organization and produce handling for all farm hands.
3. Container Utilization: Exploring specific bins or labels to streamline storage practices.

This analysis provides insight into the interns’ activities, highlights new concepts, and identifies areas for further exploration in farm management efficiency.

We weeded, planted, weeded some more, then prepared for deliveries and a CSA pickups. I am also getting better at anticipating what the farmer’s priorities will be per day and how the team will be distributed to accomplish what needs to get done.

A Usual Day at the Farm

Today it was 80 Fahrenheit and sunny, and there were 6 of us, so we were able to get a lot done. The mosquitoes and gnats were out and I got a little burnt on my face from the reflective UV index. We harvested, washed, weeded, and packed the vehicles for deliveries and market.

And there was lots of weeding…of the holes in the landscaping fabric, of the bare rows in the main field, and of the FilmOrganic rows in the dry fields.

I Am Getting Better at Anticipating Priorities

After a only few months, I’ve gotten good at anticipating activities and in what order they need to occur on days where I also know what the key outcomes need to be. But there are multiple layers of priorities (i.e. micro/daily, mezzo/weekly, and macro/monthly/seasonally).

In short, the order goes:

• URGENT: Must do right now, otherwise you’ll lose money or incur cost/time
• IMPORTANT: Must do later/soon; either blocked by URGENT or a blocker to IMPENDING/DEFERRABLE activities; if not done, will make problems worse quickly because time marches on
• IMPENDING: Activities that are able to be done at the optimal time; usually daily chores fit here
• DEFERRABLE: Things that are in the plan (i.e. legitimate needs), but the effort to do them isn’t always known or it’s not the right time of season to do them (cover crop termination, haying, etc.)

In later posts, I’ll adjust this matrix as needed and expound on the multiple layers (like a pyramid diagram) but for now, this is a good start to sharing the thought process.

The Usual Activities during CSA Season

At this point in the season, we’re representing somewhere just about every afternoon between Tuesday and Saturday. This means:

1. there’s usually something to harvest, wash, and store
2. anywhere from 10-25 share bags have to be packed
3. everything brought in the EV van has to be checked and clean

Additionally, during the day there’s always:

• greenhouse watering
• morning ‘chores’ (i.e. check chickens and pick up eggs, check/feed/water pigs, etc.)
• evening checklist (close main gate, close up barn and greenhouses, check gas in gator)

Variably every few days, there’s also:

• turn on the water to the new greenhouse with tomatoes and peppers for about 3 hours every 3 days
• process (wash and pack) eggs
• check and fill gas in gator and truck

My Personal Pipeline Priorities

If I’m lucky, I’ll get to peel off a few weekend hours and implement a better approach to the blog skeleton drafting process, which for the most part can be automated, except for the human writing part:

Still for today, this is what my AI thought I did:

Summary

Aside from the usual harvesting, weeding, and chores, I also got to talk business and needs with the farmer. Very little time to take photos, but I did get a few of the piggys.

While weeding the onions in the dry field, I got some one-on-one time to talk with Jamie about business needs and priorities, particularly options for value-added products that the farm could offer based on their current operations and produce.

What Are Value-added Products from a Farm?

The term ‘value-added products’ refers to transforming raw agricultural products into something new through processing, packaging, or other methods. These products are categorized and regulated differently than ‘farm fresh’ veg produce, eggs, and other outputs of operation like honey from bees.

Value-added products save consumers time, meet their specific culinary desires, and expand the farm’s profile. They also nicely fill out a farm stand or local general store with goodies that spark peoples’ curiosity and imagination for what their farming neighbors can offer.

What Value-added Products Could We Offer?

While we sweated under the hot sun, I thought about which value-added products we could offer:

• salsas and sauces
• jams, jellies, and preserves (peppers, tomatoes, herbs)
• pickled vegetables such as bok choy, cabbage, cucumbers, carrots, radishes
• dried/dehydrated vegetables such as kale chips, tomatoes, and carrots
• dried herbs and flowers
• honey candy

All of the above could be sold at farmers markets, CSAs, and other retail outlets without changing current outputs. It is also using surplus produce rather than hoping to sell that to local restaurants and distributors last minute.

Additionally, if the above works well, we might consider expanding productions to partner with other local agricultural businesses with a similar need for partnership, such as:

• veg-infused oils and cheeses
• home-preservation classes
• ready-to-use veg such as lettuce mixes, ‘zoodles’ (zucchini noodles), and medley of gourd spears

Side-note on Farm Product Revenue

It’s not simply the sticker price of products a farm has to think about when dividing their labor and time. Some things like eggs are not a big revenue generator, but they are an essential characteristic that CSA members value and the chicken coops are a natural source of free high-nitrogen fertilizer when rotated properly across fields. Producing hay, while not a revenue generator, is worth it for farms that already have the equipment to do so (provided this isn’t core to their produce model).

The diversity of operations on a farm also contributes overall resilience of it’s business model when one or a few areas are affected by adverse weather and other factors. Building in appropriate buffers, fallbacks, and flexibilities is the responsible thing to do if at all possible.

I Should Feed My LLM Better Produce

This is the most honest work I’ve ever done…between the real field work, the writing, the technical process development, and the daily commitment to turn around a blog post within 24 hours. I can’t wait for the summaries to get better that this:

AI-generated Summary of Field Notes

I had an appointment in the afternoon, so it was only a half day for me. I organized the walk-in, bunched radishes, washed produce, and harvested fennel.

A Time for Radishes and Fennel

The first thing to do on a new week before harvesting anything is to organize the walk-in. This makes sure that if there is any produce ready to sell, it is at the front of the walk-in. Then also there is space for new produce in the back, where what we harvest during the week will go. As we use what’s in the front, we pull forward from the back. Along with daily harvesting, this ensures customers are getting the freshest produce possible.

Today we started harvesting the fennel. Unlike much of what’s growing in the field now, I wasn’t involved in planting it (I was out that day). The smell is fantastic…the crispy crunch is phenomenal.

Three of us also harvested radishes, about 200 bunches, for about 2 hours. I think we could have done it faster, but we were also being careful to pick only those that were marketable, leaving the smaller ones to continue to grow.

There Are Many Forms of ‘AI’

Just think about that the next time a developer tries to convince you that they’re using AI to build something. If they’re not careful, what they just ‘vibe coded’ could be the end of your company quickly.

Consolidated Summary: Farm Intern’s Field Notes (June 16th)

Main Themes

The intern focused on three primary areas during their work:

1. Farm Operations: Reorganizing the walk-in, harvesting, and preparing produce for storage.
2. Sustainability: Emphasizing the importance of using fresh produce immediately to minimize waste.
3. Teamwork: Collaborative efforts in tasks like harvesting radishes and tatsuyue.

Activities

1. Reorganizing the Walk-In: Consolidating older stock forward to reduce waste.
2. Harvesting Radishes: Removing about 200 bunches, requiring several hours of work.
3. Preparing Tatsuyue: Cutting and washing produce while others harvested radishes.
4. Harvesting Fennel: Collecting the first batch after lunch for storage.

New Things Not Yet Encountered

• The term “tatsuyue” was mentioned but unclear if it’s a specific type of produce or a typo.
• No previously unknown farming practices were encountered.

Questions and Future Research Areas

1. Efficiency: How to optimize the harvesting process without compromising quality.
2. Harvest Timing: Ideal time for harvesting radishes and fennel.
3. Handling Small Radishes: Methods for managing smaller, less marketable radishes.
4. Water Usage: Amount added when storing produce like fennel.

Suggested Actions

1. Rotation Schedule: Implement a structured walk-in rotation system.
2. Grading Systems: Develop methods for assessing and grading produce.
3. Efficiency Improvements: Explore tools or techniques to speed up harvesting.
4. Communication: Ensure clear team communication about harvest priorities.

This summary captures the key aspects of the intern’s work, highlighting operations, sustainability efforts, teamwork, and areas for future improvement.

Part 1

We spent much of the day weeding and harvesting, so there was time to talk and joke and think..

From last week’s discussion with Jamie the farmer, I took on a short-term spike project to check out options for partnering with a commercial community kitchen or cannery to build out our options for value-added products (discsused in last Wednesday’s post). This is what I’ll focus on writing about today.

Terminology

• Value-added product - a product that has been processed or transformed from its raw state, adding value through cooking, packaging, or other means
• Co-packer / contract packager - a company that manufactures and packages products for other businesses, often on a contract basis (more info here)
• Commissary / community / shared-use kitchen - a shared kitchen space where food businesses can rent time and equipment to prepare their products list of Mass. options here
• Unit fee - a fee charged per unit of product produced
• Repack - the process of repackaging a product into a different container or labeling it with a different label, such as “Medely of Grilling Veg” or “Zoodles / Zucchini Noodles”

Selection Factors

These products need to be small batch (for now) so that we can run experimental recipes to figure out what sells and what doesn’t. So we want someone else who is licensed and certified to make and package these food products to do the cooking and preparation, execute on an existing recipe, and provide recommendations for improvement. Presumably since they do this a lot for others as well, they’ve got more experience than we do about what works and doesn’t when taking raw agricultural product and turning it into a finished product.

The space that is used also should be relatively local (reducing transportation costs) and licensed for producing these prepared foods. Optimally, both the contract packer and space used should generally align with the mission and economic goals of the farm.

Early Search Results

* Distance is calculated from the farm’s address in Ipswich, MA.

Other potential options:

• Boxford Bakehouse
• IQF Custom Packing
• PGI Foods

This is just a start and as I continue doing the research, I’ll likely create a page specifically for this information. Many of the sites I visited had old or outdated information, some had simply gone out of business, so I’ll also add a ‘linting’ step to check specific links for validity and alert me of changes in my site build process.

30 Minutes Later, Criteria and Options Refined

The above only took me about 30 minutes, but I’ve already learned a lot about the options and criteria. I did try to use ChatGPT with a prompt of:

As a eastern Massachusetts farmer, what are my local options for community kitchens or canneries to create value added products from my produce?

…but it largely focused on shared‑use & commissary kitchens. It did at the end however, offer a list of ‘need / goal’ to ‘best option’, which while not accurate, did give me a few ideas to discuss with Jamie tomorrow. I’ve always thought of AI as more of a fake-friend to help me brainstorm at the whiteboard than a replacement for human interaction, chatbots be damned. People use Google search for that all the time and don’t think twice about it, then complain that their ads are eerily accurate.

What AI Doesn’t Know Hurts It

AI-generated description: Activities and topics today were planning activities related to radishes, field, things.

Yesterday’s Hallucinations Are Today’s Faulty Inputs

The problem with yesterday’s summary, the complete hallucination that was Part 2, was easy to diagnose in the <think> block describing the reasoning.

“From the example provided, the origin link seems to be a hash, which might point to a specific document or data source. However, without accessing that link, I can’t directly read the content of the notes.”

It accurately guesses as to who might be asking the questions. Each summary I send also includes a simple ‘origin: [guid]’ front matter tracking code for internal purposes. So because it got side-tracked about thinking that the origin link is a hash and thinks it should link to something it doesn’t have, it guesses about generic activities:

“In the example given, the intern mentions tasks like moving calves, feeding livestock, cleaning stalls, planting trees, and applying fertilizers.”

Today it didn’t get side-tracked, and it never did before yesterday’s debacle, so I’m not sure what changed but I do know I’ll have to tweak the prompt to make it more clear what it should be doing.

---
origin: 1IassjQn1b6NBsiO_T5UKKi_U5ul5ql3v
---

Summary

I harvested and weeded a lot today. I also discussed value-added product strategy with a farmer. This was where I hoped I would be last year, but it took some hard work to get here. And it will take more to see any of it through.

Scoped Experimentation

No experiment can be learned from without at least some initial boundary conditions, constraints, or expectations. In one of my prior corporate roles, I took on the challenge/experiment of trying to create a legitimate space for experimentation. I was lucky to have a sharp, pragmatic and caring boss and direct reports. Through the years of the pandemic, we did the best we could to provide air cover and opportunity to those who were creative and bold enough to try something without fear of failure or judgement.

Scoping is vital to experimentation in two ways:

1. Proper scope ensures that there is sufficient, presumably appropriate constraints to drive outcomes
2. Aligned scope helps communication about the value of experiments (and experimentation) to decision makers

Not appropriately defining boundaries around experiments can lead to a lack of focus and resources, potentially resulting in wasted effort and missed opportunities. Positive outcomes are not the only value of experiments since you can often learn just as much from a failed experiment than a successful one. But the only things that look like work worth funding (from corporate leadership, investors, etc.) are positive business outcomes, not negative ones. (At a later time, I’ll research what this bias in formally trained management is called.)

My career experiment failed and succeeded in various ways. Many ideas were turned into features or new products (‘new product initiatives’, a.k.a. NPIs). The people involved (except for me and my boss) are still employed there in various roles. The processes hardened in this crucible are still used to this day. Many others who experienced the impact have since moved on, hopefully and meaningfully influenced by the work we did.

Tacit Approval

There are times for formal approval. Many of my conversations with very well tenured academic and industrial leaders help me understand this within the context of military, high-risk, mission critical contexts. “ATO”…Authority to Operate…is a term used in many DoD uses of specific tech stacks, on submarine installations where ‘production environment’ means life or death.

There are so many more…civilian…times where approval takes on other forms. In the corporate experimentation scenarios, most of my experience relied on implicit and short-term risk factors. Some were self-imposed rigor and others took the form of one-on-ones and midnight prep sessions for the next day of C-level big room meetings.

Today my ‘tacit approval’ came in the form of very polite and short conversation with the farmer about next steps to confirm research direction while weeding tomatoes. In short, rather than waiting a season/year to plan out ‘big’ co-packing relationships, I suggested that we experiment with a few small batch value-added products to validate the demand by local consumers. This supports the initiative for a general store in Essex beyond the growing season, not to mention the potential for increased revenue streams.

What AI Captures that I Keep to My ‘Human’ Writing

Automated outputs don’t change unless either their inputs or process meaningfully change. During the work week I have very little time to change either of these two things because I’m busy with the actual work of doing the work. But this gives me the time and gives rise to creative thinking about what changes need to take place between the work days and the (potentially) improvement days.

The immediate improvements I know I need to make this weekend are:

Consolidated Summary:

It was a very hot day after 11am field work. Fortunately, this crew and it’s farmer know how to balance what needs to get done with what can get done within reason.

Nerd Note

Balancing Worker and Weather Conditions

On a hot day, it’s nice to work with a farmer that knows the value of factoring in what’s going on in the hourly (not daily) weather report with what the given crew on hand can handle.

Weather reports and conditions, notwithstanding the imposed level of inaccuracy being incurred by the current politically-fueled mass layoffs at NOAA, are almost always NOT what you experience in a particular location. This is because weather stations, aggregated across a wide area, are not able to capture the hour-to-hour microclimate fluctuations in temperature, UV index, and humidity that can occur in a specific location.

So today, between watering the greenhouse twice and early morning harvesting, Jamie (the farmer) kept our necessary work pretty well-balanced. Hot weather encourages many early-season crops to bolt (go to seed) and this is exactly what happens with lettuces and broccoli rabe in these fields when spikes in temperature occur. Just like too much rain all at once rotting rooty veg like radishes and salad turnips, extremes in anything are generally a source of stress on the crops and a need to ‘hurry up’ about specific activities on the farm.

Sequencing and Variance

Today I spent most of the morning washing and packing crops coming in from harvesting in field. Because we’re always running tight on tote bins during the weekly harvest-to-sell cycle, every day there’s plenty of things to wash (rinse, wash, rinse, sanitize spray). Totes, harvest bins, surfaces, knives, buckets, floors, basins…basically anything that comes in contact with produce at any stage is as clean as it can be made.

In every system, there is always a sequence of events and degrees of variance, as Dr. Eliyahu Goldratt would say. In this case, the sequence of events is clearly based on constrains while the system variance is driven in large part by the weather. Balancing these fluctuations is the job of a ‘manager’ and how well they do this significantly impacts the success of the system.

Then in the afternoon I spent most of my time bunching scallions, radishes, and salad turnips. The barn has a mobile industrial fan which is *so necessary* and even on a hot day, opening both garage door sides gives a cross breeze that makes everything bearable.

But the biggest factor on a tough day is what and how the farmer calls the shots. What to rush on, what can be deferred to tomorrow/later, who to group with whom, who is wilting like the veg and needs a ‘cool down’ activity when. Who can go the extra mile on something, who does or doesn’t have the training for a particular critical set of activities.

Manual vs. Automated Processes

When you automate something, you trade off the speed and repeatability of a process with the opportunity to perform quality control and the flexibility to adapt to changing conditions. In the case of farming activities, especially the mundane and manual ones are always an opportunity to check for quality and make adjustments.

In my rushed prototyping work, I made the simple engineering mistake of changing multiple dynamics at once without verifying them separately: adjusting the process as well as foundational tech stack components…not unlike so many corporate management actors I’ve encountered over the years.

The short version is:

Approaching the Limits of a Lack of Context for LLM Work

Every AI Model Has Defective Biases

Case in point, “bolting prevention strategies” is not a thing in organic farming…sort of. I mean, you can pluck seeding heads but that’s not really addressing the fundamental issue in the organism. The time trade-offs of micromanaging natural tendencies in plant species when coupled with organic practices just isn’t worth the over-inflated sense of economic virtue that high tech (agro or digital) is optimized to value.

The genomic makeup of the plants is what determines whether they bolt or not given specific conditions. Despite lofty modern AgTech claims over what it can control about crop productivity, the accompanying chemical-heavy WW2 litany of *-icides and productized synthetic fertilizers expose the faulty scientific longitudinal effectiveness that ‘big ag(riculture)’ says it can deliver.

Dogs are a recent phenomenon. Plants are not. Did your bizarre conjunction of a Labradoodle exist 65 million years ago or was that over the past 100 years? Unfortunately, even modern organic crops service a similar human goal: convenience to humans and climate.

Today’s LLM (Deepseek-R1:8b) Generated Summary

Today was the first work day of a record-breaking heat wave in New England. We harvested in short spikes with breaks in between then set up some irrigation lines.

Farming in Extreme Heat

Even when there’s a heat advisory, there’s still work to be done. Priorities shift to ‘minimum viable’ outside work, such as harvesting in the early morning and especially things that may bolt (like lettuces). Soil that has dried out needs to be watered, so irrigation is a priority. Watering the greenhouse multiple times a day is a must.

At this farm, even though we still work, we are provided ample opportunities to hydrate, take breaks, and cool off when the heat is at its peak like this. We also rotate duties between field and barn work (such as washing, bunching, and packing) to avoid overheating.

Before and After Lunch Harvesting

In the morning, we harvested chard, fennel, and lettuce. After lunch, we harvested garlic scapes (the flowering curly shoots of the garlic stalk) and scallions. Jamie checked the potatoes by pulling some up, which ended up in my take-home produce later. Despite some potato beetle damage, he thinks most of the crop will pull through.

Then I helped Jamie continue setting up irrigation lines to the main field. This involved super-manning a huge wooden spool with some 2-inch tubing onto the back of the truck.

After we were done with that, we went back up to the barn where I stayed to bunch and wash scallions with another person. Meanwhile Jamie and the other person went up to start extending the pig pen with fencing into the forest. Though shady up there, it was 2pm and 90 degrees fahrenheit and humid, so I can only imagine it was a haul to continue the day on that task. The pigs are outgrowing the wooden shade structures at this point, so it was necessary.

Not all AI Crop is ‘AI Slop’

The populist outsider’s view and understanding of things termed ‘AI’ is often a caricature of the reality. ‘AI slop’ a the most recent example of this…content ‘creators’ who are not experts but rather rely on AI tools to generate content that is not only low quality but also often offensive or harmful.

Unless You Do Quality Control, It’s Likely Slop

Quality control (QC), quality assurance, quality management, testing, verification, and validation were all sections I had a heavy hand in pushing to completion in my prior work with IEEE 2675-2021, despite last-minute adjustments out of my hands. This took years of writing, reviewing with the team, revising, re-reviewing and balloting, then finalizing.

Pop-up Blockers to Progress

I need a more proactive approach to these ‘pop-up’ progress blockers. I don’t have time for running disaster recovery every single time I change some small part of my infrastructure. I also don’t think any of that would have helped in this case, since the problem wasn’t with anything I changed but something in the intra-dependency after a restart that was changed outside my control (the model from online download itself).

Anyway, while I was troubleshooting the ollama deployment, I realized that I had switched autoscribe (the component that produces the following notes summaries for this blog) to use Deepseek-R1:1bn (only one billion parameters) instead of the default :8B. When I get home later today, I’ll re-run it with 8B and compare the results.

I noticed that it gets side-tracked because of the tracking code at the top of the notes again, and that all of Part 1 is a hallucination. Particularly egregious is the line about Roundup PowerMax, which is not at all what we would ever use on this farm.

This is why unchecked and un-validated AI output is dangerous, not only informational but operationally. Imagine if I had used this output to make decisions about what chemicals to use on the farm!

Here’s what the AI generated for me with the 1B (small) model:

Summary

Despite the heat wave, there’s always work to do on the farm. In small batches and with plenty of water, seven of us got done what needed to get done.

Hottest Day on Record

From 85 degrees at 8am to 105 degrees by 1pm, today was one of the hottest days on record for coastal (and all of Massachusetts). Still there is work to do and a way to do it: small spikes and batches, cool-down breaks and plenty of water. Prioritization over what MUST be done vs. all the things that should be done on a day like today is so necessary.

For instance, watering the plants which don’t have deeply established root systems like the kale and chard already do, such as lettuces, radishes and baby carrots is really important. If their roots only go down into the soil two or three inches, and the top soil is dry to that depth, the plants will simply not survive for long in the extreme heat. Whole crops will be lost simply due to heat stress and lack of water.

All that time preparing beds, seeding or transplanting, weeding, waiting for them to grow…only to be done in by two or three hot days and no irrigation…would be irresponsible.

Watering the Greenhouse and Moving Irrigation

Today we watered the greenhouse for an hour three times because it was so hot that each time we checked the trays, they were dried out. The last one I did was at 4:30pm so I’m interested tomorrow morning to see how dry the overnight heat got them again.

We also moved the tripod sprinkler from the tomatoes and carrots mid-field to the high field where the radishes, lettuces, onions, and cabbages are. Mostly this took about 30mins to disconnect and drain the tubing, then moving it and reconnecting it in the new location.

Why move things at all? Because that amount of durable 2-inch tubing not to mention the tripods and sprinklers would cost a lot of money and they’re only needed in certain places at certain times (like this heat wave) per season. Yes, moving it around takes time which is opportunity cost over another activity, and muscle effort. But at this point, it’s the responsible balance between money and time spend here.

While We Worked Hard, They Worked Harder

The farm team of seven today worked very hard and sweat a lot. Meanwhile because of the ideal conditions for haying, the groundskeepers crew of four made dozens of bails of hay from the grasses terminated over the past few days and left to dry until now. They should win awards for this, like every year seemingly on the hottest day of the season, since you don’t want to bail wet or still-green grass as it will be prone to rot if you do.

As mentioned in previous posts, my other side-work around multi-participant transcription and diarization taught me that balancing process with the right tech for the use case is really important with these LLM/AI infused workflows. So this weekend, I plan to try a few incremental (and hopefully positive) changes:

1. Try the existing langchain process with a different ‘reasoning’ model
2. Adjust the process to be more like my new transcription processing work
3. Try different parts of the new process with different models (reasoning vs. non-reasoning)

UPDATE: For the first time in 3 months, I had to hold off publishing within 24hrs because I switched my attention to diagnosing and fixing what was going wrong with the newest Deepseek update. At a later time, I’ll publish the artifacts, bizarre behaviors, and contributing factors analysis, but for now I moved to a different reasoning model and many of the LLM hallucinations went away.

Final Consolidated Summary

Main Themes

1. Extreme Heat Impact: Prolonged high temperatures (e.g., 100°F) strain crops, workers, and irrigation systems, necessitating adaptive farming practices.
2. Organic Farming: Reliance on natural methods like mulching, manual weeding, and manual irrigation, avoiding synthetic chemicals.
3. Crop Management: Harvesting, weeding, and irrigation for stress-prone crops (e.g., chard, radishes, greenhouse plants).
4. Hay Production: Critical for soil moisture retention, managed under dry conditions by the groundskeeping crew.
5. Watering Challenges: Frequent irrigation required for greenhouse and stressed crops, with a focus on preventing plant stress.

Activities Performed by the Intern

• Harvested chard, radishes, and salad turnips.
• Weeded leeks and reorganized irrigation lines.
• Moved refrigerators to optimize barn cooling.
• Watered greenhouse three times daily due to extreme dryness.
• Washed eggs and observed hay-making processes.
• Assisted in pop-up market prep in Essex.

New Things Not Yet Encountered

• Greenhouse Watering Frequency: Watering the greenhouse three times daily, a new experience.
• Hay-Making Process: Observing the groundskeeping crew’s efforts under heat.
• Irrigation Line Reorganization: Adjusting systems to meet crop needs.
• Heat-Induced Crop Stress: Noticing plant stress from prolonged dryness and high temperatures.

Questions and Future Research

1. How to mitigate heat stress on crops (e.g., shade structures, watering timing).
2. Does mulching with salt marsh hay improve soil moisture retention?
3. Will upcoming rain affect the current hay crop, and how to manage it?
4. What is the optimal greenhouse watering frequency in extreme conditions?
5. How to ensure worker safety and productivity during heatwaves.

Suggested Actions

1. Adjust irrigation schedules based on heat and plant stress.
2. Monitor greenhouse watering to prevent yield loss.
3. Plan for rain impact on hay and crops (e.g., harvesting timing).
4. Install shade structures to protect crops during heatwaves.
5. Implement hydration breaks and cooling measures for workers.
6. Track mulching benefits (soil moisture, crop health) to assess long-term impacts.

Part 1

Main Themes:

1. Extreme Weather Conditions: The record-breaking heat on June 24th on the North Shore of Massachusetts.
2. Environmental Impact: The potential effects of extreme weather on the farm’s operations, crops, and livestock.

Activities Performed by the Intern:

• Observation of Weather: Noted the record-breaking heat as the first entry in the field notes.
• Documentation: Recorded the date, weather condition, and its significance (hottest on record).

New Things Not Yet Encountered:

• Record Heat: The intern may have encountered this extreme weather for the first time, which is unusual for the region.
• Impact on Farm Operations: The intern may not have previously experienced such a severe heat event, prompting a need to assess its effects.

Questions and Future Research Areas:

1. How does prolonged extreme heat affect crop yields and growth patterns?
2. What measures can be taken to mitigate heat stress in livestock?
3. How will the farm’s irrigation and watering schedules need to be adjusted for extreme heat?
4. What long-term implications does this heat event have for the region’s climate and agricultural practices?

Suggested Actions:

1. Monitor Crops: Check for signs of heat stress (e.g., wilting, reduced growth) in crops.
2. Adjust Irrigation: Implement more frequent or higher-volume watering to counteract evaporation.
3. Livestock Care: Ensure shaded areas, cooling systems, and increased feed availability for animals.
4. Record Weather Trends: Track future weather patterns to prepare for potential heat events.
5. Farm Planning: Develop contingency plans for extreme weather, such as emergency cooling systems or crop rotation adjustments.

Part 2

Main Themes

1. Extreme Heat Impact: The farm faces intense heat (100°F) affecting both crop health and worker efficiency.
2. Organic Farming Practices: No synthetic pesticides/herbicides; reliance on natural methods like mulching and irrigation.
3. Crop Management: Harvesting, weeding, irrigation, and greenhouse care for stress-prone plants.
4. Hay Production: Groundskeeping crew manages hay-making despite heat, emphasizing dry conditions.
5. Watering Challenges: Frequent irrigation needed for greenhouse and stressed crops, with a focus on preventing plant stress.

Activities Performed by the Intern

1. Harvesting: Collected chard, radishes, and salad turnips.
2. Weeding: Tended to leeks in the garden.
3. Refrigerator Management: Moved refrigerators to optimize barn cooling.
4. Irrigation Adjustments: Reorganized irrigation lines and adjusted watering schedules.
5. Greenhouse Care: Watered the greenhouse three times daily due to extreme dryness.
6. Egg Processing: Washed eggs, a routine task regardless of season.
7. Hay Monitoring: Observed the hay-making process, noting its importance in dry conditions.
8. Market Prep: Assisted in preparing the pop-up market in Essex.

New Things Not Yet Encountered

1. Greenhouse Watering Frequency: Watering the greenhouse three times in a day, a new experience.
2. Hay-Making Process: Observing the groundskeeping crew’s hay-making efforts under heat.
3. Irrigation Line Reorganization: Adjusting irrigation systems to meet crop needs.
4. Heat-Induced Crop Stress: Noticing plant stress from prolonged dryness and high temperatures.

Questions and Future Research

1. Heat Resilience: How to mitigate heat stress on crops (e.g., shade structures, timing of watering).
2. Mulching Effectiveness: Can mulching with salt marsh hay improve soil moisture retention?
3. Rain Impact on Hay: Will the upcoming rain affect the current hay crop, and how to manage it?
4. Greenhouse Watering Strategy: Optimal frequency and timing for greenhouse watering in extreme conditions.
5. Worker Hydration Protocols: How to ensure worker safety and productivity during heatwaves.

Suggested Actions

1. Adjust Irrigation Schedules: Increase frequency or adjust timing based on heat and plant stress.
2. Monitor Greenhouse Watering: Ensure plants receive adequate water to prevent yield loss.
3. Plan for Rain: Develop strategies to manage potential rain impact on hay and crops.
4. Improve Shade Structures: Consider adding shade to protect crops during heatwaves.
5. Hydration Management: Implement hydration breaks and cooling measures for workers.
6. Document Mulching Benefits: Track soil moisture and crop health post-mulching to assess long-term benefits.

It was 90 degrees as we weeded and hoed, but this was much more bearable than yesterday. Napa cabbage harvested, basil pinched, and leeks completely weeded.

Weeding and Hoeing

This just has to be done when you can get to it, so we did. Most of the topsoil had already been baked by the sun over the past few days so the light weeds in the fennel and cilantro beds were easy to address with stirrup hoes. Not so much with the leeks, which even though we mulched with salt marsh hay a few months ago, still had plenty of weeds to deal with by hand.

This kind of work seems to go so much faster when you have 6 people working together instead of just two or three. Yes, it gets done faster, but the mental burden of how much more is left seems lighter as well.

Excuse Me, You Did What to Basil?

We also pinched off basil. Basil the plant, not some guy named Basil. As basil is a tender annual, it’s best to pinch it off at the seeding tip to encourage bushier growth and extend the harvest season. Doing this in my own garden this morning took about 5 minutes because I only grow about 2 dozen Thai and other varieties of basil plants per season. In a 200 foot row with 6-inch apart and three plants per spacing, this is 1200 plants. It took four of us about 30mins to do all of them.

Trading Pallets for Tote Bins

As the harvesting season comes in to full swing with nearly every week day having something to harvest, wash, and pack into tote bins for the walk-in refrigerator, every bin becomes sacred real estate for produce. There are very few not in use at any given time now.

In general, anything that touches the floor must be re-washed and sanitized. In the early season, we were using the smaller orange bins upside down in the fridge to stack green bins on because no produce (or bin carrying such) can touch the floor. They’re easy to move around, compact, and you can always rearrange if you need one or two to come out.

While this works in the early season when there’s less produce being harvested from the fields, there’s a point where it makes more sense to use the pallets instead. So we washed and sanitized the ground-level black pallets that the totes used to sit on outside, and moved them into the fridge instead. We reclaimed about 8 orange totes this way which is more than enough to help with the tight window of variance between harvest timing and tote availability.

Watering and Consolidating the Greenhouse

It was another day for multiple waterings throughout the day, but after mentioning on Monday that I was interested in starting to consolidate various trays and seedlings in the greenhouse, I got that job today. So I spent about 20 extra (hot, sweaty) minutes moving still-viable seeding trays to the front of the greenhouse and tossing otherwise fried trays into a pile for later deconstruction/composting.

There are still large sections that can be further consolidated, but it would take two people to do it efficiently since many of these sections fit nicely together on wooden ‘ladders’. Moving these by yourself would take at least twice as long as just getting another person to help, so that’s what I’ll suggest as a 30min task tomorrow so it’s all done for the season…at least until we start seeding some fall crops.

Sometimes, You Just Need to Simplify…Even with AI

Three weeks ago, I switched the summarization part of my daily notes from using Ollama3.2 to using Deepseek-R1. I had my reasons at the time, but only time could show that this was not worth it. However, after the last few days of mitigating the Deepseek-incurred hallucinations and over-thinking tirades in the summaries, I decided to look for a model which still had reasoning built in but did a much better job of the type of prompts and context I was providing.

I also had my hesitations about how and on what the Deepseek models were trained, not to mention all the political and ethical implications of using it. After seeing it devolve into different languages, both human and programming ones, to do something as simple as extract topics and takeaways from clearly articulated and thought-through field notes…I now understand how important it is to right-fit specific AI technology to specific tasks…and to develop experience and expertise in the field of operationalizing AI.

Let’s just say that some models, no matter how thoroughly you prompt them to use English, will always go back to their roots in the language they were trained on. And I don’t speak or read Mandarin, so diagnosing the origins of the hallucinations was more difficult than it had to be.

After running some trials against the existing langchain process with a different ‘reasoning’ model, I found that the qwen3:4b model was a better fit. We will keep a careful eye on its performance and make adjustments as needed. So for today, let me introduce you to my revised AI summarization process:

Summary

Today was much cooler than the past two scorchers, so we planted and weeded as much as possible until market load-up.

A Short but Productive Day

Today I started at 10am because I had a personal event until 9:30am. In that two hours before lunch, we planted four rows of melon, two of each two varieties:

• ‘Juice Box’ (muskmellon)
• ‘Sugar Baby’ (small watermellon)

Here’s the link to the learning session I had with ChatGPT while planting in the field. A new topic I hadn’t learned about before, Brix testing using a refractometer, helps to know when the optimal harvesting time for produce like melons.

I also was thinking about how to proactively know about punctures or leaks in field drip irrigation lines, which can happen for a variety of reasons. Before assuming that there isn’t a set of good options for addressing this already, I also asked ChatGPT about it.

At Lunch, Discussion about Pandan Leaf

Someone at lunch mentioned their use of Pandan leaf (in Vietnamese cuisine) which I’d never heard about, and someone else asked about where to get seeds to try to grow it. Again, a quick ask to ChatGPT quickly lent that seeds are really not a thing, but rather it is predominantly propagated via cuttings propagation. Also as it turns out, our neighbor state of Connecticut has a few places that will ship propagated cuttings which is cool.

Fiddly Weeding

The rest of the afternoon was spent weeding a bed of dill?…carrots?…the jury was out but I think carrots because it didn’t smell at all like dill.

There was also a lot of purslane which, to most local farmers, is a weed because it A) spreads quickly and B) in places where you’re trying to grow some actual crop. In my own (small rooftop) garden, I don’t have ‘weeds’ but rather ‘overbearing plants’. Unless it’s crowding something out that I also want to make the space for, it too deserves to be somewhere in my little ecosystem.

On a farm, it’s different. Weeds and weeding sucks. They cost time, especially, which is money and opportunity cost. If you don’t deal with them, they’ll take over, no question. Nature abhors a vacuum and the ideal conditions we make for our crops are also the conditions that make weeds thrive. Some farms use herbicides, but not this one. We use our hands to make sure that the produce is as safe and healthy for people as possible.

A Well-fit LLM is Worth Its Weight in Switching Cost

I’ve said enough in the past two weeks about how much trouble Deepseek has caused my workflow to produce useful context on this blog. All I’ll say today after switching to Qwen3 and after listening to it’s outputs as a verification pass is that the finding the right tech fit takes time and is worth the effort.

In all my work over the past 10 years to help companies select and implement specific vendor solutions, I never understood how important it was that the decision was seen through to success like I do now. I truly don’t have the time to mess around with things that go bump in the night because someone out there is trying to make a buck off me or mess people up.

For today, my homegrown and hand-crafted summarization process using AI produced this:

AI-Generated Summary of Today’s Field Notes

Today was a pretty typical day on the farm during CSA season with limited crew. Chores, harvesting, bunching, washing and packing of lettuces, radishes, salad turnips, fennel, and scallions. Watered vegetables and cows.

If You Think the Heat is Unbearable, Try Being a Cow

After first watering the greenhouse, Jamie picked me up in the gator and we stopped to give water to the cows who are down by grassy space near the pond. 12 cows drink about 30-40 gallons of water each and every day, if not more when its hot.

Though I want to talk to them like their pets, the fact is that I had steak last night. People who eat meat but anthropomorphize farm animals need to work with them closely for a while and pick one or the other. Pigs are in many ways smarter than dogs, but I love a good BLT. There are times where on our way back to the barn from harvesting vegetables in the field, I think about how the cows and pigs and chickens are functionally high-inputs food compared to plants.

I know that sharks and wolves and our pet cats eat other animals. They have instincts evolved over millions of years and relatively tiny brains compared to mine. I can think thoughts and write blogs and make complex decisions about who I want to be and do what I don’t feel like doing when it comes down to it. Maybe the best thing I can do for the planet is to eat mostly veg, nuts, and bugs instead of other animals…not simply work in agriculture, but work in nature.

The thing about vegetables is that they aren’t sentient (unless you put googley eyes on them, lolz). Likewise, the thing about animals is that they aren’t vegetables (unless you raise them to eat them). You can get your veg and protein from plants. Is it as tasty as a well salted, peppered, and grilled steak? Does it take work to get veg to be the most of your diet? Can everyone afford to be vegetarian? What are other countries and cultures doing about food?

Part of nature is carnivores. Part of America is capitalizm. If people want to pay $20 for a burger or $35 for a steak, then there will always be someone willing to raise animals for food. An only-veg CSA serves mostly only-veg consumers. If I were to run my own farm, I might just stick to egg-laying hens, but I have more to learn and (no pun intended) digest about this than just 3 months has brought me so far.

My AI is in Recovery Mode

For now, below is the summary from my M1 running Qwen3:4b on today’s field notes:

Summary

We moved the chicken coops and harvested chard in the morning. Then in the afternoon, we harvested basil. It was a very hot and humid day.

Moving the Chicken Coops

This is the first time I helped to move the chicken coops. Jamie the farmer is a very patient person and this helps when dealing with someone who is not familiar with the process.

On this farm, the chickens are ‘free range’ meaning that they are not confined to a coop at all times. The coops are on dairy trailers that can be moved around the farm, and when in a location are also extended outside with a fenced and netted yard doubling the ground space the coops take up.

The process goes as follows:

• get as many of the chickens up into the coops as possible
• undo the outer layer of fencing
• undo the overhead layer of netting
• undo the inner layer(s) of fencing
• chase down and catch remaining chickens
• move the coops to the new location by towing them with a tractor
• as part of the move, align them close together (about 2 feet apart)
• re-assemble the inner fencing
• re-assemble the overhang yard netting
• re-assemble the outer fencing
• use ground stakes to secure fencing in ground gap areas
• triple check that there are no gaps in the fencing
• set up water and ladders again

Today this took two people about 2 and a half hours and is not something one person should do by themselves.

Basil Trimming

For the past 15 years in my garden, I always have a few staple herbs. Basil is one of them because it grows prodigiously and is a culinary delight when fresh.

Though I forgot to take a picture of the basil before and after trimming, the general idea is that once the plant is established (i.e. after the first two side growths produce their real leaves), the middle stem can be harvested. Basil loves a good trim as it responds quickly in season within about a week by putting the additional energy into new growth.

So four of us did this and got about 60 bunches in an hour. We also did light weeding in between the rows. At ~$4 per bunch, that’s $240 worth of basil in an hour!

Weekend AI Recovery Mode, Part Deux

Improving the LLM Reasoning Process

I’ve been using the same prompt for autoscribe for a while now, but I’ve noticed that it gets hung up at times where the langchain document splitter splits the document into parts that are too short. When this happens, there’s a lot of places in the THINK blocks where it says ‘Wait, the user is asking…’ and ‘Wait, maybe the user wants…’. In excess this is a sign that something is causing the reasoning to loop over itself.

Until I do that, we shall continue to see good-but-not-great results from the AI like this:

Summary

Lots of weeding today because hashtag organic farm. Yellow summer squash are actually rolling around the fields. Basil, scallions, chard…all bountiful. Also, my personal AI now talks to me over Slack and lives in a quantum state between not-working and working.

Schrödinger’s cat is a famous thought experiment in quantum mechanics. The title of this post is a double entendre about organic farming because it also involves many unknowns until observation and action make something a known.

Experimentation in Nature

As I was sitting on the back of the big truck today riding back from the field to the barn, I had a moment of reflection about how insignificant yet at the same time wonderfully bizarre life on this planet is. I thought of the billions of other iterations of life across the universe, each grains of sand, timelapses of decomposing dead birds, and (not despite but because of all this) how important it is for us to treat each other well while we can.

Over billions of years, nature has made inconceivably numbers of experiments. Most don’t work out and very few do…but then again what does it mean to work out? If stars spiral into each other and become black holes, does that mean the experiment failed? If creatures evolve that are silicon-based, do we consider that failed because it isn’t us? If one species of basil attracts fewer pests or resists diseases than other more wild ones, does that mean it’s better?

Organic farming is a complex experiment just like the more modern synthetic and chemical approaches to agriculture. To me the key difference is that in ‘natural farming’ the same systems that made us also make organic farming possible whereas ‘invented farming’ often fails to factor multiple biological dynamics. Invented farming practices focus on narrowly-defined objectives and often ignore their broader impact on the ecosystem.

Change as a Constant

Like in our natural world, change is also a constant in the human experience. Nature tends to change on much slower timescales than our day-to-day experience and perception so its easy to take for granted long running or slowly changing eco-systemic dynamics. Even the most generous and well-intentioned of us still evolved from deeply engrained survival instincts in a world that is significantly different almost every decade. The bias towards short-term thinking is hard-wired into our brains and is often at odds with the long-term goals of sustainability.

Most modern farming practices are based on a narrow set of assumptions that were true in the 1950s but are no longer valid in the 2020s. If there was ever a time to question the status quo, it’s now. There are many situations where the phrase “what works, works” (a truism) applies in day-to-day agriculture but much of what has worked better in the recent past came from myopic, unsustainable practices.

Personal Change Management Systems

I can’t change nature, but maybe the world…? I once thought I could, even if only a little bit. Now that I’m older, I realize that the only thing I can change is myself…if that. So at least for now, changes I can make are limited to my own life and the people around me. My decisions impact myself, my family, and the community around me far more than anything I do or say in the broader world.

Change usually takes effort which takes sustained bursts of extra energy and picking the right small changes helps to keep progress sustainable. Deadlines and accountability might be helpful for some folks, but for me it usually takes a personal commitment to do something (or stop doing it) a certain new way because I have experienced the downstream impact of my own actions otherwise. Moment by moment decisions are what aggregate up to day-by-day improvement.

What about the week-to-week, the year-to-year, and beyond that? Days turn into weeks, then months, then years. If you feel that thinking beyond a few years is impractical, ask yourself what your farm succession plan is if you own your own land? If you don’t own your own land, what are your plans for the future?

The fact is that corporations have been buying up family farms and consolidating to employer-employee relationships for decades and the trend is accelerating. Fewer people who actually do the hard agricultural work have any decision-making power that is informed by connection to the land. Having any say in how the land is used, especially if you don’t own it, is increasingly difficult for the average citizen.

Co-ops and Coops to the Rescue?

Cooperatives help businesses and communities thrive by pooling resources, sharing risks, and promoting democratic decision-making. Farming cooperatives can provide a way for farmers to maintain some control over their operations and the land they work on, even if they don’t own it.

Likewise, the average citizen (not in farming) can also participate in democratic and municipal ownership by tending their own chicken coop, gardens, and participating in events and commerce associated with local agricultural cooperatives.

Especially now, a time of peril for our national agricultural support systems and federal assistance, farmers should band together in local cooperatives and experiment with new ways of working together to support each other and the land. Change is a constant, and we must adapt to survive.

My AI Is in a Superposition of Space and Time

Last weekend I tried to short-cut a fix to a storage problem, but after 4 hours of trying to get microk8s to come back from a fraught data migration, I realized I had to start over from scratch. Space (for data storage) as it turns out, is a limited resource just as land is in farming. And just as Einstein’s general relativity showed us that energy and matter can be interconverted, the relationship between space and time can sometimes be traded on.

Case in point: dealing with a storage space constraint cost me time…low storage equals high time cost. Lots of storage equals less time dealing with space management issues. I know, on one way this is kind of a hyperbole, but in a very real way shows that two things otherwise considered unrelated might have effects on each other that can be quantified.

Molecules want to move, especially at the quantum level. Movement is a fundamental property of matter and energy, requiring us to invent the concept of time to explain how states change. But as Douglas Adams said, “Time is an illusion. Lunchtime doubly so.” Tomorrow I will be home and not on the farm. I will be working on the process, not the content, of this blog. There is no such thing as lunchtime for me, except for when I’m on the farm.

I had to laugh when I scanned through them and saw:

In some ways, working with my personal AI about farming is very much like teaching it (and myself) about the natural world. I hope some day robots that take my job find this blog useful, if nothing other than as a bizarre and curious artifact of the human experience.

AI Generated Summary of Daily Notes

We harvested a lot of beets today, a lot…5-6 crates worth. And some early celery came out too. The greenhead flies and mosquitoes were relentless under the 90 degree sun. It was a long Monday after a long weekend and I had time and impetus to reflect on my tech career to-date.

Hot Day, Lots of Bugs

Though we spend most of the morning harvesting beets and early celery, we also had to battle bugs while repairing and baiting the deer fence. Unsteady footing, 90 degrees in the sun, hands covered in peanut butter so I couldn’t easily wipe the sweat off my face, greenhead flies and mosquitos biting all day…let’s just say I was happy to get home and take my shower.

The upside is that we are getting different produce in our CSA shares this and next week. And some early tomatoes are being collected too, but we have to dial in the watering schedule to keep them from nutrient deficiency and root rot (cracking in the fruit and yellowing leaves). The ones in the new greenhouse are very different in their needs apparently than the typical field tomatoes.

Dancing to the Beet

Beets are an under-rated vegetable. While most people go directly for the bulbous root, I prefer the greens which are so packed with iron and other minerals that they taste meaty to me. They’re also crispy and gorgeous when cut into a salad of lettuce and kale.

Sodium, Boron, Potassium, Magnesium, Manganese, and Potassium are all very concentrated in both root and greens. Because the chicken coops are sat over these fields the prior season, we don’t have to fertilize much with the usual ProGro 5-3-4. Chicken manure is very high nitrogen and calcium so these beets grow quite well with minimal interference.

Yah, We Got the Beet

Harvesting these beets has been a multi-week process. At first when the weeds weren’t bad, we were able to go through and pick the golf ball to baseball sized ones which makes room for the smaller ones to grow. This process disturbs the root system a little but not enough to stunt the remainder’s growth.

Because we use Johnny’s Select Seeds almost exclusively, these beets are the classic deep red/fuchsia color. I will check whether they are Boro, Merlin, or Red Ace tomorrow since Jamie direct seeded them and thus I did not encounter them in the greenhouse of course. If I had to guess now, it would be the Merlin variety.

Beeting the Career Gong

There are two consistent things that occur to me as I’m doing monotonous and/or sweat-inducing tasks on the farm about my prior tech career:

1. Though the farm work may sometimes feel like economic exile in the wilderness from a quarter-million dollar tech job, it’s really not. It’s a paradise compared to the Machiavellian behaviors I had to navigate almost every day since the summer of 2023.
2. Potential employers that can’t figure out any way to hire me really have their heads up their asses. I am very skilled and hard working, care about others, and improve continuously. If that’s too much for someone to handle, they’re not worth working for.

I declined one job offer last year since really seriously putting the petal to the metal on my job search in November, otherwise it’s been a few really close misses due mostly to similar candidates already further down their funnel. Almost three hundred job applications since last May and I am convinced that LinkedIn and other job sites are too broken to be useful.

I was talking with a friend the other day about what they used to do when they lived in China. They said they were a general manager of their uncle’s store and that the practice of hiring family members was common. Though much of the tech hiring pipeline I’ve experienced would otherwise consider using an ‘in’ like this to get a job unethical, I’m convinced that it is in fact a major factor in getting hired these days. Not just a referral, the key is someone who has authority and the confidence to highly recommend you, if not outright hire you into a role.

I’ve been a software engineer for 25 years and have never had to have a referral. Professionally, I’ve been a sales engineer working closely with customers for a decade. I ran an incubation engineering team for 3 years. I’ve been involved in submitting patents, due diligence before acquisitions, strategic long-range planning, overnight large-scale hand-holdings with customers, trainings, and ran a mentorship group for years. Don’t tell me I’m ‘overqualified’ to make $150k, please.

And yet, I’m still pulling beets and washing celery. Harvesting knives are my ‘cutting edge’ since March.

Beeting the AI Gong

With this work, I’m not using AI because it is a buzzword…I’m using it to show my technical skill as well as where AI can go wrong. Populists have no idea what they’re talking about and developers just want to keep their jobs. I am encumbered by neither ignorance or greed, just cucumbers (‘encumbered by cucumbers’). If there’s a gong to beat, I’ll do it with a freshly harvested beet.

AI-generated Summary of Field Notes

We harvested swiss chard, basil, cilantro, and lettuce. We weeded, hoed, fixed irrigation lines, and seeded trays. It was another 95 degree, mostly sunny day with less wind than yesterday. The bugs were worse. The conversation was real.

Cilantro Like This…

“Farm fresh” is a phrase thrown around supermarkets to get people to buy not fresh things. Working on a farm, harvesting every day, working with different soil types, next to animals that help fertilize fields, paying close attention to sanitization processes…transforming seeds to truly fresh produce that I bring home to my own family table daily…is hard work.

I don’t trust things that are too easy. Hard things at least give rise to thoughts about how to improve, but easy things rarely do. The only thing I trust from ‘easy’ stuff is that someone is making a profit off the people not thinking or working through the inputs and processes that lead to the total cost at the end of the trip.

I generally don’t have cilantro in the house or my own garden. I don’t not like it, but to date the supermarket version leaves me with too much for any recipe and usually tastes more like soap than vegetable. We home cook most of our food, even after a long day. Granted, many of our favorite recipes don’t require cilantro, but many could benefit from it. I know we can dehydrate or freeze the excess cilantro from the sprawled hand’s worth we get at the store, but none of us like the flavor of that incarnation and we don’t have an overabundance of freezer space anyway.

‘Cilantro like this’, really farm fresh, comes in half the grocery size bunches, smells like all kinds of amazing citrus and grassy things, and is reasonably portioned to be a wonderful accent to one or two meals. Since we were harvesting just before lunch, I stuck an extra bunch in one of my front overall’s pockets to bring to the table. One of the other farm hands said that it was “fire” for them, and on the rest of our sandwiches…was just that.

Real Conversation

Sometimes when you work closely with people, even if it’s the kind of work you can leave at work, the conversation turns to deeper notes. Sometimes you’re ready for this and sometimes you’re not. It’s always a bit of a litmus test between peoples to see how they handle or offer that kind of talk, but when done in the right context and appropriate manner can lead to better understanding of the people around you.

People on farms come from all sorts of backgrounds. Just because the work is (close to or) minimum wage doesn’t say anything about where your farm co-worker came from, what they experienced, what they’re going through, where they’re going, or what value they put into their work both on and off the farm. Broad spectrum assumptions when present are toxic (or at least detrimental) to a deeper understanding of the people around you.

If you don’t work on a farm, my suggestion is to get to know your local CSA by being a customer and maybe even volunteer. Your food will taste better for it, I promise.

AI is Trained on People…Currently

Most of this content is made by and for humans, so naturally it is biased in some way shape or form. In my own notes, this is of course true as well. I think organic agricultural practices result in highly nutritious, safer (than unnatural farming) food. For future record, an example of my extreme bias for organic vegetable farming and related practices is as follows:

Agro-chemicals like chlorothalonil (by ground or air) is a violation of nature and betrayal of conscience for those who use it, even if that’s what his daddy and his daddy did for generations. It’s recently been linked to the decline in reproductive capability of bees, which to a farmer who knows the price of importing pollinators and produces crops who need them, should be a no-brainer ‘no thank you’. Unfortunately, many who apply broad-spectrum pesticides, fungicides, and herbicides little conscience or sense of reciprocity other than economic efficacy, and would likely draw no distinction between produce treated with this kind of chemical in what they sell to the market and what they would put on their own table.

Note: I am a ‘junior’ when it comes to farming and still idealistic. Time will likely help my transition to more nuanced and realistic positions.

The Overly Broad AI You Use is What Your Deserve

Unfortunately, the more you use current AI platforms, the more they capture and retain about you for their own proprietary use. You don’t get an option to transfer the context from SaaS ‘solutions’ like Zoom, ChatGPT or SourceGraph because in a world where your personal context is the ‘special sauce’, exporting that would be like giving away the secret sauce recipe.

Like broad-spectrum fungicides which erode the ability of the entire farming system, so does the proprietary nature of training data in AI systems. People happily working and profiting off these corporate models are then not unlike the farmers who wouldn’t serve intentionally poisoned food to their own table, but are happy to poison others’ wells and fields and livelihoods.

For today, like every working day for the past four months, I’m using a homegrown process which is augmented by an LLM to take in end-of-day voice notes in order to siphon off a simplified summary. While not my own direct writing, it often captures in detail the events of the day more than I have time for, even if it comes with the occasional quirk.

Consolidated Summary

Main Themes:

1. Weather Impact: Extreme heat (over 90°F) and intermittent rain affecting pest behavior, soil conditions, and work schedules.
2. Farm Tasks: Daily chores (weeding, post-harvest work, greenhouse seedling management) and adaptation to weather changes.
3. Greenhouse Operations: Seeding, transplanting, and seedling care in a controlled environment.
4. Adaptation Strategies: Adjusting work plans based on weather forecasts (e.g., prioritizing cooler days).

Activities Performed:

• Weeding: Removed weeds from lettuces and cilantro rows.
• Post-Harvest Work: Washed and packed crops.
• Greenhouse Tasks: Seeded lettuce, peppers, and bok choy; pricked out seedlings and transplanted them.
• Weather Monitoring: Noted rain’s effect on pests and soil, and tracked forecasted cooler weather.

New Things Encountered:

1. Pricking Out Seedlings: Transplanting seedlings into individual pots.
2. Greenhouse Seedling Care: Managing seedling growth in controlled environments.
3. Pest Behavior: Observing how weather influences pest activity.

Questions Raised:

1. How does pest behavior change with weather conditions?
2. Will seedlings thrive under current greenhouse conditions?
3. How can work schedules be optimized during unpredictable weather?
4. What soil management practices are effective in varying weather?

Suggested Actions:

1. Monitor Bugs: Track pest activity to adjust farming practices.
2. Prepare for Cool Days: Adjust work plans to accommodate weather changes.
3. Optimize Greenhouse Conditions: Ensure seedlings receive proper care.
4. Document Weather-Task Correlations: Analyze how weather impacts farm operations.

This summary integrates key insights from both documents, highlighting the interplay between weather, farming tasks, and greenhouse management while emphasizing adaptability and observation.

Part 1

Main Themes:

1. Weather Conditions: Extreme heat (over 90°F) with intermittent rain, impacting farming activities and pest behavior.
2. Farm Tasks: Daily chores such as weeding, planting, seedling management, and post-harvest work (washing, packing).
3. Adaptation to Weather: Adjusting work schedules and methods based on forecasted weather (e.g., moving to cooler days for labor).
4. Greenhouse Management: Seeding and transplanting crops in a controlled environment, including seed selection and care.

Activities Performed by the Intern:

1. Hoeing and Weeding:
   • Removed weeds from lettuces in the upper plot.
   • Weeded cilantro rows after a brief rain shower.
2. Post-Harvest Work:
   • Washed and packed crops.
3. Greenhouse Work:
   • Seeded lettuce, peppers, and bok choy.
   • Plucked seedlings with excessive seeds from trays (pricking out).
   • Transplanted seedlings into new trays (not potting up).
4. Weather Monitoring:
   • Noted the impact of rain on pests and soil conditions.
   • Forecasted future weather (cloudy, cooler temperatures).

New Things Not Yet Encountered:

1. Pricking Out/Transplanting: A technique to divide seedlings into new trays for growth.
2. Greenhouse Seedling Care: Managing seedlings in a controlled environment, including seed selection and tray management.
3. Pest Behavior: How rain affects bug populations and the need for future monitoring.

Questions and Future Research Areas:

1. Impact of Rain on Pests: How does the rain influence bug activity, and will this affect crop health?
2. Seedling Success: Will the new seeding method (pricking out) lead to healthier plants?
3. Weather-Driven Work Scheduling: How can farm tasks be optimized for cooler days (e.g., Wednesday-Thursday)?
4. Soil Management: Can the dry soil be improved with irrigation despite the heat?

Suggested Actions:

1. Monitor Bug Populations: Track pest activity after the rain to assess long-term impacts.
2. Prepare for Cooler Days: Plan tasks for Wednesday-Thursday (e.g., weeding, transplanting).
3. Optimize Greenhouse Conditions: Ensure proper ventilation and moisture levels for seedlings.
4. Document Weather-Task Correlation: Record how weather affects work efficiency and pest behavior for future planning.

Part 2

Main Themes:

1. Agricultural Practices: Harvesting, pruning, and preparing crops.
2. Timing and Weather Management: Impact of heat on plant growth, need for timely harvesting to prevent bolting.
3. Irrigation and Maintenance: Fixing irrigation systems, managing hose durability, and ensuring water access.
4. Plant Care: Managing plant stress (e.g., lettuce bitterness, basil bolting) and optimizing growth through pruning.
5. Seasonal Planning: Preparing for future planting (winter squash, lettuce) and adapting to seasonal challenges.

Activities Performed by the Intern:

1. Harvested chard (approx. 120 bunches).
2. Bunched and harvested basil (approx. 120–150 bunches), focusing on early growth to avoid bolting.
3. Collected younger lettuce to avoid bitterness.
4. Harvested cilantro (approx. 60–70 bunches).
5. Fixed irrigation lines by repairing holes and cracks in hoses.
6. Prepared for planting winter squash and lettuce (reverse hardening off).

New Things Not Yet Encountered:

1. Basil Bolting Management: Pricking flowering heads to prevent bolting.
2. Lettuce Bitterness: Recognizing that mature lettuce is bitter, prompting focus on younger plants.
3. Irrigation System Challenges: Dealing with hose durability and maintaining water access across fields.
4. Seasonal Planting Strategy: Preparing for winter squash and lettuce, which requires adjusting planting timelines.

Questions and Future Research Areas:

1. Optimal Harvest Timing: How to balance harvesting frequency to prevent plant stress and bolting.
2. Irrigation System Efficiency: What materials or methods can reduce hose damage and improve water distribution.
3. Plant Stress Mitigation: How to minimize heat stress on crops (e.g., basil, lettuce) without relying on refrigeration.
4. Lettuce Bitterness Management: Are there specific growing conditions or pruning techniques to reduce bitterness?
5. Seasonal Crop Adaptation: How to best prepare for winter crops while maintaining current harvest schedules.

Suggested Actions:

1. Monitor Plant Health: Regularly check for signs of bolting in basil and lettuce to adjust harvesting schedules.
2. Upgrade Irrigation Systems: Invest in durable hoses or alternative irrigation methods (e.g., drip systems) to reduce maintenance.
3. Document Harvest Timing: Track when crops reach peak quality to avoid overharvesting or underharvesting.
4. Experiment with Lettuce Varieties: Test different lettuce types or growing conditions to reduce bitterness.
5. Plan for Winter Crops: Develop a planting schedule for winter squash and lettuce, ensuring adequate water and nutrients.

We planted winter squash, harvested basil and cilantro, and prepared for market. And I got to drive the John Deere 1140 tractor for the first time.

Just Like That…Weather, Basil, and Tractor Operation

The past two days, and much of last week and the week before, have been hot and bug-filled. On Monday while baiting the deer fence with peanut butter in the lower field, I got swarmed by mosquitoes and greenhead flies. I was wearing long sleeves and pants, but the bugs were still a bit overwhelming. But today while planting in the same field, the weather being much cooler and breezy, in a freshly tilled and slightly damp soil bed, there were no bugs at all. Just like that, conditions for certain activities on a farm change.

The basil has already been seeding up for two weeks, but we’ve been pinching it off before and during harvest. Now it’s full on harvest or loose it time. We’ll probably get another week in the CSA out of it, even if the top leaves are slightly bitter, the rest of the leaves aren’t for some reason and you can make the best kind of pesto with it. So just like that, the basil is pretty much done because it doesn’t keep for long after picking.

I got to till a 300-foot field row for the first time with the big tractor. Jamie had me stand / hold on to the side of the tractor while he drove it and showed me how to operate the various controls. I am from the ’80s and manual transmission was always my first automotive love, as with the throttle on the sit-down lawn mower I grew up on. But the 3-point hitch attachment…the power drive and up/down controls…along with keeping straight to the row was a good amount of things to learn how to balance together at the same time. Just like that, I did what would take dozens of hoes hours to do by hand in minutes.

Things that don’t go ‘Just like that…’

There are many monotonous, tedious, and slow-going activities on a farm that don’t lend themselves to the ‘just like that’ mindset. It’s okay, you get used to them, but the whiplash between hurry up and slow down due to the weather, the seasons, and the unpredictability of nature is definitely part of the learning curve.

Case in point, pinching basil or cleaning up celery plants is a slow process. It took three of us maybe 10 minutes to harvest about 50 plants. It took me about 30 minutes to clean them. You have to…cleaned ones are so much more ready to sell than quickly-rinsed ones, but it’s tedious work. Just like washing eggs, just like planting leggy squash, just like weeding…at least you know there’s something tasty to show for your efforts.

Just Like That, I’m a Farmer

After writing my mind in yesterday’s post about use of broad-spectrum herbicides in organic farming, I’ve been thinking about how hard a line I drew. Yes, I’m only a year into farming, but I’ve been very pro-natural and anti-chemical for at least 15 years of being a gardener and parent.

But I had to ask Jamie about it, why he draws the lines he does on this farm, how he would engage other farmers (or not) about their use of chemicals, and as expected the conversation was more nuanced than my simplistic and junior view. In short, there’s no good discussion you can have with someone about their practices if you come at it abrasively or in an accusatory manner (this, I knew). The best way you can share your views is to listen to theirs, and then show them what you’ve been doing differently and how it might be feasible for them to adopt some of those practices. Organic farming business models are still evolving, and there’s no one-size-fits-all approach.

In this process, I mentioned that I still don’t feel like I can speak as a ‘farmer’ because I’ve only been doing this for a year or so, and I’m still learning the ins and outs of the business. Jamie was quick to point out that I’m already a farmer in my own right, and that my perspective and experiences are valuable contributions to the farm. That was nice to hear.

Just like that, My AI Assistant Loses Its Bullet Points

Maybe a more nuanced approach to diagnosing the source of the problem is warranted. If I had the time, this would include:

1. Reviewing the prompts and input used (maybe the document splitting didn’t work the same as it has done on other days)
2. Re-running the summarization process with the same input to see if the output is consistent
3. Adding observable moments (logs, metrics about artifacts, etc.) to the process to see if there are any patterns or trends
4. Adding a check for the output format to ensure it matches the desired format

AI-generated Summary from Field Notes

We moved chicken coops, harvested the first early onions and garlic, and prepared for all the cabbage that will be coming in from the fields for the next few weeks.

Back to Reality

The weather was really nice…70s, sunny, and breezy. Perfect for just about anything you have to do on the farm. After a long, traveling weekend for a memorial service, it was nice to be back home and back to work.

We started by moving the chicken coops, mostly because the new hens had arrived late last week and it was time to get them outside for the first time in their little lives. As part of the new setup, we used one of the fences between the old coops and the new coop so that the birds could get used to each other without any sort of skirmishes developing between them.

Then we moved over to harvesting the first early onions, so juicy and sweet right now that you get sticky bunching them. By then it was lunch time.

A Garlic Harvest, Part One

After lunch, we harvested the first of the garlic. Jamie said to ‘pull with your knees and legs’…I quickly found out why. A few hundred later, if you used your back to do all the bending and pulling, you’d be in trouble. Also, positioning almost directly over the garlic shoot and being patient prevents from snapping the shoots off without properly getting the bulb. We drove the truck full of garlic to the barn, where we bunched it then drove it over to the utility shed where on the other side we hang the garlic to dry out a bit before putting it in CSA member shares. Tomorrow, we’ll pull the rest and do the same.

Then we harvested some celery and fennel. By this time in the season, the fennel is starting to flower, so we only pulled a little before we decided to call it. The rest of the day was spent washing and storing the produce.

AI Summary from Field Notes

You can check out the results in the gallery. Though not all tags are correct, this handles many of the photos well and then I can more quickly go back and update the tags as needed.

And as always, my personal AI-based summarization process produced the following summary:

Summary

We bunched chard, onions, basil, and garlic. Photos today had a technical problem, but the old farm dog Jenkins did not. He was as attentive and ready to deliver as ever.

No Photos Today

My primary phone has suffered it’s fair share of damage over the past two years and this month has not been any kinder than the last. My prior phone’s cracked inside screen still allowed it to be usable as a ‘farm phone’ for photos and syncing up to my self-hosted Immich instance. Today I decided to use the latter of the two and none of the photos (I thought) I took showed up. A fresh new unlocked phone is on the way, but for today this means that there are no new pictures to share, despite all the recent work on the gallery to make it easily searchable by tags.

Fortunately, most of the things I’d otherwise take snaps of today were the same things we’d already been doing yesterday and last week…harvesting garlic and early onions, washing swiss chard, and bunching basil.

Jenkins is a (Very) Good Dog

Jenkins is the older of the two farm dogs. While the younger dog is about 4 years old, Jenkins is 7+ and a very good boy. He’s the one who will actively respond first to oddball sounds coming from across the field where the hens live, and though both are great little farm companions, there’s something about Jenkins that I connect with more than the younger one.

When I was a boy, we adopted an Alaskan husky pup. This was way too much dog for 7 year old me to walk and we would often find her stuck somewhere after breaking loose of whatever she was connected to when we were away. My wife grew up with border collies, the very opposite of my early dog experiences but despite that, we both share a love of animals, pets or other humankind.

Today when we were headed to lunch, I took a chance and whistled like Jamie their owner does when he wants both dogs to come up into the truck. They came immediately and we headed down, the whole (human and dog) farm crew together. I don’t like to head-fake or pretend anything and I don’t have little treats in my pocket to hand out to dogs or colleagues alike. So when I finally catch on to the behavior of creatures around me, dogs or people, to get them to understand it’s time to move on to the next thing, I feel like I’ve finally earned the right to call on them when they’re needed.

If I was a farm dog, I’d like to think I’d be like Jenkins. He is attentive, anticipates human movements, checks for human confirmation, takes initiative when appropriate, and is occasionally ornery like the best of us are. I would say this is due to the kind and patient nature of his owner alone, but I know that all dogs are different just like the farms some of them are raised on, and much of it is due to the individual dog’s personality more in the long run than their early and often short-term training.

Where Was the Training in DevOps Pipelines?

In my prior career, I spent a lot of time writing automation and helping others do so. While ‘Infrastructure as Code’ (IaC) was an important step in the right direction for its hayday, in retrospect it was never enough. The learning, the training, was only potentially in two places: 1) the human brains of the engineers who wrote the code, and 2) the proprietary platforms that had full access to anything done on them (no matter what the contracts might have said).

Most, if not all, software delivery pipeline platforms and technologies missed the mark when it came to what mattered most…even the early adopters of ‘AI’ in 2021. The pipelines we wrote were often brittle, hard to maintain, and required a lot of hand-holding to keep them running. This doesn’t work in today’s AI-infused world because much of the tuning now has to do with the prompts, the context, and the (necessarily) flexible processing patterns required to make them work reliably. The leading platforms still miss the mark by providing AI features designed to wow executive audiences, not by providing the practical features and as importantly the training needed to make them work in practice

Unlike Jenkins the farm dog, IaC-forward technologies and platforms like Jenkins, Harness, Pulumi, and certainly tech like Terraform, Ansible, and CloudFormation, are not at all self-training. They require constant human attention and effort to keep them working as expected when even the slightest things change in enterprise IT environments. One missing character from a tag deep down in a Helm chart can not only cause software delivery pipelines to grind to a halt, but possibly the production environment as well.

All I know is that if there was a CI/CD version of the farm dog, it would be attentive, anticipatory, factors in human behavior, knows when to need confirmation vs. take initiative, and only when really appropriate be ornery to us.

What My AI Doesn’t Learn Won’t Hurt It

AI-Generated Summary from Transcribed Field Notes

We harvested Alisa Craig onions, carrots, and garlic this morning then we did odd jobs like irrigation repair and bunching before packing the vehicles for markets.

Only a Tenth of the Way in to All the Onions

We planted a lot of onions this year…like, a lot. The number was 13,500 when I asked back in mid-May. After two weeks of harvesting the earliest variety from the top of the main field, in my estimation we’re only about a tenth of the way through the harvest because those rows are about one-fifth. The rest are in the dry fields and of a different variety that will be used for storage.

It felt good to finally be done planting the onions. It will likely feel good to finally be done harvesting them too.

All Done Harvesting the Garlic

It’s nice to see the entire lifecycle of something that over-winters from year to year. Garlic is one of those things, having planted and mulched it in the fall last year, seeing it grow up and race the weeds, harvesting and hanging it to dry…I’ll be very happy for its tastiness in the winter.

First Eggs from the New Hens, Already!

Surprisingly, we got our first eggs from the new hens today. This is less than one week since they arrived and Jamie says that usually it takes two weeks or so. This is a good sign because our egg production will ramp up quickly now that we have a total flock of over 500 hens. And how the new 250 are rapidly growing…their appetite for anything edible is easily seen when you drop overgrown trays of seedlings off for them to eat in the morning and by mid-day its all gone.

Of all the farm animals I’ve experienced, I think chickens are the most productive. They fertilize the soil and provide eggs. They aren’t as dirty or smelly as pigs, and compared to cows they don’t require as much space or food inputs. Though they aren’t a source of high profit (despite the price of eggs), I think they’re worth it for every vegetable farm that can support moving mobile coops around every few weeks.

Everything Is in Production

The greenhouse is a ghost town, not a plant in there now that everything is out in the fields and temperatures are high. What’s left of the seedlings is small half-dozen successions of lettuces and other greens that will likely be planted in a few weeks to head in to early fall.

On a farm, everything is in one stage or another of it’s producing journey. The vegetables that start as seeds and become seedlings all have timings for when best to plant them which are dependent on the weather, the soil, the labor available, and the intended harvesting window. Necessarily things have to be incubating in order to reach production, but not everything incubating will. Some abundance eventually has to become compost or feed for the animals.

Unlike my prior experience in software engineering which only wanted to see things in production rapidly and never anything that looked like incubations, I think the lack of competence around appropriate buffers and timing in software development leads to problems in process variance and what to do in times of scarcity. Having “credible investments at the ready” was the idea from Dr. Geoffrey Moore that I built a small and short-lived practice on…and at the wrong company…which might have looked like a few extra seed trays more than was in production, but there’s a reason why the farmer builds a little extra in to the plan.

Yesterday’s Tech Problem, Fixed

Speaking of what happens when you don’t have a backup plan…

Whatever the reason my farm phone wasn’t retaining the photos I took yesterday, it worked fine today. My real phone took a tumble a while ago, but now the outside glass cracks are so bad that they spread to the camera lens so I can’t take pictures with that one without using the inside/selfie camera…not optimal. Both are Moto Flips and I’ve ordered a duplicate 2024+ model which will become my ‘don’t ever take this out of the car’ phone when I’m at the farm. Phones that fold up don’t take kindly to the dirt and water frequently encountered in the field.

The photos are important to me for a few reasons:

1. They’re a visual to accompany this writing
2. The farm business can use them freely, as can others
3. They are future training data for Agro-specific vision models

Summarizing Field Notes with AI in Production

Summary

We harvested cabbage, celery, carrots, and basil. A surprise spider and tomato plant were encountered in the field while weeding. It was another hot day, but hopefully the last of a three week heat wave, at least for a little bit before August.

To my corporate readers: what is your employer’s approach to addressing bugs, either in operations or in its products? If it’s not managed properly, chances are that broad-spectrum methods like lay-offs won’t keep the company afloat for very long.

Bugs Around the Farm

There are many, many types of insects on the farm. Not all are ‘pests’, but many are. There are cabbage butterflies (a.k.a. Pieris rapae) whose larva leave little poop and bite marks on the outer leaves (which we throw away anyway). There are aphids, leafhoppers, and other soft-bodied insects that suck the sap out of the plants. There are beetles, caterpillars, and other insects that eat the leaves. There are ants, bees, and other insects that pollinate the plants. There are spiders, centipedes, and other insects that eat the pests. There are greenhead flies and mosquitoes that eat on us and the animals.

One of the photos I took today was of a big brown spider in the basil eating what looks like a young leafhopper. As a gardener and farmer, I’m not a big fan of aphids or caterpillars, and any natural predators that help to minimize these pest populations are most welcome. Like I mentioned in my previous post, applying agro-chemicals like chlorothalonil (by ground or air) has recently been linked to the decline in reproductive capability of bees, and I can only imagine it has similar effects on other insect-preadators as well.

This farm doesn’t spray pesticides, which is really good for many reasons, but it also means that we have to deal with the bugs. Even with other farms that use pesticides, there are non-affected bugs. Organic gardening practices would tell you to do things like knock down aphid and mite populations with a water jet spray, apply a vinegar and/or neem oil spray, or even introduce nematodes to keep specific pest species. On a farm at scale, all of these methods get cost and/or time prohibitive.

There are always bugs in nature…and especially in software. Some are your fault in your own code, some are not (in dependencies you bring in), but you always eventually have to deal with them.

Bugs Around My Homelab

Not all bugs manifest themselves in the same way. Some are more obvious, annoying, destructive, or dangerous than others…just like bugs on the farm. In my opinion, the priority of addressing bugs goes:

1. Mission-critical functionality
2. Security-related vulnerabilities
3. Data destruction/loss issues
4. Performance and reliability issues
5. Everything else, including new functionality

The reason for ‘new functionality’ being at the bottom is not that it’s not important, but everything above it undermines new work if it’s not addressed. If I were a 10-person team then we could split up these responsibilities and work on them in parallel…but of course the more people, the more communication and coordination become a factor. Right now, my team is just me.

Cancellable Components in Autoscribe

About two weeks ago, I noticed that sometimes Autoscribe would stop summarizing audio files. I was able to track this down to a bug in the way the Langchain and ollama libraries don’t provide signals to cancel ongoing requests. When diagnosing this in my development environment, I also realized that they don’t handle SIGTERM or SIGINT signals well either (the other way around vs. the original problem).

At that time, I looked at how others tend to deal with this and found that wrapping these types of calls in async subprocesses/threads using the ‘asyncio’ library was the conventional approach when it’s a code dependency used by the main application vs. a shell command with it’s own subprocess ID to kill. While functionally bug-free, this introduced an even more insidious and inconsistent bug that only happens in production after the container is running for a while and doesn’t happen when automatically retried again…the dreaded ‘Event loop is closed’ error I now see in Slack notifications as output of monitoring the process.

Canceling Workforces to Maintain Profitability

This weekend, I know I’ll fix this and move on, but it’s an example of unanticipated time-spend due to using a 3rd party library that I don’t control. This happens all the time in software development, not because of human error or oversight, but because of the nature of the work itself. You can’t cancel bugs…or the need for skilled tech workers…just because balance sheets and speculative management thinking say it’s the easier way to deal with maintaining profitability margins.

Who’s going to solve for all the work AI will never get right…the unanticipated, the insidious, the undocumented, and only operationally evident bugs? I just hope that companies stupid enough to be laying off skilled and needed workers will be the first ones to drive themselves out of existence when their lack of organizational and cognitive bandwidth hits critical mass.

Both expansion and contraction in workforces, in their pay, in requirements are all part of every industry. It’s the nature of the beast. Specifically, management culture is the primary driver of this. Mass lay-offs are not easy decisions, but they’re a damned sight easier on those making the decisions than on those being laid off. You can’t cancel your board or shareholders expectations for constant growth, or the real root problem which is that many tech companies were never built to survive contraction cycles. Companies might cancel workforces, but market forces cancel weak companies.

AI Only Gets Right What We Feed It

What isn’t bug-free is the summarization phase despite adding try-catch handling around cancel and finalization calls. The code problems I will fix, but I’d really like to get back to the real work of improving the quality of the summarizations by adjusting prompts, context, and models I use. Harvesting, in this case would be making progress on higher-order ‘business’ use cases such as easy print-outs of my blogs to send to my extended family members who aren’t on social media or email much.

For today, Autoscribe finally produced the below AI-generated summary…after the customary first-attempt failure of course:

Consolidated Summary

Main Themes

1. Weather: Overcast and humid morning with warming afternoon conditions.
2. Crop Harvesting: Multiple crops (cabbage, celery, carrots, basil, tomatoes) harvested with varying methods and maturity stages.
3. Farm Maintenance: Weeding, irrigation line adjustments, electric fence checks, onion washing, and pig watering.
4. Observations: Unusual natural occurrences (a large brown spider, a lone tomato plant).
5. Farm Operations: Packaging electric van, battery maintenance, and crop tracking.

Activities Performed

• Harvesting: Cabbage, celery, carrots, basil, and tomatoes collected.
• Weeding: Removed weeds from celery, summer squash, kale, and pepper rows.
• Irrigation: Moved sprinkler lines and hoses.
• Maintenance: Washed onions, packaged electric van, watered pigs, replaced electric fence batteries.
• Observations: Noted a large brown spider and a lone tomato plant.

New Things Encountered

1. A lone tomato plant growing without support.
2. A large brown spider (species unknown).
3. A specific tomato variety (likely wild or hybrid).

Questions for Further Research

1. What species is the large brown spider?
2. How did the lone tomato plant grow without support? Is it wild or hybrid?
3. How long do electric fence batteries last? Are any from last year or earlier?
4. Will the lone tomato plant continue to grow or produce fruit?

Suggested Actions

1. Document the Spider: Photo and consult an expert for identification.
2. Monitor the Tomato Plant: Track growth and fruit production over time.
3. Inspect Electric Fence Batteries: Test lifespan and replace outdated ones.
4. Record Weather Impact: Note how weather affects crop growth and harvesting.
5. Expand Weeding Efforts: Target persistent weed areas (e.g., around summer squash or peppers).

Part 1

Main Themes

1. Weather Conditions: Overcast and humid morning with a warming afternoon.
2. Crop Harvesting: Multiple crops (cabbage, celery, carrots, basil, tomatoes) were harvested, each with distinct methods and stages of maturity.
3. Farm Maintenance: Weeding, irrigation, and equipment maintenance (e.g., electric fence, sprinkler lines).
4. Observations: Unusual natural occurrences (a lone tomato plant, a large spider).
5. Farm Operations: Onion washing, pig watering, and battery maintenance for the electric fence.

Activities Performed by the Intern

1. Harvesting:
   • Cabbage: Picked mature, undamaged heads.
   • Celery: Harvested two blue buckets worth, noting growth progression.
   • Carrots: Used pitchforks to extract carrots 2 inches below the soil surface.
   • Basil: Cut 2 inches above soil, harvested in bunches for $5 value.
   • Tomatoes: Harvested ripe cherry tomatoes and hoop house tomatoes.
2. Weeding: Removed weeds from celery, summer squash, kale, and pepper/shishito pepper rows.
3. Irrigation: Moved sprinkler lines and hoses.
4. Farm Maintenance:
   • Washed onions.
   • Packaged electric van.
   • Watered pigs.
   • Checked and replaced electric fence battery.
5. Observations:
   • Noted a large brown spider (species unknown).
   • Discovered a lone tomato plant producing green tomatoes.

New Things Not Yet Encountered

1. A lone tomato plant growing without a trellis or support.
2. A large brown spider (species not identified).
3. A specific type of tomato plant (likely a wild or unmarked variety).

Questions and Future Research Areas

1. Spider Species: What is the species of the large brown spider observed?
2. Tomato Plant Origin: How did the lone tomato plant grow without support? Is it a wild or hybrid variety?
3. Battery Lifespan: How long do the electric fence batteries last? Are some from last year or earlier needing replacement?
4. Tomato Plant Growth: Will the lone tomato plant continue to grow or produce fruit? How does it compare to cultivated plants?

Suggested Actions

1. Document the Spider: Take a clear photo and consult an entomologist or online database to identify the species.
2. Monitor the Tomato Plant: Track its growth and fruit production over time to determine if it is a wild, hybrid, or cultivated variety.
3. Inspect Electric Fence Batteries: Test the battery lifespan and replace any that are over a year old.
4. Record Weather Impact: Note how the weather (overcast, humid, warming) affects crop growth and harvesting efficiency.
5. Expand Weeding Efforts: Focus on areas with persistent weeds (e.g., around summer squash or pepper plants).

Because of the good weather, we harvested lots of carrots, beets, celery, eggplant, fennel, and peppers. Similar to the way We empty beds to make room for new crops, I also have divested effort in some of my prior technology works.

Good Weather, Good Harvesting

Even on an organic farm, there are many timing ‘knobs’ that can be adjusted to make sure that you’re harvesting at just the right time. Mostly this involves knowing when the crop is ripe and ready to harvest, a long-cycle outcome of planting and caring for the crop well in advance of the actual harvest.

Weather however, is not something you can control, even sometimes with greenhouses. So it’s nice when nature cooperates and gives us a few uncharacteristically cooler mid-summer days to work out in the fields this week.

Yesterday we harvested lots of carrots, beets, celery, eggplant, fennel, and peppers…all before 3pm. The bugs weren’t bad because of the breeze. It was a lovely Monday for hours of physical activity outside. If only every day could be like this one.

All Things End

Some of the beds on this farm host the same crop season-long while others are rotated to different crops even within the same season. Either way, they have a lifecycle that ends at some point, but this is so they can begin again.

The heavy farming season around here will end at some point likely in October. This gives impetus for me to line up my winter tech work and contracts with clients to keep me busy until at least the spring. I may or may not also take that time to narrow in on a new career path and/or startup.

Indeed though, I have been putting many seeds in soils both on the farm and from a career perspective over the past year.

Specific Things that Must End

Last year was my final year for a few things. Specifically, I left an employer to pursue possibilities in agriculture. I volunteered on some farms and found one that I liked enough to work for the following (this) year and this year I no longer volunteer with others because there’s just not enough time to do so.

I also amicably ended my volunteer work with the Boston DevOps meetup, event organizing and non-profit board behind it. I divested because after 5 years I wasn’t seeing the growth I needed to justify the time investment. During the pandemic, I served as the president of the board, and put in place a succession plan that worked as intended.

In that time, we donated over $50,000 to local charities because we had done a great job in prior years of maintaining a strong community and fundraising. Unfortunately, there was a long-tail economic effect of the pandemic that caused many of our sponsors to divest from community events such as ours. It got harder and harder to sign on new sponsors in a model that never had that problem before 2020 and like many of our sibling events across the world, we saw a major decline in attendance and sponsorship.

All throughout that time, I was ‘planting seeds’ for how that community would rotate responsibilities and incorporate new folks while maintaining its core values. Some of those seeds blossomed and there are new community members and leaders now. Some of them, more deeply personal ones, did not. I did more than what most would to make sure the trust and respect was maintained, but I have to move on.

Divest to Invest

The fact is that I have limited resources…we all always do, some far more than others, but no one has infinite money and/or time. I have to divest from some things to invest in others.

I knew this in 2013 when I left the mundanity of cubicles at a college for high tech and higher pay, in 2016 when I switched from a toxic company to something higher paced, in 2021 when I handed off a lucrative front-of-house role so that I could run an incubation engineering team, and last year when I decided to leave the corporate world to carve a new path.

Like the tilled up radishes and carrots, you have to let some things go before you can move forward.

Daily Operations of This Blog

In the past two weeks, I’ve worked out a homegrown automated tagging strategy for all the images on this blog that are used in the gallery page. This is a process addition that takes a few minutes each time I add a blog post so it fits within the time constrains I set for myself each day.

For today at least, here’s the AI summary from my field notes:

AI Summary from Field Notes