Episode 172: Systems, Software, and Stewardship with Matthew Fitzgerald

Hi friends, this is John. Welcome back to the Regenerative Agriculture podcast where we have all kinds of fun conversations related to improving our soil health and improving the health of the things that we're growing. And in today's world, today's macroeconomic climate, perhaps one of the most important conversations is how do we maintain and improve the financial health of our operations? Because obviously that is a foundational prerequisite upon which everything else rests. Because if we don't have the resources to be good stewards, then we all know what happens. Stewardship suffers and declines and the things that we're responsible for managing necessarily go into decline. And this has been a macro trend for the last century or more. You could argue even longer that there have been all kinds of externalities in agriculture and that there has been this extractive, this financial extraction that's occurred from agriculture, but this has really been accelerated in the last 50 years or so and is continuing. And so it's important for us as stewards and as farmers to be thinking about how do we put ourselves into a different position, how do we manage? This is one of the topics I'm going to be speaking about next week here at the ACRES conference. But anyway, too much of a monologue. I'm delighted. I'm joined here today by Matthew Fitzgerald. Matthew, thank you for joining me. Tell us a little bit about your story, your background, the scope of the things that you're working on on your farming. Minnesota.

Yeah. Hey John, thanks for having me. I appreciate that opening comment because, well, I'll tell our story and how it connects to financial stewardship and we'll see where it goes from there. So I'm a second generation farmer. My parents, Anna Joe moved to Minnesota from Iowa in the year 2000, so 25 years ago. And we are 100% organic grain farm. And I'll share a little bit about the origin story of our farm and how we as beginning farmers, kind of I would consider myself a 1.5 generation because my folks started in their 40s, they were kind of midlife crisis, wanted to become farmers, raise, raise kids in a farm farm context, but weren't multi generation. So how did they get into it? And really that was the organics niche. And our farm has really paralleled the growth of the organic industry. And now as I return to the farm thinking about this next generation, regenerative is kind of the next thing that comes along. But I'll give some more background about our farm and how we got started and what we're up to today. So the Farm was originally owned by a woman named Mabel Brelia. And Mabel was an obgyn doctor. And she was really interested in women and children's health from her practice. And so in her retirement she moved back to her family's farm and took over. And so in her 70s she was farming organically like in 1996 as an experimental creative, pushing the edges of science, thinking about the whole system health. So she had converted her family's farm in the late 90s to organic production. Met my parents at a farm conference. Want to, you know, sit down at lunch, stand next to somebody and say, who are you? What are you up to? Hit off conversation. And my, my dad had grown up on a farm in Des Moines, Iowa. Big family, 10 kids. The farm was lost in the 70s. Parents said whatever you do, don't farm, you know, stay away from it. It's a tough way to live. But there was sort of this through line of, of being connected to the land, wanting to raise a family on the land. And so my parents met Mabel at this conference and struck up a conversation. And she was in her 80s, ready to re retire. And every, everyone in the neighborhood said, why don't I come by your farm? And she said that sounds great. It has to stay organic. And nobody wanted to do that at that time. So she was out looking for somebody and met my parents. So we started, we moved to the farm in the year 20, 25 years ago. Very humble beginnings. 200 acres, borrowed money and a 4440 and was really, really the early days. And so that's, that's the origin story of our farm. And I share that because getting into farming is like the mafia. You know, you need to know somebody or have a lot of money. It's incredibly difficult. And it's an issue that I'm passionate about thinking about beginning farmers and repopulating the land and connecting communities of food. And you just can't do it in conventional commodity agriculture. And so that was the story of our, our farm is that how could you support a family? And so we looked at the organic niche. We're considering it for environmental, ethical, religious beliefs, wanting to be stewards of the land. But then also the economics at that time, a higher, higher price for the commodity niche, markets, direct relationships, all those things. And so our family was able to start with basically nothing. And over the last 25 years we've grown to about a 2,500acre farm. So really parallel the growth of the industry. Um, much of that growth has been still in the grain world. So we produce corn, soybeans. Those often end up in the feed grade market that end up in feed for chickens and livestock. Some is food grade, some distilling, some food grade soybeans for end users like tofu, soy sauce, wheat. We have some canning crops available to us. So we do peas and sweet corn and then edible beans, black beans and kidney beans, which are really fun and then some cover crops in that rotation. But we're, we're here on the edge of the prairie, so our ecosystem is really right where the big woods hits prairie pothole and it gets flat and wide. And so both the landscape changes and the type of farming changes. So 20 miles to the east of us, it's rolling hills, old dairies and silo harvesters and 40 acre fields. And 20 miles to the west island, you know, grain infrastructure that looks like it could be a whole cooperative and it's just one individual farm and they're 20,000 acre farms. So we're nestled kind of in this geography, cultural change, and we've wiggled our way through it as kind of a 1.5 generation farm with the organics. So that financial stewardship and really observing the loss of the small family farm and the, the mega farms emerging over the last 25 years.

So we've touched on the financial management piece and the need for kind of economic independence in this very challenged world. You also touched on the difficulty, the barriers to entry, the difficulty of getting in as something that you're quite passionate about. What's the intersection of those two things and how are you thinking about them in the context of your operation today?

Well, we, we do a couple things on our farm, so we're trying to create opportunities for other farmers. Pay that forward. I think that's like a real, a real commitment that there was a, a gift from Mabel to our family and we want to pay that forward. So we've dedicated ourselves to helping other farmers get into it and especially folks making the transition to organic. So anytime somebody gives me a call and says, hey, I'm thinking about organics, the door flies open and we invite them in, start having that conversation, talking about strategies, transition, market relationships. My father Joe has really mentored dozens of folks over the years and we've helped get folks started and really try and continue that, that opportunity creation that was given to us. Pay that forward. I think about grain farming specifically as a more sustainable way than the, the vegetable world. So a lot of my friends, other, you know, idealistic millennials that were out there trying to change the world, got into vegetable Farming. And what I see is a pretty consistent pattern of folks going to intern on a farm, learning about farming. Maybe they're 17, 18, 19, into their early 20s, they may be figure out how to buy 20 acres of land. In their early 20s, they hit 35, they've got two kids, their bodies start to break down, they realize, oh, this isn't working out, the economics aren't working out. So at least here in the Midwest context, I'm always encouraging folks that are beginning farmers or first generation farmers to think about pushing beyond just the quick natural solution of like being a CSA vegetable farmer. I think that's a really important part of our food system, but it's still undervalued. And so grain farming is a huge opportunity for beginning farmers because there are systems in place, there's scale to a certain amount, there's mechanization that helps protect your body over time. Um, and economics are still a little bit more robust and a little bit more stable. So at least start with grain. Think about maybe adding livestock for diversification. If you can complete the whole cycle with, with vegetables or you know, direct to consumer stuff. That's, that's kind of the sweet sweet spot, I think, for a really sustainable economic farm.

So when, when you think about grain farming, of course, even organic farming, there is still a substantial barrier to entry relative to a fruit and vegetable farm, relative to the size that is needed, the acreage that is needed. So how do you solve that barrier to entry?

Good question. It's, it's not easy. I certainly don't want to suggest this. Just, you know, go buy a tractor and get going. We have an opportunity because there's a huge generational shift. Right. I think half of the land is going to change hands over the next decade. And so that, that changing hands is the opportunity, I think, because just going out and being able to buy a farm is darn near impossible given how expensive land is in our context here. Minnesota land is 11 or $12,000 an acre. Tractors can be hundreds of thousands dollars combines. So I think where beginning farmer can think about partnering, where they can maybe offer their labor, their energy, their human capital over their financial capital in partnership with farmers, retiring folks, folks who are aging out. One of the things that we did here in Minnesota was worked on a beginning farmer tax credit. So in 2017 I worked with the National Young Farmers association and we worked on a beginning farmer tax credit that incentivizes the sale or lease of agricultural assets, that's land and equipment to beginning farmers. So there's some Kind of policy, financial mechanisms you can try and do. One of my marketing techniques is to find folks who are retiring and mail them boxes of our product. So as an example, like we have black beans. I literally mail people cans of our black beans and say, hey, I'm in the neighborhood. We're a beginning farmer. Here's our story. We raise food and it's good, it tastes good, and this is what it looks like. And if you're interested in something different, I'd love to talk. So it still requires kind of being scrappy and being creative to get your way into the industry.

Well, you know, what you're describing is particularly that your, your last thing that you described of reaching out to people, you, you were already established, you already have a foothold. And I think there are, there are many young people who want to get in, who want to get started, who've become quite cynical about the older generation. You had an exceptional opportunity in connecting with someone who cared more deeply about mission than they cared about money.

Yeah.

And the reality is many, and this is not true for everyone, but there's this cynical perspective by the younger generation that the older generation is just going to cash out, sell out to investors, get as much money as they can, that they, they care more about the money than they care about the next generation. So how do we navigate that landscape?

I kind of partially agree with the cynicism. It's not just made up. We operate within a capitalist society that drives the most return on investment, the highest amount of money possible. It's very tempting. So I can be empathetic with the folks that make that choice because that, that's very, very powerful. And it, you know, frankly, it's going to be hard to beat. So I think you do have to just place yourself in spaces where people are mission aligned. You know, think about the reason that that meeting happened between my parents and Mabel was they were at a conference with like minded folks thinking about organics. So if it's the Acres conference, if it's Marble Seed, if it's a community gathering around sustainable food, if it's going to your farmers market and introducing yourself to the folks that are at the farmer's market, I think placing yourself in spaces where you maybe find those like minded folks is probably the best way to do that. The cold calling that I described is really difficult. That's, that's a, that's almost a used car salesman gig. It's, that's a, that's an uphill, that's an uphill battle. And I don't know if I would.

Describe it that way.

Well, I've sent lots of cans of black beans and I've only had one or two opportunities come out of that tactic. So it's not like a perfect tool. But yeah, your point's well taken. It is incredibly difficult and I think it is okay to be realistic about that. And so creating opportunity, creating luck, creating those connections, set yourself up in an environment where that could happen.

Yeah, we could have a lot more conversations around that topic, but I want to dig into the way that you set your farm up for financial success, both historically and now. I think a critical part of this story, particularly in today's environment, although I would argue it's been true for the last couple of decades, is the need to decommoditize yourself by any one of dozens of possible different mechanisms, which I think you have done. So let's talk a bit about your thought process. How do you think about your farm's economic viability and sustainability? You've, you've also grown, you've grown 10x over the last couple of decades. How were you able to accomplish that?

There's, there's, there's probably two or three key components. I'd say one is, one is the partnerships in your team. The second is that decommodification. Why don't we tackle those first two things and I'll come back to the third thing I'm thinking about. But the first thing is the partnership and team. And both, it's like your internal team. So that scale requires having folks who are on the farm every day, who are bought in, who believe in that mission, who are willing to work, who are bringing their whole selves. And so we have a really, really strong team on the farm. I farm with my father, Joe. We have one or two full time folks, we have some retired farmers who help us out, who are excited about the organic thing. It reminds them of how they grew up. And when we meet, we talk about why we do what we do, you know, our care, stewardship of the land, our commitment to proving that sustainable agriculture is competitive. You know, we play the underdog card a little bit, know, we motivate everyone around that, proving that we can do it. And then it's, it's the other partners that we have. So it's, you know, financial partners. We, we bank with a group called Matt Agriculture, which is an organization that's committed to organics and regenerative agriculture. So they share in that vision and mission. So some of their financial tools are more built and designed for organic and regenerative farms, as opposed to just a conventional commercial agricultural lender that maybe doesn't have the context, doesn't understand the cash flows. It's having peer farmers. So on my phone are the speed dial numbers of farmers who are like me, that I can talk to and work through problems, questions. I'd like to say having a good agronomist, but that's really difficult. So I'm excited to talk about.

Let me translate, let me translate that, that call that's called higher aea.

Hop on field hop on. But it is, it is, it is like there's a myth that a farmer has to be everything. And, and we know that that's not true with the plants that we raise, that the, the, the crops are raising our whole systems, and we have to treat ourselves and our businesses as whole systems, and we have to bring in all the right elements and have them complement and interact with each other. So that's how I think about the partnership that allows that, that growth, that's that financial stability. Because we have hiccups, we have droughts, we have crop failures, we have grain buyers that, you know, go broke. You know, there we are not without problems, challenges and, and things going sideways and, and.

Well, I think one of the things that you said that stood out to me was your, your discussion of the team of people that you have around you. And this is one of the things that I see as being so foundational. And many farms have this to varying degrees. But, you know, in, in understanding plant communities and microbial communities, we're, we're developing this emerging understanding of quorum sensing and how, how the entire system changes, the behavior of the entire system changes. When you have enough diversity of perspectives to establish quorum sensing. And it's kind of the same way in human organizations or human communities when you have enough diversity of perspectives, like you need enough different points of view, if you only have five or 10 people and they all share the identical point of view, then you only have. You have a much more limited perspective. You have much less resilience.

Yeah, yeah, it's hard to do it too. You know, I think that it's not an, it's not an easy thing, so. And it's long, slow work. Just like stewardship or soil health, it's long, slow work to build that. And so I absolutely personally benefit from the work that my, my, my parents have done before me. And I feel like the work I'm doing today is for the next generation or for the next farmer. And if you Just see yourself as part of this larger arc, this larger ecosystem. It brings joy to the work too, frankly, because, like, sometimes this is just hard. There's no way around it. It's just hard. It's stressful. Um, for me at least, it's. It's the human element, the community element is really grounding and life giving. Yeah, yeah.

I want to not lose sight of the third thing that you had in the back of your mind.

Oh, yeah, yeah, yeah. So. So I talked about partnership decommodification, really briefly. So I would say we're. We're still probably. Probably 60 to 70% of our crops are what you might consider an organic commodity. You know, they go to a broker that we don't know personally that gets shipped on a rail car to the east coast or west coast and ends up. We're slow.

There are degrees of decommoditization, obviously, but I would argue that the fact that you are growing a crop that's organically certified already to some degree has removed you from the mainstream commodity marketplace.

Yeah, yeah, I would say so. Certainly our corn is worth twice as much as kernel corn sold on the Chicago border trade. And it's a lot more nutritional value and love and effort put into it. So I think that's a fair price. But yeah, the third component I'd say for like, how do you manage the scale or how do you de. Risk things is this concept of doing the right thing at the right time and in the right way. And what I want to talk about is a project that I've been working on for the last three years called Farm Flow. And I think, John, you'll share this insight, which is like, alternative systems require intensive management. It's not a passive endeavor. And I think that's often what people think organic or regenerative or permaculture or something. It's sort of like, well, you just let the nature do its thing and it's going to work its way out. And it's not that at all. Nature is absolutely doing its thing, but you're an active part of it.

Actually, I'd like to touch on that in just, just a moment for. Just from a very high, extremely high level philosophical or principles perspective, as I've been learning more about quorum sensing and how ecosystems function, I think there is an ass. It's possible to develop an ecosystem and develop a landscape where nature just does its thing and there is much less management involved. I think one of the early inspirations of this outside of the ranching and grazing world was Masinobu Fukuoko and his one straw revolution and natural farming and some of the work that he was doing. But it's becoming increasingly clear to me, as I understand and look at different examples of this, that in order for that principle to even possibly hold true requires an abundance of different species and requires an abundance of diversity. That doesn't happen in organic grain cropping system. It doesn't happen in most farming systems. And so I think that principle can be true, but isn't true in most grain cropping environments or even all of them.

Yeah, yeah. And there's this transition period from what you're describing, this quorum sensing that like requires active participation, active management, active.

So management intensity. Absolutely. Most, most grain cropping, to do it really well, to do regenerative management, organic management really well, requires much more knowledge and it's much more knowledge intensive and much more management intensive. You've, you've taken away all the easy buttons.

Yep, yep. So three and a half years ago I had a friend who's a data scientist and I said, hey, I have a bunch of information about my farm and I'd like to think about the question, this is a natural farmer question, how could we raise more yields? How could, how could we raise a bigger crop? And, and what could we do? How could we figure that out? And so he, he was willing to help me on this project. And so we put together data set and this isn't a university level research with randomized trials, it's just our own multi year large scale farm. And we looked at planting date, variety, crop fertility, soil suitability, a whole bunch of different factors on our farm, including we looked at the number of passes, so the number of times we were in a field doing some kind of operation or scouting or intervention or cover crop or whatever it might be, and the frequency. So how often were those passes occurring? We ran this multivariable regression which basically helps us understand like which of these variables has the largest impact on that yield. And we ran that regression. The strongest variable was the frequency of passes. So if we were in a field every four days versus every seven days there's a five bushel increase in corn yield. So the insight was the more often you're in your field, the better your crops and you're going to go, yeah, duh. You know, you talk to any good gardener and they say, yeah, if you weed your garden, the more often you're weeding, the better your tomatoes. But that insight really made clear to us like the active intentional management of our farms drove outcomes that we wanted, whether it was weed control, whether it was yield, whether it was like the success of a cover crop, whatever it was. And despite being good farmers and having 25 years of experience, we weren't always doing that ourselves. We kind of knew this insight. The regression proved something we already intuitively understood. But we had a gap. And so we came up with a solution for that gap, which is the farm flow concept. I can share more about that, but I can see your gears turning.

Yeah, I have a couple of flags that ran up the pole for me here. So one of them is this. When we look at plant nutrition management from a physiological development perspective and the phenological growth stages, it's become extremely clear that there are windows of time where if you get on the necessary application, let's say I'm just making things up here, but let's say you get on a critical side dress application on corn at V5, and if you miss that, if you have like a three to five day window and if you miss that window and you're a week late, then your yield response is down disproportionately like you can, that can cost you 10 bushels in yield. And there are all these various windows. So on one hand, when you talk about, I understand the critical need to get certain management activities done within very well defined windows, I understand how important that is, but that's not exactly what you said. You said the, the frequency of and the number of times that you're in the field. And so if you take that all the way to the extreme, you could say, well, if you're in the field every three days, then for the entire season that should produce the highest yield. Right, but I don't think that's exactly what you're saying.

No, it isn't. And that's the limit of a farmer run data set and experiment. There's an insight that can be gleaned, but if you take it too far, it falls apart or it's limited in its utility. So from this Insight, we created a $300 ideas. I think it's worth a million dollars and we can give it to everybody for free. Um, but we, we basically took this insight of like, okay, we need to visualize the work that's being done on the farm and, and catch those gaps to get that side dress at that right time. And so we took a 4 by 8 whiteboard and we took tiny little tape and we laid out, we made a grid 4 by 8 grid, and from top to bottom listed our fields. And from left to right, counted out days. So individual squares equal to individual day on an individual farm. And then we just simply use magnets and we place the magnet with a color code. So for example, planting was green, tine, weeding was blue, manure application brown, so on and so forth. And what we're able to do was literally just chart out the work that was happening on the farm by placing magnets. So you could walk into the gathering shop and look at this giant whiteboard and without even reading any details, you could visualize, here are these patterns of work that's occurring from seeing the, the colors and then the gaps in the colors. And so really quickly that team, you know, retired farmers, part time, seasonal, full time folks, all could quickly understand like what was happening on the farm, what was going on, what was getting behind. And all of a sudden what changed was we were able to respond like maybe a 3 inch rain occurred and projects were delayed, or a piece of equipment broke, or a farm that was maybe further away or had a grumpy neighbor, that we had a subconscious bias, no one wanted to go work at it. We could identify those things and respond to them. And so people had a sense of their role, their participation, their impact in the whole system. Folks that didn't were new to organics could start to understand the sequence of these interventions, these passes and how they complemented each other and how they built on each other. And as a result of just this $300 whiteboard with magnets up on it, we changed the way that we were thinking about the farm. Everyone had a high level buy in, we were executing much more quickly. We're catching those gaps, the management gap. And it changed the flow of our farm. And so that was the farm flow concept.

So let's, I want to talk about the results that you observed and what changed. But before we go there, let's dig into a bit more on what this chart actually looked like on the wall. So I'm imagining that over the course of a season, you have some gaps that are three days apart and you have some that are 30 days apart.

Yep.

Do you deliberately try to bring things closer together and have greater frequency or are you just trying to, to hit those optimal time windows?

It's, it's seasonally appropriate. So, you know, the activities that are happening say in August and September are much more spread out than they might be in May and June on our farm and a grain system. So in May and June I might expect something to be every three to five days. And by the time we get to September, Might be once every 30 days. You know, we might be applying fertility and seeding a cover crop and that might be the only activity within those 60 days. So I can, I don't know if you're. It won't be helpful to the listeners, but I could share an image of it for your benefit, John.

Just, just describe it visually, I think.

Yeah, yeah. So there's a limit to four by eight whiteboard because you can't fit 365 tiny little squares or on a four by.

I think that means a second four by whiteboard.

Yeah, we just keep stacking together or you know, take a photo, you wipe it off. I also have a 2 year old so we have to be careful about how high those magnets are because she can come in and move them around pretty quickly and the whole plan's off. So yeah, the, the tool itself really became a gathering point or a conversation for you were talking about the, the diversity of opinions. And so we all knew what, what the reality was. But then we could talk about in the morning like hey, what's going on? What are people seeing? What do we think we need to do? And it gave us a reference point that we could start to talk about as a team of planning out. Because before really everything lives in your head or maybe if you're hyper organized, maybe on a spreadsheet or on a.

Piece of even a spreadsheet. It's a spreadsheet that one person looks at instead of 10.

Yeah, yeah. And so it just totally changed the dynamic of what people's ownership was of the, of the farm. And it became a teaching tool. And I think that's part of, part of our joy of our farm is like the teaching element.

So you've been working on this now for the last three years. How has this shifted and how has, how have things changed on your operation as a result of it?

Yeah, so we got to the end of the first season and we realized that everything up on that whiteboard we were going to have to type up for our organic certification. And so previously what we were doing was just having a giant paper, you know, an office supply paper calendar where we're scribbling today. We did this and then that was part of our certification process. Totally fine, totally legit works for record keeping, is accepted by the USDA for that process, but kind of labor intensive. And if we wanted to go back and refer like hey, what did we do in 2012 when we had similar circumstances, you know, how could we reference that? And so it became clear like this physical whiteboard solved the initial problem, but there was a natural evolution to maybe a digital tool. Could, could we build a piece of software that functions like this whiteboard that could assign tasks to team members? We could pull in weather so we could geotag it and have weather pulled up automatically. Could we have multi year references so we could look back two or three years in similar circumstances and see what worked or what didn't? Could we have details about who did the work so if it looked like the, the cultivator was set really well, we could all learn from that or. And so I started the process of developing formflow, a piece of software and that's been the last three years journey. I am a non technical person. I have no computer bones in my body and so it's been a learning process there. The first year I hired a team on upworks, which is a online gig economy tool. I hired a team in India to build the software. I thought, oh boy, ten grand I could build this thing that's gonna be amazing. It'll change the world. How hard can it be? Boy, was I wrong. So between the barriers of like language, farming context, the farming context in India versus the Midwest grain farmer, wildly different and really just being non technical, just there's a whole language that's spoken about how do you develop software? And what I might visualize in my mind communicating that to a team was it was just a mismatch. So I spent a year, a chunk of money and I built a bicycle that didn't have wheels or a seat on kind of looked like the thing, but it was not going to work. We restarted the process here in 2024 with some grant funding. So the project today is a grant funded software tool and I'm working with a team in Minneapolis. So a US based team have a product manager who helped build a large software tool that vegetable farmers use. And we're working with four or five farmers here in the US to kind of test it out to learn from what they need in their farms. And so we're, we're restarting, we're relearning, but we're, we're building a tool that takes that whiteboard concept and starts applying it to a digital world. And eventually we're thinking about AI. So I'm, I'm excited to talk with you about AI too.

Yeah, I, I want to dig into the AI conversation, but I want to not lose sight of my original question quite yet, which is what's changed on your farm in the last three years? As what, what in terms of crop responses or crop management. You spoke about this being kind of the, the nexus point for conversation and for facilitating discussion around different points of view. Which is, which is awesome. But has that, has that resulted in actually improving yield, which was your original question?

Yeah, it has improved yield. So the first year that we used the Farm Flow board, we had our record yield up to that point. And of course, weather, fertility and everything else has a large role play. So I would be very, very remiss to say that if you put a 4x8 whiteboard in your farm shop, you're going to have the best yields ever. That would. That simply is not true, folks. That will not happen. But what has changed is that the ownership of the operation has changed. And I think as a result of that, we have better execution day to day. And as a result of better execution day to day, we have more consistent yields. And what it is doing is like reducing our lived experience of stress. What it is doing is like taking the record keeping process and making that much simpler year to year. And you know, this will be our third season with the, with the tool in place, so I'd have to look at, you know, 25 years and say, have. Can I chart that out? Hard to say exactly. You know, again, weather, weather, fertility, crops, everything else comes into play significantly, but it's changed the day to day experience in our farm significantly and helped our team.

Got it. All right, let's talk about AI. You've mentioned Field Arc twice now and I have conversations with people all the time who are using AI for a variety of purposes and in usually partially delighted and partially terrified and partially. Yeah, there's just a mix of emotions about the results and the performance that they're getting from it. There's delight and occasional frustration in equal measure, particularly from some of the more mainstream tools. So what have you been playing around with?

Well, I did have an opportunity to use Field ARC when it launched. I was excited to see it and we talked earlier about.

Oh, were you one of the, you were one of the original beta users?

Uh, maybe, maybe I signed up. Yeah. Um, we have a, we have a great long time conventional agriculture agronomist who I called the help get soil samples and if something seems really confusing, I might call them or send them a photo. But beyond that, we are on our own. Uh, and so I uploaded all of our soil test results into Field Arc and I started asking it questions like, okay, let's, let's walk through this. What can you do? And I had a more robust conversation with field lark than I've had with my local agronomists over the last six, seven years in about 30 minutes. So in this way, it was really exciting.

Out of curiosity, what was the nature of the questions that you were asking and what was, what were the types of responses that you were getting back?

I start out, I'm just pulling it up here. I started out very generic. So I said, attached to is my soil samples. What can you tell me about my farm? And I got a long response. And then we started digging into it. So then I started saying, you know, on our farm we have some peat soils which are really high in organic matter. And so that was throwing off some of the model instead of said, okay, remove the fields with, you know, super, super high organic matter. Let's, let's dig into some of my target yield goals. You know, what can you tell about micronutrients? What suggestions would you have for increasing these results? Um, I started asking about foliar boron and zinc on corn and, you know, got interesting answers there. So it was. It's been fun. I think my experience with AI has been, I think about it maybe as like a college intern where it has access to knowledge, but not quite the context to offer original insights that it can assemble and answer things that you maybe give a framework for, but it's not generating for me, at least, totally new and original concepts that the, the temptation. My experience, and this is, I'm talking about ChatGPT or other large language models, not necessarily Field Arc. But my experience with, with AI is this, maybe this same sort of delight and distrust that initially it's really tempting to interact with it and just think that it's going to do the thinking for you. But where I found AI to be most helpful is if you come to the tool with a question with context and have done the work of thinking before. It's helpful if, if you're approaching the thing with, with not much thought or without much inquiry or, or kind of context, the, the results you get back are mediocre.

Yeah, yeah, I think there's, there's a fundamental difference. Obviously, I've, I've thought a great deal about the development of AI and the work that we've put into field arc. And there's a fundamental difference between the way humans respond to answering questions and the way a machine responds. And there are two key ones that I'll focus on for the moment. One of them is, okay, you're an expert organic corn soybean farmer, and for your local environment, you've got, I mean there's no question, you know way more about that than I do. And I ask you the question to say, hey, I want to grow organic corn. How many, how many cultivation passes do I need to make for effective weed control? And you're going to say, well, it depends on this and that and the other thing. It depends on rainfall, it depends on, there's, there's a whole bunch, bunch of variables like you, you, you can't give an answer to that question without knowing a whole bunch of context. So knowing the context is important that any human is going to say, well, I don't know the answer to that. It's like there is no cut and dried answer.

Yeah.

And we will respond, humans will respond by asking for more context or describing that more context is needed. That's one significant difference. And the other significant difference is we'll ask about one thing and for someone who is an expert or has experience and knowledge in a, in a certain area, they will say, well actually while that might appear to be a good question, you're actually completely missing this thing over here. And what you should be thinking about is you shouldn't be thinking about X, you should be thinking about Y. The machine can't do that. Yeah, doesn't do that or very does it? Does it very poorly at this point. And so the, you know, I had this experience some of, I've had some of these interviews both here on the podcast and in and in person who that haven't been publicly released yet or interviewed people from a military background about research work that they did decades ago on plant pathology and so forth. And when you get into some of those conversations that have national security implications, the answers become very direct and concise. They become yes, no. It's like they will only answer the question that you asked and they don't elaborate. And so in order to have any type of conversation or any type of dialogue, you have to be smart enough to know what questions to ask. Because if you don't know the questions, they're not going to volunteer the information. And I think this, this same thing is so true of, of these AI tools like Field Lark is you have to, to some de extent you have to be smart enough to know the question you ask the question to ask. And this is one of the things that I've tried to develop in thinking about how do we develop these tools to help people become better, to help people ask better questions and so forth is in Fieldmark we have these follow up questions as part of each response and that is our initial attempt to inspire more of a dialogue and to help all of us ask better questions and remind us of questions that we might perhaps should be thinking about, but that we're not thinking about right now.

Yep. John, as you've been building this tool, how do you, how do you balance the, what I might perceive as risks of AI, you know, either energy consumption or data privacy, or even kind of intellectual, what you're describing, kind of intellectual risk of keeping in the creativity of the human element. How do, how do you balance that with the opportunities or the utility of the tools that you're building?

Well, there's several questions within the question that you've asked, but I'd say broadly I've taken a fairly pragmatic approach to just say AI is what we're currently calling AI, which is largely defined in terms of LLMs, these large language models. And this is, obviously this whole space is rapidly iterating and evolving, but there's a, there's a very pragmatic element of saying this is going to exist whether we create field dark or not. This, this, this whole space is like the guardrails are off and we don't have control of this horse anymore. So there, there is an element of we, we. If you can't beat them, you might as well join them. And let's create something that is better than what currently exists for a specific use case. Let's create something that is better that, that helps people better, that helps farmers and agronomists be more effective and that has significantly more powerful for a specific use case than chat, GPT or GROK or Gemini or any of these other tools that are out there. So that's, that's the one consideration and then the, the other. Another aspect of the question you asked about energy consumption that this kind of relates to, my initial response is we are perfectly comfortable in many domains of life using significant energy and you could argue using energy wastefully for outcomes that we perceive to be valuable. So for example, I'm trying to forget, I'm trying to recall who did this analysis. I want to say it was Wendell Berry, but I don't think it was Wendell Berry. Wendell might have talked about it though in Unsettling America. But something to the effect that our modern agriculture Today we use 3 to 4 calories of petroleum energy to produce a single calorie of food energy, which is not sustainable on its face. It's working for a temporary moment in history. But when you look at overall human history in terms of, of centuries or millennia Then I think at some point we will look backward and say that was a very wasteful use of energy. And similarly, so we have to look at these, these LLMs have been created. What are we going to use them for? How are we using that energy? And so I would argue that using that energy for the purposes of facilitating a more regenerative agriculture is a valuable use of that energy. Much more valuable than what it's probably most widely being used for on a consumer audience perspective.

Yeah, yeah. Looking up a recipe for spaghetti.

Exactly.

Not helping the world. Yeah, yeah. Well, it's. One of the principles that I find is true at our farm is that we've, we, we have done well by being early adopters, by being willing to ask questions, by being able to lean into things, whether that was organic or certain technologies, say GPS for tractors, being able to drive straight so you could cultivate better. Whatever, whatever it may be. We've, we've generally leaned into being early adopters. And so that's been. My approach to this tool in question too is that there needs to be a healthy amount of skepticism. There are always trade offs and at the same time it is a tool and. We know that commodity agriculture is going to be using these things too. And, and it will be to kind of perpetuate those systems. And if we're not thinking about how we're going to participate or respond, I think those conventional systems will continue to dominate. So yeah, I would say I kind of take a similar perspective about using the tools for good, for, for driving change.

Well, in the way that we've developed Field Lark, I'm actually not thinking about it in terms of conventional systems versus organic or regenerative systems. We built it, I built it very deliberately to be, what's the word that I'm looking for, to be practiced? Agnostic or to be, but to instead function from a first principles perspective and to function from an outcomes perspective. And when I say an outcomes perspective, fieldlark is directed to take second and third order consequences into consideration as well to the degree that they're understood or known. And so it's quite fascinating. I spent quite some time thinking about and developing the ethical framework of what's the ethical framework and what are the first principles framework that field ARC is to operate within and what are its foundational directives. And so when you, when you, and it's interesting because we didn't give field ARC any directives to for example, to say don't make recommendations for pesticides or don't make recommendations for anhydrous ammonia or a variety of different things that I could point out. But what is intriguing is that fieldlark will not make, at least not readily on an initial request, it will not make recommendations for those types of things because of its ethical framework and its first principles framework thinking. Or it takes into consideration second and third order consequences just all of a sudden say, well, actually, yes, anhydrous ammonia can be a very valuable, useful form of nitrogen from a crop and agronomy perspective. But anhydrous ammonia also has all of these other externalities that are perceived to be unhealthy to the environment and to the ecosystem. And therefore Field Lark, without being specifically told to, it will selectively, it will make preferential recommendations for other forms of nitrogen other than anhydrous ammonia. So it's not that it doesn't make recommendations for nitrogen, it's just, it's, it has, it has that first principles approach which is so intriguing because we deliberately asked it directed Field Arc to have no biases, to remove all biases in favor of, of any system. It's not in favor of organic or not in favor of regenerative, or not in favor of, of contemporary agricultural systems, but instead to operate exclusively from a first principles perspective and not take any biases into consideration. And that very fact makes it appear to some people as if though it were biased.

Yeah. Challenge challenging our own worldviews. Yeah, yeah, yeah. When you, when you reflect on what you've learned in the last couple months of releasing Field arc, what's been unexpected?

That's a good question, I would say perhaps, and this is a reflection of my own biases as you described, I tend to be a technology early adopter. I've leaned in very strongly into testing these various GPT platforms when they were first developed. I've been thinking about them for over a decade. I had some early exposure to the technology that became these large language models long before they were publicly released. And so I've been thinking about this for some time and I've leaned heavily into using them and, and building them to work for me. And the piece that I'm perhaps the most surprised about is that many people, many users of Field Lark and I, based on what I can see, I'm assuming other GPT models as well are still using it as a glorified Google search. They're asking more specific questions, but they haven't really understood the implications of what this technology can do for them once it is trained for them. And you mentioned this at the beginning of the con you uploaded a bunch of soil samples. But what you can actually do is within fieldlark, you have these different projects. So you can have a project be a specific field or a farm or a specific crop that you're working on, or just say you use it for your whole farm so you can write up a document or even better. And this is the way that I use Fieldmark and other tools as well. When I'm building out an initial profile, I won't build a profile. I'll use the AI tool to help me build the profile. So I'll say, hey, Fieldmark, I am farming at this location with these crops, with these soil types. Here are my soil characteristics, here are my dominant weed species. Help me build a profile that describes my context that I can use to train an AI tool and you'll go back and forth, you iterate half a dozen times, and I end up with usually as much as a 20 to a 30 page document that describes my context in detail. Here are, these are the common weather conditions, here's the climate, here's the USDA growing zones, here's all this. And let the AI actually help build out that context rather than you trying to do it yourself. And then you take that information and you upload it into a project, put it into a document, put it into that project, and now it will constantly reference back to that training data set and you can, you can constantly add to that and build it. So what happens over time? The information that you get back becomes ever more customized and focused on your specific operation as you keep building and as you keep training it. And so the quality of the answers and the quality of the information you get back out of it becomes very valuable. And the reality is, what I'm talking about in terms of training these AIs to work for you is this can be as simple as 30 minute exchange back and forth. This, they are very good at actually identifying what information would be valuable for them to understand your context.

Yeah, where, where I've used that similarly is with this Farm Flow project I mentioned. It's been grant funded. I'm also not a grant writer, so I've been using it to apply for grants by giving it context. So like sharing previous grant proposals or previous language, previous documents, the request for proposals, you know, pitch decks, kind of all, everything. And I've created a project in ChatGPT and then that, you know, when I get off the farm and tuck kiddos into bed and then it's time to start working on a grant at night and I've got two brain cells left. It's been really helpful to kind of be that partner, to kind of take it to that next level.

One of the projects that I did, just to give give you an example, this is actually one of the projects that I did that I then refused to ever use again. Just from a principle to perspective. I trained an instance of Gemini to, to sound like John Kemp, to talk like John Kemp, and to write like John Kemp. I said, okay, here are. Here are 50 web. Here are the transcripts from 50 webinars and dozens of podcasts. Here are a long list of articles that I've written and blog posts that I've written. And so I trained it on all this and I said, analyze all of this and write a. I forget what the proper term is, but kind of a brand tone document of how does John's language and his expressions and write a document to train an AI so that it speaks and writes like John Kempf.

Oh, no.

And it generated like a 50 page document with this extensive detail about the tones and the type of voice inflections, and it analyzed all of it, wrote up this document, and I provided the document as a training tool, as a training set. And it came back and I said, all right, behave as John Kemp and write an article on quorum sensing.

Yeah.

And I looked at that article like, this is freaky. That's done.

Because.

Yeah, well, it's, it's not that my concern was not that it sounded too much like me. What I was concerned about is that if you start using those tools, it's just very similar to once maps apps were prevalent on every phone. People lost their sense of navigation. They can no longer. People generally, there are exceptions, of course, but people generally no longer have a good sense of direction or a good ability to navigate. And you start using tools like that for writing articles and so forth, you'll start losing your ability to think critically. And so there are ways, I think we need to think carefully about how do we use these tools, like in the way that we've built Field arc. I've thought a lot about how do we use this to not replace people, but to make people better. And there's only so much of that we can do as a tool designer. Another part of that has to be picked up by the tool user. And the tool user has to say, all right, I want this to help me do the things that I'm not great at. It can help really expedite research, for example, or to learn things that I didn't already know. But then when I learned those things, there is still this, this critical thought process that we need to engage in personally, because if we don't, we're going to lose it.

Yep. Your, your comments are making me think about one of the tools we're building. So for Farm Flow, we have the whiteboard tool. We call that grow. We have a tool that we call track, which is just managing paperwork. So your bill of ladies, your grain inventory, we use AI lightly there, just image recognition. You can take a photo of your grain contract extracts, it can generate bill weighting numbers. Pretty simple stuff. The thing I'm excited about is the future. We're calling Farm Flow Guide, which is basically a SOP generator, standard operating procedure generator or best practice generator. If I was to say, john, I want you to go plant corn and the you would have a particular way of planting corn at your farm. Maybe it's based on your soil type, maybe it's based on the equipment you have, maybe the variety you have. And the way that I would plant corn at my farm is going to be wildly different, even if we have the same piece of equipment. And every farm is unique in this way. And so what we want to do is build a tool that partners with the farmer to capture the processes, to document the on farm best practices. And the purpose behind this is to do a couple of things. One, to capture and help retain the knowledge of wisdom from retiring farmers. Farmers are like great innovators, great thinkers. They're not necessarily great teachers or they don't think about themselves as great teachers. And so this tool can be kind of the partner to help the farmer become a teacher to teach next generation. It could help a farmer scale. So to train employees, interns, whatever it might be, that you can take your practices and give these sops to your employees or your team to help them learn and they can interact with it. Rather than just a static document, it's a SOP that you can interact with. But the part that I'm most excited about for Farm Flow Guide is that I think it could actually be an unlock for adoption. So more than an incentive payment, more than an inspiring video that you might watch. If a farmer is going to adopt a new practice, it's probably because they see their neighbor doing it. They see somebody that they know or have context that's similar to theirs doing something different. That farmers trust farmers. And when I think about adoption of organic practices or gender practices on my own farm, it's anything that we've adopted or brought onto our farm, it's probably Because I've seen another farmer do it successfully or they've shared an idea or concept. Maybe it was leaning over the back of a pickup truck at a field field day and we were swapping ideas on cultivators. But we want to, we want to use an AI model that's trained to think like a coach, to think like extension educator and help document the practices that are on that individual's farm and then create these SOPs that can either be used internally or externally to adapt and adopt new practices. So we're calling this thing Farm Flow Guide, just in the early, early stages. So we're working with the same set of farmers. We're really wrestling right now with the data protection. That's an unanswered question for me. And I don't, we're, we're spending time talking with experts in the field. They kind of think about how we protect this data because we don't want to just train a model and then get that extracted that value in that process that's intimate and important and really the intellectual property of that individual farmer. So that's something that I don't have good answers to yet. But we're thinking about that SOP guide tool as kind of an unlock for farm flow and for adopting new practices in our context. So that's, that's to your point about the, the tool and the user of the tool. I think this is like the, the next level for us as we think about AI.

Well, it's certainly a very rapidly evolving and rapidly iterating space. And simply entering into the AI space means that you've entered into a race and you'd better be prepared to move very rapidly because the capabilities of these tools are already today. It's, there's not a good way for me to quantify it for a, for a general audience, a farming audience, of how, how rapidly improving these tools are in their, in their mathematical capabilities, in their logic capabilities. But there is, there is such rapid improvement. What I can say is you cannot look at a tool such as, let's say Gemini or ChatGPT or Grok. You cannot look at a tool six months ago and have tested it, say, okay, I'm going to upload this type of data into it and I'm going to ask it to help me analyze it and come up with a set of recommendations. You cannot evaluate that tool six months ago and say, oh, that failed to give me anything useful and therefore it's worthless and I'm not going to bother using it again. If you test it Again, today it is going to be substantially better than it was six months ago. It's not, not a little bit better, not 5 or 10% better. I'm talking dramatically better. So the whole space is iterating and evolving so rapidly that even if you used it six months ago or two months ago and you said, ah, this thing wasn't all that great, you need to stay on top of it, you need to look at it, take a constantly be evaluating how you are using it, how to get the best results out of it. Because it is, it is growing at a rapid pace.

Yeah, yeah.

So that means that all the users of Field Lark are going to be continually delighted with all the new things that they get out of it from a month to month basis and from farm flow perhaps as well.

Yeah, yeah, yeah, it is. The trick I think is to stay grounded in it too. You know, to be aware, to interact and then to stay connected to the bigger version.

Absolutely. The why and be very deliberate about what you use it for and what you don't use it for. Yeah, wonderful. Well, Matthew, this has been an awesome conversation. We haven't gotten to talk about the farming yet, but I want to be conscious of your time and of all of our listeners time. So thank you for being here, thank you for the work that you're doing and let's stay connected.

Thank you, John. I hope you have a good winter. Stay warm.

The team at AEA and I are dedicated to bringing this show to you because we believe that knowledge and information is the foundation of successful regenerative systems. At aea, we believe that growing better quality food and making more money from your crops is possible. And since 2006 we've worked with leading professional growers to help them do just that. At aea, we don't guess. We test, we analyze and we provide recommendations based on scientific data, knowledge and experience. We've developed products that are uniquely positioned to help growers make more money with regenerative agriculture. If you are a professional grower who believes in testing instead of guessing, someone who believes in a better, more regenerative way to grow, visit advancingecoag. Com and contact us to see if AEA is right for you.