I think that "mixed" farming like this article describes has a huge potential for automation by the software field.
Monocrops[0] (homogeneous fields of crops like corn or soybeans) are easy to build automated robots for. We've been doing that for generations now, starting with combine harvesters[1].
But there are many problems with growing monocrops[2]. Pests are a problem -- they breed and spread disease. And on the soil nutrient side, you have to keep re-fertilizing (and often over-fertilize[3]) every year to keep it going. And still nutrients continue to decline in the food that Americans are eating[4].
If, as a field (of software engineers), we can figure out a way to build robots that allow growing and harvesting plants in heterogeneous environments, that will be a significant step forward to fixing these problems. I'm optimistic that it will be possible in the next decade to build it and make it profitable. :)
If you don't have to use fertilizer and pests don't eat your crops because you haven't created the ideal breeding ground for pests + disease, then this problem should be solved by market forces alone. At least, I hope so!
I was always confused by this as well. Then I spoke to a person working in agriculture, apparently chemical fertilizers are cheap enough to make crop rotation redundant.
As a layman I assumed that was the case. But I understand the bigger issue now is that both soil and crop nutrients are dropping in farms due to soil depletion[0].
Yeah. And not only that, the "regenerative agriculture" mentioned in the article isn't just about sustainable practices, where we put back into the soilwhat we take out. Regenerative practices includes practices where soil fertility is built up year after year as a result of the agricultural practices. In other words, you can take a depleted land and bring it back up into productive land, and potentially, become more productive than non-regenerative practices.
You can grow more than a single plant together, and depending on the combination, they can synergize in a way that helps both of them grow. Some of that is due to pest management (where one plant species deters pests that raid the other). Some add nutrients (nitrogen fixers and other dynamic accumulators). Some can grow at different canopy layers (such as growing ground cover to help retain moisture while taller plants are used as trellises for vining plants).
Some plants have deep tap roots that break up compacted soil and help draw out nutrients to make them bioavailable. Many of them are pioneer plants and are considered weeds or considered invasive.
Much of big Ag is an effort to shortcut all this. Use a chisel plow to break up compacted soil. Add ammonia. Use earth movers to create more flat land with useful soils. Etc.
The advantages are clear: you can decide when to do these things, and do them on a short schedule. Its hard to argue waiting 5 years for some remediation project to maybe work, when I have a chisel plow right over there in the shed.
Yeah. Big Ag is not incentivized to be sustainable, let alone be regenerative. Big Ag is certainly not resilient. Its very reward system (like modern commerce in general) is based on the extraction of resources (whether that be in the earth, or people resources), and then controlling access to them. Overharvesting always happen, because we're culturally conditioned to define wealth and status in terms of how much one has hoarded more than other people.
I think the main path out of this is to have decentralized food systems, and this is where regenerative and resilient design patterns really shine. I think it starts with a shift in the mindset, where instead of seeing wealth as coming from extraction and hoarding, to seeing wealth in terms of being stewards of regenerative processes.
We have pseudo-regenerative processes in the financial sector with compounding interest, but you can't directly eat money. I can eat the stuff from my backyard.
The beautiful thing about participating in building a decentralized food system is that I don't have to convince you, or anyone else on HN to affect collective action. I can start doing these things in my home, and trade things with other like-minded people. I can offer up the excess abundance to my friends, family, and neighbors. Some of them may want to take up the practices themselves.
Just to be clear, I used shortcuts too. I am using raised beds and importing compost and manure (buying them from big box stores). But I'm also taking the time to plant edible perennials. I only found out later, there are many ways to get even quicker turn around time -- sprouts, fast growing annuals, succession planting. I'm learning. Next year's growing season will incorporate more.
>Yeah. Big Ag is not incentivized to be sustainable, let alone be regenerative.
Can you expand on this point some? I'm trying to understand but I don't really see the problem. Whether the nutrition comes from the soil or the fertilizer doesn't really seem that important. I get what other posters have said about vegetables in the past being more nutritious but it's not clear to me that this matters terribly much compared to the accessibility gains from produce being cheaper and still nutritious (ableit less so).
Ammonia created from fossil fuels accounts for 2% of carbon emmisions, but it seems likely that ammonia created from air and water with renewable energy will replace it in the future. Theres some pilot plants in existance.
It isn’t as if regenerative agriculture is not putting nutrients back in the soil either. They are, usually in the form of compost or animal manure.
However, soil fertility is not just about the chemical makeup of the soil. There are soil bacteria and mycellium that symbiotically live at the roots to help plants make use of those nutrients. Root systems of plants grow together, and plants biochemically communicate with each other. Roots also need oxygen, so aerated soil matters. Plants that die back in the soil leave their roots in place, slowly decomposing and releasing other nutrients for use of other plants, and other organisms such as earthworms. Those same roots — both larger tap roots, and fine branches of the root system, changes the absorptive qualities of the soil so that it can hold water and release them.
In other words, healthy, fertile soil is very much alive, and helps store and regulate water and nutrients for the plants that live and die on the soil.
Healthy, fertile soil can support life directly in a way in a much tighter feedback loop. Using amendments requires more intervening steps. If the top soil gets eroded away, that top soil has to get imported, creating a longer supply chain. The longer the supply chain, the more fragile it becomes. It becomes less adaptive, more sensitive to volatility.
Another is that the optimization toward short term yield is often at the cost of long-term fertility. We build these fragile systems that are optimized for scale and yield as if that fertility will not run out. If the soil is depleted somewhere, importing soil amendments, or even top soil, has to come from somewhere. Even if the supply chain is not cut, at some point, the resources to maintain that longer supply chain will run out.
This setup also conditions people into a situation where they now have to rely on currency to supply basic needs. That, in and of itself, drives the growing wealth inequality of today.
To address that wealth inequality, we are now seriously considering universal basic income. But if you looked at universal basic income as just one type of energy and nutrient flow, it is ridiculous that an easier method for obtaining those basic needs can be had by allowing each household, or neighborhood to supply some of those needs locally.
There is a small town in Canada where the community got together to convert one of their public spaces into a food forest. Volunteers came in to dig and plant. Local businesses stepped up to provide transportation to bring in the compost, the perennials, shrubs, and trees. The food forest was designed in a way so that the natural rainfalls can sustain the growth. The perennials all create a long-term supply of food. Once the initial capital was invested, there is little ongoing maintenance.
People in the community can then go into food forest and get free food. There are probably some education on identifying what is edible (and there are plenty of it). Food grows on trees and on the land. A food forest doesn’t require currency to feed people, and as such, is resilient to a deflationary contraction of the economy.
And that is just one regenerative system. Individuals who have access to multiple, local, regenerative food systems have a greater food security than those that do not.
And that is just food, one out of a handful of the foundational survival needs. There is also shelter, warmth, clothing, water.
Permaculture design looks at the whole system, such that all of the basic needs can be partially, if not wholly, met with regenerative systems. A individual within a community that has access to multiple regenerative processes that can meet all of the basic needs do not require universal basic income. When something like the pandemic happens, they are in a far better place than individuals who do not have access to any regenerative processes.
Such an individual can build the higher needs (I am referring to Maslow’s Hierarchy of Needs) on a foundation of resilient and regenerative processes. With such a solid foundation of abundance, there is less reason to hold others in contempt. Or to cast out and alienate another group out of fear and anxiety over survival needs.
That fertility of the land extends into creativity of the people. The permaculturists I encounter tend to generate many new ideas. There is a liveliness and exuberance that is a contrast to the general malaise and loneliness of many urban and rural residents.
It isn’t as if building this is not for free either. This does require a significant shift in the mindset, in how one views and experiences the world. There is hard work required. Your hands get dirty. But I would say, it is also more fulfilling — not just nourishing the body, but also more intangible needs for people to have meaningful and purposeful lives.
I love "decentralization" as much as the next person on HN, but one the problem is not pricing in the externalities, I am not sure centralization is the problem.
A few bleeding hearts can run a parallel economy for good X with an alternate value function, but this doesn't really scale in and of itself.
Now, if someone combines alt grid / alt farms with lobbying for punitive tax in the smallest surrounding jurisdiction that can enforce it, perhaps that might work.
While they can pump in nitrogen and phosphorus, they can get away with trashing the soil as much as they want. But if phosphorus becomes scarce they'll be suddenly left with ruined soil and no cost-effective work-around.
> we're culturally conditioned to define wealth and status in terms of how much one has hoarded more than other people.
Do you have an alternative to this? This is basically animal behaviour, it's not unique to humans. Most of that hoarding is out of an effort to get a mate, too.
I wrote the alternative: seeing wealth as being stewards of regenerative processes. I'll elaborate.
Plants are generally more cooperative than they are competitive, and groups of plants that cooperate will do better together than by themselves. Trees will establish a network of roots, and shunt nutrients to trees within the grove that are not doing so well. Just because some plant species will engage in chemical warfare with some species doesn't mean it won't cooperate with others, and there are many examples of such cooperation. This idea that nature is based on competitive pressure is not wholly true; there is a whole aspect of cooperative processes that occur in ecology.
Animals (an insects), if you actually observe them, can also be more cooperative than competitive. It depends on how you set it up.
Some examples: chickens are micrograzers, and sheep are grazers. They don't compete for the same niche. Due to the size of their mouth, sheep cannot eat plants past a certain height. Chickens will also eat bugs. That includes eating the bugs off of the sheep, if you let them.
Yeah, my chickens get into my tomatoes sometimes. They scratch at the green mulch I put down in the garden. They will fly over the small fence to get to it.
But a different perspective: I let them out into the main yard, and they like the forage there more than they like it in my garden. They help keep the cricket population down (which helps keep the black widow spider population down. The fertilize my backyard (which has poor soil fertility from the previous owner). Their scratching to get at things help aerates the soil, and bring in more fertility.
So I don't have to compete with my chickens. I can shape the circumstances such that their natural instincts further the overall health of the system.
We can do better than that. You can, for example, plant nitrogen fixers directly as a companion plant for the plants that are heavy feeders. (Example: https://en.wikipedia.org/wiki/Three_Sisters_(agriculture)) (But you'd still have to rotate out tomatoes and other nightshade family plants because of accumulated toxins in the soil).
There's a way to do this kind of a thing with perennials, and better yet, in the backyard or in the neighborhood garden. That is as local and "farm to table" as you can get.
At least in my experience you plant cover crops there and plow them under at the end. My relatives used field peas. And the four-field system is considered a strict upgrade over the three-field system last I heard, basically just 3 productive fields and 1 planted in cover crops.
- Someone figured out a way to grow comfrey (and other nitrogen fixers) in a separate field with low maintenance to supply mulch material for the main crop, without having to rely on animal manure.
- There are people who built their animal barn at the top of a hill and dug swales to transport runoff from manure downward. They plant edibles all along the swale. Once established, low maintenance.
- I saw another practice that involved growing grasses, and using grazers and micrograzers to increase fertility in depleted soil. It’s part of an overal system where fences and chicken trailers are used to move grazing around.
Some of these practices may not be practical at scale, but that is not the point — these practices work well for decentralized food system
Three kinds of evidence point toward declines of some nutrients in fruits and vegetables available in the United States and the United Kingdom: ... 3) recent side-by-side plantings of low- and high-yield cultivars of broccoli and grains found consistently negative correlations between yield and concentrations of minerals and protein, a newly recognized genetic dilution effect.
I think more automation is only going to make the nutrition content problem far worse. Industrial agriculture has optimized the strains largely based on consumer preference - consumers vote with their dollars for fruit that look consistently like their ideal vegetable or fruit (i.e. round, bright red tomatoes). They pay a lot of money for produce out of season so it's far more profitable to ship long distance so they have to look good and have a decent shelf life after a long transport too. We've been running breeding programs to optimize for those variables for a century or more without ever considering their impact.
These evolutionary pressures combined with yield optimization seem to strongly select against nutrition, sugar content, and flavor. For example, my parents only started planting their homestead a year or two ago but already this year the peaches, raspberries, and apples are leagues better than anything I can find in stores or farmers markets and bring me back to the food I remember before I moved to the United States. I can't easily test the nutritional content but I bet it's much better, even with the over-fertilization that they have to do (despite being in California).
Once we start optimizing for automation requirements as well as everything else - to justify the extra capital expense - I'm afraid the quality of the food will get even worse. Every operation, even the "organic" ones, falls prey to loss of quality beyond a certain size and it's largely because of economic pressure that I suspect will only get worse.
I once bought some Californian peaches (I live in NZ). They were very large, good looking, and absolutely tasteless. Watery with only vague hints of peach flavour.
I do not believe the solution is to increase robotics to support this field, I think the solution is much more about a paradigm shift from over-optimization aka mono-cultures to a more holistic approach that increases the human-nature connection. We have to heal ourselves to heal the planet and while technology can help, software engineers will not be the heroes on this front, it will be more traditional stewards of the land.
There is something deeply broken about our human experience and that comes from becoming abstracted away from the ecology. It isn't just that we have to heal the plant; we also have to heal ourselves. (And frankly, if human civilization self-destructs and humans get wiped... the planet will recover just fine, eventually. Outside of maybe nukes going off everywhere).
We have come to think of ourselves as apart from the ecology rather than a part of the ecology. We think that the ecology is a factor in our economy, rather than our economy being a part of the over-all ecology. Software engineers in particular, have this weird aversion to getting their hands dirty with hardware, and so we abstract ourselves away from the hard, dirty bits. It blinds us to simpler, more practical solutions.
I think there is a place for technology, but only in terms of how it supports the direct human experience with the land, rather than getting in between. For example, I think the free and open web is highly vital to helping people share ideas and methods. Every single locale is different, but there are design patterns that may work. Seeing all the forums, subreddits, podcasts and blogs on this subject ... we already have the technology we need to pull all of this off.
And none of this involves VCs, Big Tech, or aggregators (from Stretechery, Aggregation Theory). It will be the land stewards. Those aggregation of wealth are fundamentally based upon the extraction of resources (be they from the earth or the people), and then controlling access to them. It's to the point where we have to depend upon buying into the civilization in order to live and survive, but it is breaking down and now the best we can come up with is the universal basic income. (When right there, in your backyard, you can be a part of something that nourishes you directly!)
The irony is that, the same skill and mindset software engineers use to design an appropriate architecture from a myriad of design patterns is the same kind of mental process one can analyze a site and design resilient, regenerative, and decentralized food systems.
The parent is precisely arguing against mono-cultures. I don't see how your reply is really discussing any of the various relevant points he raised.
Right now, all that I understand from your opinion is that you think robots are bad because of mono-cultures, but the parent would argue equating those two things is incorrect.
I'm definitely not anti-technology as a whole, but I very much think many of the environmental problems we're facing in agriculture were caused by an over reliance on technology in the first place. We abstracted ourselves out of the process too much and it has had horrible consequences.
Biointensive gardening and earth-centric/organic/whatever methods definitely have their place, but it would be a huge shift to grow everything that way, and it's far from clear that it's even workable for the volumes needed for the big staple crops like corn, wheat, and soy.
The idea of bringing companion planting, cover cropping, etc to conventional agriculture by way of increased automation is a fascinating one, in part because of how pragmatic it is. It could allow those methods to scale much more than they do today, and allow a conventional farmer to adopt them incrementally, realizing the savings piecemeal (better yield here, reduced fertilizer use there, etc). This is going to be way more realistic than trying to reboot the whole system from scratch, especially to a method that requires 100x the labour inputs.
And it's definitely closer to the land than the other high tech direction, which is closed loop indoor hydroponic/aquaponic system— essentially doubling down on the monoculture approach by isolating the whole operation even more from unwanted natural inputs (pests, weather, and so on).
Thanks for your great comment, it's helping further the discussion and my own understanding.
> but it would be a huge shift to grow everything that way, and it's far from clear that it's even workable for the volumes needed for the big staple crops like corn, wheat, and soy.
Agree completely. But I think a huge shift is very much needed, and this shift may decrease our dependencies on staples like corn, much of which goes to cattle feed and fuel uses when we have other, more ecological methods like pastured beef, solar etc. https://www.scientificamerican.com/article/time-to-rethink-c...
RE 2nd paragraph, also totally agree with, minus the "100x labor inputs". If done properly, bio-mimicry in Ag should decrease labor inputs as you leverage the advantages of natural plant/soil nutrient cycling, reduced watering and tillage from cover cropping and soil development, reduction of pesticide application by attracting animal/insect diversity etc.
From my understanding, "conventional" farming (I say ironically as we've really only been practicing these methods for ~60-70 years vs millennia of small scale farming that is the real "conventional"methods lol) has forced us to increase labor inputs - adding fertilizers, pesticides, annual tillage, managing irrigation systems - these are all things that can be eliminated, if the proper systems are put in place.
Gabe Browns book, Dirt to Soil was an eye opening read. Cheers.
We don't necessarily have to scale this. I think a better approach is to build decentralized food systems using these practices. At the neighborhood or household level, these practices are very practical. It does not require rebooting the whole system. It does not require venture capital, because you're not trying to make a big return. It does not require waiting for conventional farmers to change their practice.
That process is under way, somewhat— I have several CSAs serving my city, and two major farmers markets. But relevant to note: all of those outfits are focused on fruits and vegetables, where the impact of "fresh" is the greatest. All of the corn being grown around me is sweet corn for immediate consumption, or for animal feed.
No one is growing wheat on a small scale, much less in the kind of quantities that would allow me to have several dozen kg with which to bake through the winter. Same with legumes, as far as I know.
In any case, I'm not trying to impose a solution here. Quite the opposite; the food system is a large problem that requires multiple irons in the fire, and that may well look like reinventing large-scale cropping to be less destructive and chemical-intense, while simultaneously moving fruit and veg production closer to home with smaller scale operations.
Legumes are interesting in that there are many crops that can be companion planted. Can they get to quantities that can sustain a family? I don’t know. But pole beans are known to grow better as a companion plant to corn than by themselves. There are many other types of legumes — the Mesquite tree here in my neck of the woods (lower Sonoran) is a nitrogen fixer, with some varieties producing bean like pods that can be harvested. (They require a mill to get the full use pf them).
My point is that there is a lot more options out there, and what is possible is a lot greater than people think.
As far as wheat goes, I have not heard much from permaculturists doing that. I hear more about cultivating amaranth and quinoa as nutritionally dense pseudo grains.
Generally, I also hear a lot more about winter squashes. That follows the design principle of “Catch and Store Energy”, with squashes being a very durable store of sunlight and nutrients. Speghetti squashes can be made into pasta-like noodles. Not so much with bread.
Cauliflower can be processed to make a kind of flour that can be used in a lot of baking. Again, I don’t know the yields that can sustain a family. Depending on where you are, they are a fall crop. (They are a winter crop for the lower Sonoran, zone 9b/10). Probably grown in succession so that they can be had for a while. Maybe the flour can be stored over winter.
There are also sunchokes. These are in the sunflower family, and are considered invasive. They are very hardy, and their tubers can be harvested during a time when few other plants can be harvested. Their flowers open late season, and also nourish pollinators at a time when they don’t have a whole lot of options. They will grow in poor, depleted soil, and will probably edge out other pioneer plants. So not bread for the winter, but a good source of calories nonetheless.
Maybe that will the challenge my wife presents to me: finding the a good set of crops that would allow her to bake.
There is an element of truth to this. I agree -- people are out of touch with the world around them, on average.
But it's a double-edged sword. The abstraction has been hugely beneficial to society as a whole. Specialization lets us create entirely new industries.
These problems can't be ignored forever though. Eventually the dollar figure behind them will motivate even the most conservative actors. If it's cheaper to use better technology, and that technology doesn't have the same side effects, then there is no reason to continue to harm.
Having been wrangling this over the years, I've come to conclude that these abstractions create more problems than it benefits society.
I think these very abstractions are the cause of class inequalities, and by extension, racial and gender inequalities as well. Many identity issues come from that specialization. We've defined worth to society to how economically productive someone is. There are many things that society as a whole and individuals can benefit from, but society don't value them economically to pay for them. People involved in that (such as artists and musicians) are constantly struggling with the tension of authentic expression as an artist and creating something popular so they can make a living wage.
I won't even go into how these abstractions are a detriment to mental and emotional health, let alone, someone's spiritual life.
I am not advocating for things to go back "to the way they used to be".
Rather, I think there is a way to move forward. I think it will involve decentralizing our food system using resilient, and regenerative practices at the household and neighborhood scale. People will have to become more involved with growing their food, and that is generally a good thing. We have collected a number of design patterns that can yield nutrient-dense foods without as much labor, and reducing transport costs to ... going to your backyard (or going down the street).
Our society is already moving towards more remote work (for some segment of our populous at least), and that will drive a trend towards the hyperlocal. The big missing piece is the decentralized food system.
I have a coworker who grew up on a farm. When asked about the experience he has almost nothing positive to say about the experience. Working on a farm, even in modern times, does sound like a significant step down from the relative comforts of city living.
If advances in technology can decrease the burden on farming then we may see people willingly choose that lifestyle.
"Growing up on a farm" can have vastly different meanings across places - from living next to an industrial-scale operation to an idyllic, quiet small farm surrounded by nature.
Well, that goes to show it. Not every country has gotten rid of small-scale family farms. I used to visit a place just like that in Brazil, and a friend's parents also still live in such a setting in France.
There has been an incredible resurgence of small scale farms, in the range of 40-300 acres, that are a combo of country house and working farm. Some are capable of feeding thousands of community families. Not all food comes from industry.
The environmental problems we're facing in agriculture seems more closely related to our need to improve yield given limited land / proximity with customers.
If you hate on agricultural technology, you're just hating on the effect, not the cause.
Most people are against your so-called "environmental problems". It's just that solutions have to be financially sustainable besides being environmentally sustainable. The original comment you replied to was suggesting how that trade-off can be fixed with technology, just like many other trade-offs before it.
Top soil depletion, nutrient deficiencies in crops, watershed damage from excess fertilizer run-off, insect population decline from overuse of pesticides; these are all real issue. I'm accurate in hating the effect... the cause is ignorance and shortsightedness.
I know what you mean by "Luddite", but historically, they were not anti-tech either. They broke machines to gain leverage and put pressure over employers to give better working conditions, and not because they hated machines.
There are many design patterns in regenerative agriculture that works, but are low-tech or no-tech. It isn't that you need to be anti-tech; it is that we don't need high tech to solve these problems. They are low-tech because you are allowing natural processes to do things on its own, for the most part. Nature, not technology, is the "automation".
I believe we are of similar understanding here. Great to read your responses here. Here's to the wish that we could share a conversation at some point. Trade notes.
Send me an email sometime (talktohosh at gmail). This stuff I am deep diving —- permaculture design —- is still fairly new to me. But I had started to reevaluate the existing tech in my life in the context of the main ethical principles of permaculture (care of earth, care of people, fair share).
BS. It can, and often times is, more profitable than traditional farming. I encourage you to learn about Joel Salatin and the amazing Polyface Farms in Virginia.
http://www.polyfacefarms.com/
By layering different farming enterprises like beef, hay production, pastured pigs, chickens, rabbits, and more, they have a higher profit per acre than many larger farms, and are doing so in a way that benefits that land, rather than degrades it.
Or, I can plant things that repel specific pests, plant things that attract the natural predators of that pest, or plant sunflowers and tobacco as sacrificial plants. I can let the chickens go after the pests. I can recognize that sometimes, the pests are a solution to a different problem, and adjust things accordingly.
All of those solutions are more resilient than microdrones, computer vision, and laser. These high tech solutions require a long global supply chain. They don't make them as robust as as they used to. Further, having multiple, redundant pest management systems means that when one of them fails, they won't all fail.
the whole efficient markets argument requires pressure in both directions. no one is short selling farmers.
just like how if a house is overpriced and real estate investors believe it that doesn't mean there's going to be adequate downward pressure (in expectation) to prevent an unlucky family from overpaying on the house. (there is upward pressure, though, because of house flipping)
and since entry into the agriculture industry is pretty steep, it's not like anyone who has an idea can just build a farm with a cool robust idea that is more optimal than the current solution.
the other difficulty is the extensive time horizons to implement the solution suggested by hosh. farmers can't just drop everything they're doing this season, implement regenerative ag ideas today and reap the benefits come harvest this year.
it's the same issue with organic. transitioning to organic is a time and cost intensive process for farmers (it can take a few years to be fully certified). only a small percentage of farms produce organic even though it commands a higher price.
But most pests are rather microscopic. I doubt we will ever construct silicon/metal hardware better able to operate at the level of, say, spores than the biological systems that are natural to that environment.
But we're already killing them in vast quantities using chemical biocides. Micro Laser Drones (tm) are only killing pests that are already doomed. They just don't seep into the water table or cause cancer.
There is a short explanation on why heterogeneous cropping is not the norm : Harvesting is just extremely difficult. It's not just a software automation problem, it's building the machinery and tractors, training people, gathering feedbacks... Even if you manage to repetitively grow successful combinations of produces, that's only the beginning of the journey.
Farming is an far older industry than software engineering, don't you think humans already tried that in the past 10 000 years ?
It is a hard problem. Historically, people often grew plants in heterogeneous environments because they grew their own crops. Even if it wasn't _all_ of their food, it was still a subset of their food. But it was very labor intensive.
The scale of monocrop farming today with modern tractors is a very new field as well (no pun intended).
With machine learning specifically, I can see a robot arm being able to farm a lot of mixed crops. This is already happening for the most labor-intensive crops like strawberries[0]. Robots are cheap and can work 24/7.
I think it just a matter of time before computers advance enough to be able to perform the basic, repetitive tasks that humans are able to.
It took one of the protagonist ten years of testing to find a machine capable of harvesting special-bread peppers. The breakthrough your are looking for isn't for software engineering only - or even at all -. It's a combination of mechanical engineering, biology and hacking. Venture-led startups often lead to short-lived niche products made to impress silicon valley. It's not sustainable, and it's not the future.
> don't you think humans already tried that in the past 10 000 years ?
Often what you are saying is true. Except in this case we have access to new technologies that have only existed for a small percentage of that 10,000 years. It might be worth re-examining what we think we know about growing food in light of these new technologies. That is, things that would have been ideal but were practically impossible may now be possible.
Software gives the agriculture industry new tools, it helps gather information, it helps build better machinery and can even help train people. Of course software engineers won't be the only ones contributing to solving these problems, but farmers are very resourceful and tend to welcome any new tools to improve their businesses. They've been improving for the last 10,000 years, why stop now?
On a similar note, is there any way to get money to fund amateur crop research? I know you usually need to plant at least 1,000 trees to have a decent shot at creating a new commercially viable cultivar, which I'm not interested in doing at that scale, but I'd plant and monitor 100 as a hobby if I could get someone to pay for that as 1/10th of a research plot.
The thing is, "market forces" (= large corporations) might "solve" it by relying on the use of cheap energy from fossil fuels for doing online vision processing. And then you've converted one sustainability problem into another.
(Which may not be a terrible trade-off, but still.)
That, and the fact that we don't really revisit the modeling approaches when the constraints that led to them no longer apply. For instance, incremental evolution based on a branch where work was done "by hand" into animals, simpler machines and then very complex machines might lead to a very different result than evolution from a branch where the ability to design some types of complex machines is there "from the start".
Note that this is different than saying "let's completely ignore the past".
Monocrops[0] (homogeneous fields of crops like corn or soybeans) are easy to build automated robots for. We've been doing that for generations now, starting with combine harvesters[1].
But there are many problems with growing monocrops[2]. Pests are a problem -- they breed and spread disease. And on the soil nutrient side, you have to keep re-fertilizing (and often over-fertilize[3]) every year to keep it going. And still nutrients continue to decline in the food that Americans are eating[4].
If, as a field (of software engineers), we can figure out a way to build robots that allow growing and harvesting plants in heterogeneous environments, that will be a significant step forward to fixing these problems. I'm optimistic that it will be possible in the next decade to build it and make it profitable. :)
If you don't have to use fertilizer and pests don't eat your crops because you haven't created the ideal breeding ground for pests + disease, then this problem should be solved by market forces alone. At least, I hope so!
0: https://en.wikipedia.org/wiki/Monocropping
1: https://en.wikipedia.org/wiki/Combine_harvester
2: https://en.wikipedia.org/wiki/Monocropping#Monocropping_diff...
3: https://en.wikipedia.org/wiki/Fertilizer#Environmental_effec...
4: https://journals.ashs.org/hortsci/view/journals/hortsci/44/1...