Hacker News new | past | comments | ask | show | jobs | submit login
What vertical farming and ag startups don't understand about agriculture (thinkingagriculture.io)
348 points by kickout 12 days ago | hide | past | favorite | 314 comments

My father is an ag soil chemist of 50+ years.

I'm an industrial systems eng. w/ a specialty in polymer-textile-fiber engineering. (Mostly useless skillsets in the US now)

Gonna share a few lessons here about agriculture that I try to convey to EECS, econ, Neuroscience, and the web developer crowd.

- You can only grow non-calorically dense foods in vertical farms

- It takes 10-14 kwh/1000 gallons of water to desalinate. More if it gets periodically polluted at an increasing rate.

- Large majority Agrarian populations exist because the countries are stuck in a purgatory of <1 MWh/capita annum whereby the country doesn't have scaleable nitrogen and steel manufacturing.

- Sweet potatoes and sweet potatoes are some of the highest satiety lowest input to output ratio produce. High efficiency.

- In civilizations where you are at < 1MWh/capita annum - there is not enough electricity to produce tools for farming, steel for roads, and concrete for building things. The end result is that the optimal decision is to have more children to harvest more calories per an acre.

- Property, bankruptcy, and inheritance law have an immense influence on the farmer population of a country.

I remember telling some "ag tech" VCs my insights and offering to introduce my father who has an immense amount of insight on the topic from having grown things for as long as he has....My thoughts were tossed aside.

> 1MWh/capita annum

Oh this is fascinating! I never thought of this but of course energy consumption per capita is going to be an indicator of how industrialized a country is. I briefly checked the two countries I am a citizen of (Canada, Hungary) and counterchecked with one of the poorest countries I know of (Chad) and the numbers are as expected: 14.6, 4.1, 0.013 (oof).

Now think about what it means to do with less energy.

Energy availability is wealth.

On the topic of energy, I find this correlation fascinating as well: https://jancovici.com/wp-content/uploads/2016/04/petrole_gra.... Oil consumption predicts GDP. Complete article: https://jancovici.com/en/energy-transition/oil/is-the-price-...

I highly recommend Energy and Civilization[1], the most engrossing dry recitation of facts I've ever read.

[1] https://mitpress.mit.edu/books/energy-and-civilization

Yes, and for this reason (IIRC) GDP/capita scales linearly with MWH/capita.

> You can only grow non-calorically dense foods in vertical farms

Purslane (Portulaca oleracea) grows on the sidewalk already, and often next to some wild amaranth (Amarathus Hybridus). What is the point of more efficiently producing specific crops, when there are all these underutilized nutritious plants growing without any human input (or should I say growing despite human input)? This is another problem that I see with the technification of food production in general (including the Green Revolution). Some food wants to be free, but people keep looking for whatever makes the land produce more money in the short term, not what makes it produce more nutrition, etc., because the latter does not adapt so well to the market.

I think this is the type of magical thinking that pervades a lot of these ag startups.

Just because you noticed "Hey, this unpopular grain grows so easily it's popping up from the sidewalk!" does not mean you can actually then go and scale that to providing food for hundreds of millions of people.

We developed modern agriculture the way we did for a reason. Some of those reasons are no longer valid, true, but a whole lot of them are still very very pertinent.

I think vertical farming / rooftop farms / etc are at this point largely an exercise in virtue signaling more than they are actually improving the food system. They might be improving it for high-income Whole Foods types who are upset Amazon owns their favorite store now (ie: us here on HN) - but those people aren't exactly the ones who need to see improvement.

Yes I agree with you, but maybe I did not express my point with enough clarity. I mentioned this two species because they count as food of negligible caloric value (sidewalk amaranth is mostly useful because of the edible leaves). As far as I understand, human caloric needs at present population levels require the large scale conventional methods of production (including the Haber-Bosch process). But caloric needs aside, there are populations, which are not consuming enough greens or fruit, even when it grows in front of them. I have seen it in the (warm and tropical dry) countryside here, they think they have to eat lettuce, when they have lots of hibiscus plants growing in their garden. It turns out the leaves of hibiscus aren't just edible, but their nutritional value is similar lettuce (I suspect higher in Mg since it's a lot greener). The answer of some technologists to this problem is to grow more lettuce. Another plant that grows plentifully here certain months of the year is Ipomoea Triloba. I don't think anyone should try to produce them at scale following the conventional methods. In the case of cities I imagine there could be gardens with a synergistic diversity of plants, which try to keep a high total yield, but not a high yield of any specific plant.

Not the original person, but I agree with you. A decentralized food system does not require as much of the distribution network we have, nor does it require scale.

There are food sources that can be had if they were grown local, and grown in a way that makes sense for that locality. By "local", I mean at the neighborhood, and home level. They can be grown synergistically. Even the old, but simple idea of (thoughtful) companion planting can create better yields than monoculture crops.

Our current food system is optimized towards the kind of scale that allows an easier way to control and meter food resources. VCs chase after multiples on investment, and are not incentivized to truly decentralized food systems.

For others following along, there _are_ already people who have guardens with a synergistic diversity of plants. And not only that, have practices that, over time, create yields that require less time to maintain. The permaculture design community have been exploring and implementing these ideas for over 50 years, and people have gotten them to work. They may not have achieved 100% self-sufficiency, but you would be surprised how much they are able to achieve by synergizing not only plants, but also animals, and human social inclinations into the whole system.

Some people here in Australia do pick dandelion leaves and other wild greens, but the general advice (unsure if backed by evidence) is you probably don't want a lot of it in a larger city because of roadside pollution.

I was happy when parsley sprung up like a weed around my house, but it all died out in the drought. Hopefully it comes back soon.

The human input is providing the moisture trap (the concrete lid on the ground preventing water vaporising into the air), so of course plants will thrive there once the level of lime leeching out is low enough to be tolerable.

But in the end you have several tons of concrete covering tens of square metres of ground, to allow one or two square metres of cereal grains to grow. That is not particularly efficient use of time or resources.

I'm not sure why you think the plant will only grow with a concrete lid on the ground. That plant thrives in the marginal spaces in the sidewalk with poor soil quality, and can probably do better if allowed grow in the lawns where we normally cultivate inedible, ornamental grass.

Furthermore, there are several techniques to help trap moisture that does not require a concrete lid. I'm in the Phnoenix desert, and I use the trimmings from my zucchinni plant as green mulch. Along with the palm fronds that my neighbors love to trim, they create very effective moisture traps. I need to trim my zucchinnis anyways for their optimal health.

I could also plant things in a way to take advantage of the vertical spaces as well. Okra, for example, thrives in the low desert, and provides dense shade that already traps moisture. If I add something that likes shade, and stays closer to the ground, that forms a kind of living mulch that does not require a lot of ongoing maintenance. And if I plant pole beans to take advantage of the sturdy okra stalks, they fix nitrogen and help keep the soil fertile.

I didn't come up with these ideas. They may not be widely known, but people have come up with these design patterns and used them well.

That's right. I've seen purselane thriving along with other weeds away from any concrete, growing taller and with bigger leaves during the first month of rains. A week later someone comes with a machete, and that's the end of it. I do think however that most edible weeds require some sort of disturbance that human presence tends to bring.

same exact boat.

the VC that use to approach us for insights would just never listen. my father literally knew the researchers that tried it in the past and failed.

it did not stop this VC from investing his LP's money in a vertical farm. although i suspect his willingness to allocate other people's money in this manner, for this particular company, had more to do with the social side of things re the founders and other investors.

> Sweet potatoes and sweet potatoes

Is this a typo? Did you mean to write a second vegetable?

Probably regular potatoes; they are incredibly calorie-dense and resilient.

Great comment, thanks.

HN just keeps delivering. It is almost impossible to believe how much embedded technical knowledge is lurking here. You could colonise Mars with it.

> HN just keeps delivering. It is almost impossible to believe how much embedded technical knowledge is lurking here. You could colonise Mars with it.

Of course, in this context you should be prepared for the possibility that if you could tap all that expertise what you'd actually get is "reasons that colonizing Mars actually can't work". ...hopefully not, of course, but beware mixing hopes and dreams and reality.

Something as complex as colonizing Mars would really need to be tried before anyone should claim to have expertise on if it's possible. It looks like our current technology and knowledge is up to the task, so I'd say it's worth seeing if humanity can do it.

Re: <1 MWh/capita annum

Which brings up another issue, which is energy density/m2 of land. To support industrialization/high density urbanism the only fuel sources that do this are currently fossil fuels, or nuclear, but none of the renewable fuels have the energy density.

So if these countries want to increase the amount of MWh/capita, the most efficient (only?) pathway is through high-energy density fuel sources, which right now is being achieved through the use of fossil fuels. To me, this is (one of) the main reasons nuclear energy needs to be prioritized as a climate change solution.

ETA: And, now that I think about it, another way to squeeze more effectiveness from your grid is to build super energy-efficient buildings that reduces the overall and peak grid energy consumption.

Why is energy density important in that context? If all other variables were identical between a high density an a low density solution, the high density one would of course be preferable. But if the low density solution is cheaper and relies less on pre-existing long-distance grid infrastructure, why would high density still be considered the most efficient, or possibly the only viable pathway?

IANAE but the general idea from the discipline of urban geography is that industrialization, and also the knowledge economy relies on the economies of scales that come from high density, mixed-used urbanization. These regions rely on extremely high power that must be supplied from scarce land resources.

Hmm. If the target is 1MWh in a year, that seems trivial to deliver in a place like Chad. Insolation in Chad should be right around 6 or 7 kWh per sq metre. [0] Chop off 85% of that, because we can't harvest all of the solar energy that falls on a square metre of land. We can only reliably capture maybe 15 to 20% of it. But even if you harvest that 1.5 kWh for only 4 hours a day, (which is pretty easy in the middle of the Sahara where you should get about 8-12 hours a day), it comes in around 120 kWh per month. Well above the 1MWh per annum target.

So on the distribution side, they have a population of only 13 million, with a geographic size of 1.2 billion sq metres, which lands us at roughly a thousand times the amount of energy they need to meet the 1MWh per annum per Chad citizen. If they were harvesting it that is. (And that's using 20% efficient solar panels alone. No wind, etc.)

Chad's problem is a lack of anything to trade in exchange for the equipment to harness the sun and build the storage and distribution network. (Grid distribution infrastructure, maybe a liquid air storage facility or 3, solar panels etc) But the sun itself provides them more than enough energy to meet that 1MWh per annum per capita target.

[0] https://en.wikipedia.org/wiki/Solar_irradiance

Solar panels themselves are only 10% - 15% efficient at converting solar to electricity. So I think your estimate here is off by a factor of 10.

If you google “solar panel efficiency”, you will instantly find several sources that all indicate that most (cheap!) solar panels are 15% to 20% efficient, which is 50% higher than the number you’re providing. Top end residential solar panels are over 20% efficient, somewhere in the neighborhood of 22%. Specialized/research solar panels can reach efficiencies of over 40%.

Also, energy density is not nearly as important as you seem to fundamentally believe. It’s pretty much all about cost and availability.

In countries that use very little energy per capita, even just a single solar panel (with a small inverter and a lead acid battery) per person (or even per family) is life changing, giving them enough energy to run a small refrigerator, a light bulb, and a place to recharge their smartphone without having to go into town and pay someone.

In deep urban environments, space is definitely at a premium, but I struggle to imagine that many people are excited to set up a dirty diesel plant in the town square when air quality is already bad enough due to heavily polluting vehicles running in the city center, let alone set up a small nuclear plant, even if wealthier countries would allow such a proliferation of nuclear technology and fuels. Once the nuclear fuel is spent, then you have this extremely toxic, dangerous waste that has to be put somewhere, and those people probably have a lot of other things on their minds, so you could just be rad poisoning their town by giving them nuclear if they don’t dispose of the waste properly.

On the other hand, the rooftops are prime locations for solar panels and solar water heaters, and because 175W/m^2 of solar converted electricity is actually a ton of energy, it’s still plenty of power.

Unfortunately, cost and availability are king. Wealthy countries are happy to sell old assets at a steep discount, and so shiny new solar panels have to compete with second hand fossil fuel plants / generators that don’t lend themselves to great air quality.

Solar is an extremely sensible solution. Nuclear is not a clear answer at all for developing countries, unless you happen to have the design for a clean, portable cold fusion reactor in your pocket. Fission has a lot of problems that are manageable, but managing those problems takes significant money.

Your comment about a solar panel and battery reminds me of my van with solar panel and 100ahr lead acid battery and a 2000W inverter.

Things I've done with it. Run a Engel Cooler. Run a microwave. Boil water. Probably 1-2 gallons a day. Charge batteries for power tools. Run a skill saw. Power lights indefinitely.

With just that you're getting close to a middle class lifestyle.

Grandparent estimated capturing 15%-20%, so they’re only off by a factor of 0x-2x from your estimate range.

I'm not going to provide a source, but my understanding is that one square meter can provide roughly 1 KW of electricity at noon on the equator. Looks like that's about 4MW per acre. Five acres for the "traditional" family farm. I bet where I live that irradiation equals 1MW per acre, and of course it varies seasonally. Sounds ridiculous, until you've tried to water any sizable amount of square footage, and contemplated the heat energy required to evaporate all that water which doesn't go into the plants and isn't runoff.

1 KW sounds high to me. Granted I work in higher latitudes (42). Rule of thumb is 1KW for peak solar radiation, which translates 100W from a solar panel (which are 15% - 20% efficiency).

And the sun in eight light minutes away. The sheer power is mind boggling.

So you can't grow potatoes vertically? Can you elaborate? Is it a function of physiology, i.e. calorie dense vegetables need far more leaves and supporting stems than can be practically stacked vertically?

I imagine space is a factor, but energy will be a big one as well. Calorie dense foods will likely need more space and energy (light) inputs. Vertical farms are very water efficient, so I don't think that matters much.

Vertical farms make a lot more sense with fresh vegetables like leafy greens that grow quickly, command high prices if grown organically, and benefit from being closer to market.

Potatoes are the exact opposite. If it ever becomes more cost effective to grow corn, wheat, and potatoes in virtual farms then outdoor agriculture is dead. While I don't agree with the article that it will never happen, it might require energy advances like fusion power or drastically higher _rural_ land values and water prices.

Greenhouses make sense long before vertical farming, just look at agriculture in the Netherlands, it's mind boggling how much they produce for such a tiny country.

Can you expand on this?

I get that to store a calorie in a potato I need to supply a calorie of energy from somewhere else.

But why is fusion power required instead of better UV lamps in my vertical farm? (Assuming I had enough electricity to run them)

What powers the UV lamps?

Sun + water is cheap and plentiful. Small scale farms can sell potatoes at $0.50/lb or less. Amish farms with oxen can go a little less.

Capital and operational costs for vertical farms don’t seem to make sense, unless there’s some disaster in the Colorado watershed or a trade war that makes hothouse winter produce a viable business again.

This is totally unrelated, but I saw people plowing fields with oxen in Cuba. A scathing indictment of socialism if ever there was one[1]. It's easy to forget at times how large parts of the rest of the world live.

[1] Don't get me wrong, it's a spectrum and some socialism is a very good thing. But not like in Cuba or the old USSR.

>But why is fusion power required instead of better UV lamps in my vertical farm?

Because of the second law of thermodynamics. Your UV lamp is not going to produce light that contains more energy than the electricity you used as input. That energy needs to be produced via solar panels if you want maximum efficiency. If we ignore nuclear or fusion all energy on earth is derived from sunlight.

And geothermal.

I think the suggestion is that fusion power here is a stand-in for electricity that is too cheap to meter -- basically free power.

Yes, that's exactly what I mean, sorry for skipping a step in my logic.

The total amount of electricity to power those UV lamps should be on par with what the Sun sends to the potatoes fields. Maybe that's the reason for fusion. It didn't do the math.

Actually no not really. Plants only absorb two wavelengths of light. It's currently more efficient to convert sun into solar power via panels and then to light LEDs supplying only the wavelengths that plants use. Despite the seeming inefficiency here, the fact is that plants are even more inefficient at absorbing light not at the right wavelengths than solar panels.

Even if artificial lighting and natural lighting were equally efficient you would still have to cover the same number of acres in either solar panels or plants. In other words, the denser your vertical farm the more land it consumes. Clearly vertical farming is meant for some really exotic situations in which you might have access to electricity but are in an environment in which you can't grow the plants you want. That situation would probably be a mars colony or a fallout shelter.

Well, it's more efficient (I forget by which factor). So, if the factor were 2, you would need 1 acre of solar panel for every 2 acre-equivalents of planting space. Since you're okay with going vertical, the acre-equivalent could be much less than an acre. So yeah... electricity being much easier to transport than produce, it would actually make more sense to cover a large part of unpopulated area with solar panels, and then farm in very very small portions of the earth.

From an environmental perspective, this is certainly a 'win'. Reducing the amount of land needed for agriculture is a win to both consumers and the environmentalists and the farmers.

Could one imagine a material that would absorb solar spectrum and emit the preferred frequencies? Something like a polymer one could stretch over fields to get more from the suns rays.

>> "Actually no not really. Plants only absorb two wavelengths of light. It's currently more efficient to convert sun into solar power via panels and then to light LEDs supplying only the wavelengths that plants use. Despite the seeming inefficiency here, the fact is that plants are even more inefficient at absorbing light not at the right wavelengths than solar panels."

> Could one imagine a material that would absorb solar spectrum and emit the preferred frequencies? Something like a polymer one could stretch over fields to get more from the suns rays.

Would you call that a "solar transmitter"?

https://en.wikipedia.org/wiki/Transmitter :

> Generators of radio waves for heating or industrial purposes, such as microwave ovens or diathermy equipment, are not usually called transmitters, even though they often have similar circuits.

Would "absorption spectroscopy" specialists have insight into whether this is possible without solar cells, energy storage, and UV LEDs? https://en.wikipedia.org/wiki/Absorption_spectroscopy

(edit) The thermal energy from sunlight (from the FREE radiation from the nuclear reaction at the center of our solar system) is also useful to and necessary for plants. There's probably a passive heat pipe / solar panel cooling solution that could harvest such heat for colder seasons and climates.

Also, UV-C is useful for sanitizing (UVGI) but not really for plant growth. https://en.wikipedia.org/wiki/Ultraviolet_germicidal_irradia... :

> UVGI can be coupled with a filtration system to sanitize air and water.

Is that necessary or desirable for plants?

https://www.lumigrow.com/learning-center/blogs/the-definitiv... :

> The light that plants predominately use for photosynthesis ranges from 400–700 nm. This range is referred to as Photosynthetically Active Radiation (PAR) and includes red, blue and green wavebands. Photomorphogenesis occurs in a wider range from approximately 260–780 nm and includes UV and far-red radiation.

Photomorphogenesis: https://en.wikipedia.org/wiki/Photomorphogenesis

PAR: Photosynthetically active radiation: https://en.wikipedia.org/wiki/Photosynthetically_active_radi...

Grow light: https://en.wikipedia.org/wiki/Grow_light

Are there bioluminescent e.g. algae which emit PAR and/or UV? Algae can feed off of waste industrial gases.

Bioluminescence > Light production: https://en.wikipedia.org/wiki/Bioluminescence#Light_producti...

Biophoton: https://en.wikipedia.org/wiki/Biophoton

Chemiluminescence: https://en.wikipedia.org/wiki/Chemiluminescence

Electrochemiluminescence: https://en.wikipedia.org/wiki/Electrochemiluminescence

Quantum dot display / "QLED": https://en.wikipedia.org/wiki/Quantum_dot_display

Could be possible? Analyzing the inputs and outputs is useful in natural systems, as well.

Amateur hydroponicist here. It depends on what you mean by 'vertically'. You can certainly grow potatoes in containers that are stacked vertically. You can also grow them hydroponically. However, the issue I've noticed in the hydroponics community is that no one is interested in growing potatoes. That is really the problem with these vertical ag startups and such. They focus on ridiculous foods like greens, which -- while nutritious and easy to grow -- cannot form the bulk of a human diet.

As a community, vertical agriculture need to focus on high calorie crops like potatoes or sweet potatoes or at least something useful like beans.

But circling back to the beginning. You can't really grow a potato with less industrial input vertically than you can with regular land, so unless you are really out of land (and the United States at least is not running out of land anytime soon), it doesn't really make sense to do so. Potatoes are really easy to grow -- you stick them in the ground and dig them up a few months later. Anyone can just buy a few acres of land, fertilize it, stick in some seed potatoes, and get a pretty decent crop that more than covers their costs. This is currently way easier than the amount of setup it would take to use containers. If you were to use conventional growing containers, you would need to import large amounts of soil / substrate. If you were using hydroponics, you'd also have to buy large amounts of hydroponic substrate or expensive nozzles for aeroponics. Either way, it's more expensive.

Anyway... wish me luck, I'm starting some potato growing experiments this summer to see if I can develop new container, vertical, and hydroponic techniques. I'm particularly interested in growing potatos without a substrate and without expensive aeroponics. Currently investigating 'aeroponic' drip systems.

Do you have a blog or a youtube channel where we could follow your endeavor?

No not really unfortunately. I'm not great at that kind of thing.

You probably can grow them (you can "grow" a potato in a cup of water on your counter), but probably not profitably. Potatoes have a fairly low commodity price relative to their light and space demands. Additionally, they store and transport really well.

> The end result is that the optimal decision is to have more children to harvest more calories per an acre.

For a photo-essay about this there's the New Humanitarian article here: https://www.thenewhumanitarian.org/report/94947/lesotho-weat...

It describes the interaction between climate change, HIV/AIDS, and poverty.

(It used to be called "Too poor to farm").

I guess I was always under the impression that vertical and urban farming would be done for "specialty" crops like herbs or kale or something, never for high volume cash crops like potatoes or corn. I can see a benefit for these "specialty" crops because they aren't done to the same scale (maybe I'm wrong about that)

The parents comment certainly is accurate for most staples, cereals, and even meat (raised on corn) but entirely misses the need for more fresh veggies and leafy greens in the US. In the states it generally costs more to buy food made with fresh vegetables. That’s where indoor and vertical ag shines. Many developed countries have plenty of empty calories. Too many. Indoor Ag provides an opportunity to provide scale and freshness to a number of non-grain foods, with less pesticides and preservatives. I hope the field is successful at that goal.

So you plan is to replace calories with mostly empty (calorie and nutrient-wise) greens...

Yeah that was my take as well. I think there's some sort of trade-off point here, though I don't know where it is. Yes modern outdoor agriculture is hyper efficient, although I think the author's comments about cash flow self-sufficiency gloss over a lot of government subsidies and bank bridge loans. In any case, indoor ag should be able to exploit the lower weather/pest risks, lack of need for damaging pesticides, consistent conditions, 365 day growing season, proximity to markets, etc. at least for some combination of products.

Even for specialty products, you’re competing with cheap travel and labor.

Guatemala or Peru are short hops.

What is it about vertical farms that prevents calorically dense foods? Also, I know something like lettuce would not be dense, potatoes probably dense, but is there a cutoff/metric, e.g., calories-per-gram for this determination?

Thanks for the insights!

I'd imagine it's sunlight / artificial light + growth to harvest time. It takes energy input to make food calories, there's really no getting around it with technology. The only way to increase the calories in the harvested food is intense direct light in a short time, or less intense (shaded) light over a longer time. In vertical farms, that's a higher cost of production.

Hi You write: Sweet potatoes and sweet potatoes ?Could this be a typo what's the other one please?

> - You can only grow non-calorically dense foods in vertical farms

How about seeds and sprouts?

I grow sprouts on my kitchen counter. Mostly they require lavage. For the last day or two they require light to turn green and possibly develop flavor; they don't need much.

> How about seeds

You need adult plants to produce seeds, defeating the purpose.

> and sprouts?

Alfalfa sprouts? That's very much a non-calorically dense food.

Just one thought, everybody can buy an imported tractor these days. And steel too.

Around 6 years ago I quit my job as a developer to dive into agriculture. I learned about syntropic agriculture systems and felt in love with it because:

- You are able to work with space and time in a way to maximize yield (not 1 crop yield, but but multi crop) - It focus on being biodiverse - It builds forests

So in this systems you will see rows of trees intercalated with rows of beans, corn, soy anything "weedy" or grasses... Harvest this small plants for many years, after a few years you harvest fruits, and after 2 decades you harvest the wood and start over. All with extensive pruning.

This way you end up with better soil each time without machines or fertilizers (sure you can speed even more the process with them), its a type of agriculture focused on nature's processes instead of inputs.

There's an interesting video about it showing some big farmers here trying to build machines better adapted to this kind of agriculture, this is the biggest bottleneck to scale because right now most machines are very focused on monocultures: https://www.youtube.com/watch?v=gSPNRu4ZPvE

The phrase "It builds forests" is so powerfully, simply descriptive.

I think that framing agriculture's transition (hopefully) away from mono-culture into a more ecosystem focused idea seems like a tractable optimization problem. If we look at the reasons for mono-culture, I would argue part of the reason is that traditionally bigger yield is linked to bigger tools -- tractors are much larger than horses, spraying a chemical is easier when only one thing needs to survive. Monoculture makes it easy to apply big things, harvesting one row of corn is easy to scale to ten rows of corn just by making the combine harvester wider -- the harvester's problem statement is generic and scaleable in this way.

The hard problem, that you raised at the end, is how do we scale harvesting non-mono-cultures. The constraining variables are quite different when we need to perform a set of ten actions with no locality guarantees (Monoculture just guarantees locality of similar actions). I think one natural perspective is to look at how we do things non-locally at scale, which effectively reduces down to a distributed systems problem.

edit: few small changes

It’s pretty amazing what the land can do if you’re clever. I have a friend from college who makes a decent living as a flower farmer who also does horse boarding, etc.

When he slows down for retirement, he has a few million bucks worth of hardwoods that he planted right out of college on land that wasn’t good for other purposes. Mostly black walnut and maple, which he also produces syrup with and may start making booze with!

I grew up in a scenario more similar to this, though my parents didn't have the foresight to plant hardwoods for harvest (Though we did chop firewood to keep the house warm in the winter from our woods). My father still sends a supply of maple syrup each year, which is important because the syrups sold in stores are pretty questionable (The texture is too thick and the sweetness is one-dimensional).

Definitely a tension I've found in life between working in an urban software world and a more bucolic, fostering atmosphere of a farm. I hope more people find a balance in life like your friend, seems they have found the best of a couple worlds.

It is still hard, though, to scale this approach in the way that modern factory farms (Or even small family farms, to be honest - harvesting 400 acres is still non-trivial compared to the average of 150 acres in the 30s [1]) have done with the monoculture.

I would probably assume your friend has a small family style farm, which is what 90% of the farms in the US are [2]. Total farm output has tripled [3], a top of the line tractor costs nearly half a million, it really makes the equation of making an integrated farm a much more complex scenario. If I start doing a more bespoke culture, these tools are probably much less effective - that's the core challenge I think that needs to be solved, how do we increase output of heterogenous cultures.

If I had to bet (And I don't go to Vegas often for a reason), there's likely an inflection point where micro-technologies come to farming in a more direct way, possibly supplanting the way we do a lot of things today. I hate the appeal to nature, but it does seem prescient, in that a bunch of small organisms (Bees, butterflies, and birds) contribute so much to the overall health and harvest of an ecosystem. Maybe there is some, excuse me, cross-pollination to be found between that world and the one we've constructed.


Interesting history of farming I found https://www.thoughtco.com/history-of-american-agriculture-fa....

[1] http://usda.mannlib.cornell.edu/usda/AgCensusImages/1969/02/...

[2] https://www.ers.usda.gov/amber-waves/2017/march/large-family...

[3] https://www.ers.usda.gov/data-products/chart-gallery/gallery...

Keeping the soil productive with more fertilizer costs a lot in some ways- fertilizer production is a big energy consumer. Basically attempting to accelerate using fossil or nuclear what the sun or things that eat organic matter do more slowly.

Hey man no way, you have been living my dream!

First I wanted to just grow berries, then I realized, pesticides and so on, so add another plant to fight that attacker instead of pesticide, then add another plant to protect that plant by being attractive for those other bugs which kill the bad bugs. Then I realized, this would eventually be a forrest with just more berries and edible fruits than normal. Thats where the problem appeared, reaping it would be hard to scale, indeed even planting such a forest would be hard to scale with current mechanical means.

I have a few designs for robot-like planting and pickery, yet all I currently have in realization is 2sqm dirt with potatos, carrots, strawberries and another pot of blueberries. :-/

Then another depressing realization, even if I made this on 100ha of land and produced a lot of nice fruits, berries, roots, the pay-off in money would probably not be worth it.

The good and the bad news is that eventually, when the rest of the land goes infertile from overuse, your method will be worth the money :)

I'm actually harvesting water through condensation machines and storing them in a man-made underwater resevoir now.

Soon I'll be pumping it into the house for plumbing, etc.

My reasoning is, in 20 years I'll have a cool project to talk about, or I'll be sitting on a fortune in freshwater.

You don't do sustainable farming for the money. It can certainly pay the bills and some. Plus subsidies and incentives can help.

If you want to make money you start a vertical farming AgTech, which will not be profitable, but will attract the trendy funding. Pay yourself well while it last.

Smart Farming is the way! You need to concentrate in flavor and quality. Yield is a race to the bottom.

Why wouldn't the pay-off in money not be worth it? You can charge a "responsibility premium" to local hipster stores.

Hipsters and eco-paying-premium people dont have money, not even close to the money provided to vertical ag tech farms.

Capital intensive to develop the automatic reaping bots, which will also cost more to operate compared to current "pour diesel into tractor and pull the earth up and shake it to extract so many potatoes". Destruction just costs less.

Hello, also a developer who's interested in agroecology. I actually also left development (as a job, not as something I do) in order to pursue a more human-centric approach to agriculture. With automation, it seems obvious we'll be seeing way more unemployment than what's happening right now, which is already alarming; small farms with synergistic crops & forestation seems like a no-brainer to achieve food sustainability. Plus, chemical pesticides are usually not used in syntropic systems, which makes it good for your health too.

I dropped out of Agroecology course in 2018 but I actively work with it or did before the pandemic at least.

I am skeptical that "more farm workers" is a trend that anyone really wants. Maybe at small scale you can sell produce at vastly higher prices to make up for the higher costs, but I don't think that what you're suggesting would be good for agriculture if adopted broadly.

Like others said, from a pov of global economics and current geopolitics, it might not make sense. But when you factor in sustainability, independence from the system and health, things begin to make more sense. Mono-cultures degrade the soil, up to a point when it'll no longer be able to sprout that culture anymore, so what do these millionaire farmers do? They just log more and more of our forests in order to plant. That's where all this logging in the amazon rainforest comes from.

All of this happens due to the green revolution & mass automation. We have papers plus empirical evidence you can turn any used up soil into good farming soil, if only we mimic the way nature does it, creating micro-climates with different cultures next to each other. One of the good outcomes of this method is that you don't even need chemical pesticides, because policultures are inherently more resistant to plagues. That and with this method, we attempt to use natural predators to cope with them too. It's basically a method of rebuilding forests, which is why it's called an agroforestry system

How much of this not "good for agriculture" is a result of a mispricing that doesn't factor in the unsustainability of the current mainstream approach? Like many areas this may involve more human workers before later transitioning to smarter machines in the long run.

Sure, that's fine. I was too unclear, I don't think jobs should be a reason to intentionally make farming less automated and that if fully manual or mostly manual farming somehow became the dominant approach it would simply not scale.

I am aware that family farms are more productive per acre and more sustainable usually, but there just aren't that many farmers or people who want to be farmers as a percentage of the population... it's hard work and exactly the kind of labor I'd expect to see automated right back away again ASAP.

Helping farmers with new automation tools that enable sustainable farming seems like a far better option than trying to disrupt farming in a way that intentionally increases the labor required to feed people. If the goal is to help people get back in touch with nature that's a great goal. It's just not a goal I think could be widely adopted.

Farmers are very smart, as the article mentions. If you give them the tools they need, they will use them if they make sense. Heck, farmers are pushing hard for the right to repair and modify their equipment (i.e. http://repair.org/agriculture/)

Edit: In case this is still unclear (it's hard to phrase right), I'm trying to make the point that you're better off trying to create a win-win with existing farmers rather than trying to start from scratch. If they are given better tools they will generally prefer to make their farms and soil healthier because it improves their bottom line. I don't think it makes sense to flip it around and completely change the agriculture system twice.

Might not be good for agriculture economy, but more farmers means more people with the means to feed themselves. Sounds like something I want.

Trade is fairly efficient at that too.

> With automation, it seems obvious we'll be seeing way more unemployment than what's happening right now

I fully support the underlying message, but automation has been happening at large scale for 70+ years now, unemployment rate doesn't follow automation, jobs are just shifted to other industries/sectors.

In other words, "Automation hasn't increased unemployment in the past, even though some pesky scientists and economists said that it would eventually be a problem. Some of those people were wrong in the past, therefore automation will never increase unemployment, ever."

This topic is boring me to death already. I'll keep it short. Automation+competition mean cheaper potatoes (or anything really). Cheaper potatoes means you save $10 per month. Now you can spend $10 on something else. Maybe on a movie or you save the money and go to a theme park with your kid. This represents a new employment opportunity (less farming more entertainment). Therefore automation isn't causing structural long term unemployment.

However there is a darker side to automation. What if automation is used without any competition? e.g. your company has a first mover advantage and it takes 4 years for the competition to catch up. What happens is that the potatoes stay at the same price but the company is increasing its profit margin which benefits the owners/shareholders of the company at the expense of workers. This isn't about unemployment. This is about wealth inequality. When a company replaces a worker with a machine it becomes more profitable but the worker gets nothing.

Society needs to change in a way that the laid off workers benefit from automation to the point that people are hoping their job gets automated or they decide to automate their own job. If someone gets laid off by automation for the third time that person should be happy, not sad.

Contrast this with J.S. Mill.

"Hitherto it is questionable if all the mechanical inventions yet made have lightened the day's toil of any human being. They have enabled a greater population to live the same life of drudgery and imprisonment, and an increased number of manufacturers and others to make fortunes. They have increased the comforts of the middle classes. But they have not yet begun to effect those great changes in human destiny, which it is in their nature and in their futurity to accomplish. Only when, in addition to just institutions, the increase of mankind shall be under the deliberate guidance of judicious foresight, can the conquests made from the powers of nature by the intellect and energy of scientific discoverers, become the common property of the species, and the means of improving and elevating the universal lot. "

Yes; if people claim that X will cause Y, and over decades of doing X it continues to not cause Y, I'd like some very compelling reasons to suddenly believe that X is going to start causing Y.

Wow, that's so cool. I have long been interested in permaculture, which this seems quite similar to — how would you describe the difference? Answering my own question I'd say that immediately the focus on automated harvest of non-monocrop is very important, ad the main arguments against permaculture that I've come across (here and on e.g. Reddit) are that it's not scalable with automation. Thanks so much for sharing

They are very similar actually, but permaculture is about more than just agriculture, agriculture is one of the sides of permaculture. For me syntropic agriculture is that side, some people also call it agroforestry but this term is used for other kinds of agriculture, which builds forests but differently. On syntropic the main difference is very high density of plants and extensively pruning. The video I posted in the first comment you see a few people doing research on automating this processes, there's also some people Swiss investing into this, sure with less biodiverse but its being working great for them, so yes, can be automated, also lots of machines used on fruits crops can be used on this system, specially to speed up pruning bigger trees. And usually on syntropic its not common to find "key" shapes beds and stuff that we see from permaculture, its usually straight rows, which helps a lot with automation I guess.

Introducing an idea, in case you haven't encountered it elsewhere already. There are some estate-farms in England that are arriving on the same conclusion of permaculture/syntropic ag, albeit from a different angle.

They were spending a lot of money to extract marginal agricultural products from soil that wasn't well suited to monoculture, and some eccentric estate managers decided to stop spending the money, and allow the estates to "re-wild": no more shrub pruning, earth-moving, etc.

One of the "big ideas" they had, which might be useful to you, is to reintroduce "mega-fauna" to their ecosystems (in their case, ponies, "wild-ish" bovines, pigs, goats, and deer. They found that these fauna did an excellent job of pruning the wilds, but they had a new, second-order problem in pruning the fauna; they'd like to reintroduce wolves, to do the culling for them, but can't for somewhat obvious NIMBY reasons. :)

They're at the point where the estates more or less run themselves; they mostly make income from selling flowers, culled remains, and ecotourism. Anyways, all this to say that you might consider leveraging some organic automatons to do some of the "extensive pruning" for you. Herds of goats in particular are very efficient pruners, and they can pay for themselves.

Sources (great reading/listening): https://www.newyorker.com/magazine/2020/02/17/can-farming-ma... and https://www.econtalk.org/isabella-tree-on-wilding/

That's very interesting and useful, thank you for sharing

How did you support yourself economically when setting this up and ongoing? Was it economically viable and sustainable?

I had a small reserve, and I cut my living costs a lot. I wasn't trying to make money from agriculture in the beginning, was all about learning, I volunteered a lot and did a few courses later. This year we started actually selling produces and I get lots of calls to pruning jobs which I do decline because 2 years down this line I started working with development again because I got out of money. We would be able to live from the land today for sure, but also having economic security and being able to invest in better tools and such is also very good.

I'm now looking to merge this two worlds and work as a developer on solutions for agriculture/forests. I have a product in mind which I'm currently working on, lets see :)

I think you've just explained why farmers become coders and just buy their food from someone else. The idea of the Gentleman Farmer has been around a long time, but it's hard work and the pay is terrible.

After watching that video, it seems like you could just mulch large areas of "dry land" and it would have a similar effect more quickly. The pruning (and rotting of the wood) is what is fixing the soil right?

Yes exactly! Its what happens naturally, trees dies, falls, takes others with them with the fall, make space for newer trees and wood decompose... Natural succession.

If you don't have woody material, just leafs works too, the key is organic matter build up and photosynthesis. So we tend to cut weeds (when they start to mature/flower usually) very cleanly for them to grow bigger and better, not killing them, focus is to build soil for more demanding plants.

I am thinking of a similar route, as a data scientist i am eager to know what graduate level courses would you recommend ? Especially for agriculture in cold climate (Canada )

Hi - coming from a PhD in agriculture (focus on sustainable ag), graduate level courses are going to be tough to jump into unless you have a strong background in ecology. Most of my grad-level courses assumed years of training in e.g. genetics, soil science and chemistry, plant physiology, ecology, weed science, entomology, etc. Agriculture is a very broad life science field.

That said, if you want a quick primer, "Crop ecology: productivity and management in agricultural systems" is a good primer on most of the basic ecological systems in agriculture. I've read it cover to cover many times.

However, you don't need a grad-level education to farm (believe me, I have been reminded this endlessly) - this is more for people doing research. For applied/actionable specifics for cold climates, your best friend is going to be local crop-extension services (in the US, most land-grants run an extension service). They will have tested techniques for your area and will be able to point you to good resources for farmers, not people researching agriculture.

One last thing - to be a successful farmer has very little to do with growing crops. Take business classes - the rest is relatively easy to figure out.

I don't know what to recommend you. I know a few people doing syntropic agriculture in Portugal which is as close as a close climate that I know. There is a guy in Florida, he have a company called GreenDreamsFL, hes the only one I know in the US doing this. But sadly this is not very much taught in academic courses down here in Brazil, but anything related to agroecology is very close, also understanding deeply plants biology helps A LOT when working with this systems, so we see a lot of people from Biology with a focus on Botany and Plant's physiology, and "florest engineering" I couldn't find a good translation to it, but its an academic course found here in Brazil which also helps a lot on understanding forests processes.

It's not a graduate course in the traditional sense, but Paul Wheaton runs a number of hands-on permaculture classes and courses on his land in Montana. Maybe not Canada-cold, but there's a large focus on shaping land and designing buildings to use energy more efficiently.


If you want to learn about agriculture, find a farm you can support close to you, and enroll in a summer program. You will learn more if you get your hands dirty.

Ok, this is very cool, but it looks like as of now it requires orders of magnitude more manual labor than existing agriculture. This means increase of produce price to about the same extent. Unless a huge paradigm shift will happen (which takes decades without a major disaster), I don't see this as feasible any time soon.

I have so many questions. Do the rows run north-south or east-west? Is there a formula for how widely spaced the rows need to be? Is the pruned wood buried or left on the surface?

Looking forward to the product you mentioned in another post too! :)

what advice would you offer to another developer that longs to get into agriculture?

If you are wanting to do it as a commercial venture, then livestock (particularly beef if you are in the US) is about the only way to go unless you can purchase vast tracts of land and the equipment to run it.

If you are considering vegetable farming commercially, don't unless it is an extremely boutique product like truffles or exotic mushrooms, the economies of scales are crushing. The other option that is still viable is small plot that produces and end product. e.g you own a vineyard but you are not selling grapes you are selling wine. You own a pepper farm but your end product is hot sauce. Those are still viable for small plot.

The best thing you can do with a decent tract of land is to plant it full of expensive hardwoods such as black walnut and occasionally prune the trees to promote straight growth for lumber.

I have 7 acres and I planted 4 of it with African Ebony, one of the most expensive woods in the world. They are not native to my area so there is no issue with harvesting them and they require little in the way of care. They will provide a nice cushion for my children when they mature given that a single tree is worth between $300,000 to $1,000,000 (at current market) depending on size and quality of lumber. I planted about 50 trees per acre. The math is pretty self evident and it is the best use of land agriculturally if you are looking to maximize profit via small plot agriculture.

My wife uses some of the other land for personal farming but that is her gig, I grew up on a farm (citrus) and after NAFTA swore I would never scratch a living out of dirt again. I told her she was on her own with the vegetable farming other than helping her with where to plot certain vegetables and when to plant them.

Aren't there stories where 30 years ago lots of people had the same idea and planted similar species, resulting in a price crash when the trees matured? I'm not a tree expert so don't know the species, but I feel like I heard something like that happening in the southeast.

How did you decide on your species and how do you know other people don't have the same thing in the ground right now?

Yes there where but that was mainly from timber stock which matures faster than true hardwood stock, if you are very young say in your 20's you could plant hardwood stock for retirement but generally hardwood stock foresting is a generational investment.

Some of the limitations on everyone planting is that most of these species are protected, so you have to be able to plant them in a similar environment where they are not native or you run the risk of having to pay impacts for every tree harvested. Others are land availability and the other is many people don't want to encumber their land for a return they personally will never see. Most of the stuff you hear about from 30 years ago where faster growing trees in the pine and oak families which you could see harvestable maturity in 10 (pine) to 20 (oak) years and while it did cause a price crash, those people did make money. Just not FU money.

Contrast this with any African blackwood and you are looking at 50 years minimum till maturity and possible as long as 100 year. I don't need the money (not that I am rolling in it) but it is generational insurance for my children and their grandchildren. For a little back story I own a house that sits on 7 acres on the ocean, I plan to will the house to my descendants and keep it in the family as a place to come back to and congregate, for all generations to use. The trees are the hedge that their will be money to support that vision, as well as provide for the family if need be.

That being said the whole thing could flop, but at least I planted some trees that are in serious danger of going extinct in their native habitat and my descendants will be in possession of some really resilient hardwood.

What changed after NAFTA, if you don't mind my asking?

I mentioned this in another post on another topic but there was immigration reform in the 70's and 80's that opened up migrant work to the conglomerate farms. Which drove down prices most family farms where able to survive this onslaught but it kicked the legs out of any cushion they had. They then lobbied for NAFTA which allowed them to buy up tracks of farmland in Mexico that could not survive due to the new Mexican labor shortage created by the US workforce immigration reform, they then moved production to Mexico, drove prices down to an unsustainable level for small plot farmers, those farmers bankrupted, the conglomerates came in and bought up the small tracts that where now available, they then parceled those tracts together. Then they lobbied for more immigration reform and brought in workforce for their new US farms and that is why you do not see an American field worker nowadays. It's not because they do the jobs that we don't want. It's because they systematically destroyed the opportunity to do so and hold the cards to keep wages suppressed. If wages go up for farm hand work in the US, they shift scale to Mexico, if Mexico is unstable they shift scale to the US.

Most of the politicians on both sides of the isle where and are complicit in it because they view food pricing as a national security issue. The government has a vested interest in keeping the price of food and necessities low as people tend to become pretty violent when they are starving. That being said, it was a huge transfer of middle class wealth to large conglomerations.

I assume the little "Product of Mexico" sticker on all my supermarket bell peppers and that they don't have a $14 minimum wage down there.

Buy a copy of Farming Simulator.

Agriculture is a brutal, pitiless world of perfect competition, commoditisation, and winner-takes-all consolidation. There’s an old farming joke: “What would you do if you won the lottery? I’d farm until it was all gone”.

farm sim is the most broken game ever.

Even with the more realistic mods, which brings the most basic things like seasons(!!!), it is a futile fight against the bugs and bad UI.

farm sim is nothing but an advert for tractor brands.

Watch a couple youtube videos but NEVER pay it. you've been warned :)

Start by getting your hands dirty. Grow some herbs in a window box or something simple. Once you reap the rewards, you may get the green-thumb itch and keep going. Getting started is easy: seed, dirt, water, sunshine

> Start by getting your hands dirty.

Same advice I'd give someone in agriculture looking to get into code.

decide what kind of agriculture you want to do and check what is time and money requirements and seasonality is. Next step could be doing internship to see what it it feels like. There are many options from wwoofing to more job like situations.

Well and then you are ready to decide. Being small farmer is tough: not a lot of money and a lot of work, but it is rewarding by many means.

I personally decided to be in more play farm: few acres of vineyards, small wine production. It is still professional operation but I don't expect to be making full living off it.

Just do it, start getting your hands dirty as other said. I personally started with composting and now I have a system where my food waste becomes forests, I eat lots of vegetable/fruits and I just throw the bucket on a specific place, cover with mulch and food grows. Avocado, papayas, limes, cucumber, tomatos, lots of them grow easily here just by doing this.

If you look for "agroforest academy" in youtube you may find a video course in english on this syntropic agriculture topic too.

I think that's the first time I've actually seen farmers have an interest in improving their soil. Conventional practices are basically strip mining fertile land as if it was some finite resource.

I would like to transition to this in the next 5-10 years. Starting off with a garden on a plot of land I just purchased.

That syntropic agriculture video was powerful.

Any way to contact you to understand how you've made the transition?

For sure, will be a pleasure, you can find me at alissonpatricio at gmail

Thanks! I sent you an email. It's from my personal domain, so might have gone to your spam.

Nice to hear it can be done .... after 20 years in technology I am on the same path to get into the agriculture sector.

I can see that being critical in Brazil, near the rainforests, but does this strategy work in the great plains?

Might not count as the great plains, but Mark Sheppard's New Forest Farm in Wisconsin is a good demonstration of a similar approach in a different context.


Not to belittle the difficulty, but as someone who just drove from WY to WI over the last couple days... east of the Mississippi is drastically wetter than western MN+.

Yeh, there's a chicken and egg situation…

Having trees increases rainfall but need to grow the trees with little rainfall first!

"The willow submits to the wind and prospers until one day it is many willows - a wall against the wind."

Yes, the method applies anywhere in the planet. But for that you need to deeply understand the plants available for you, I mean those that are able to grow there in the beginning, native or not, here we use lots of african grasses and eucalyptus to start. There are a few people replicating this all over the world in very different environments.

your description reminds me of Jane Jacobs in Death & Life of Great American Cities

The author here has a good premise, although glosses over many things. Yes, "vertical farming" is over-hyped. That said, the author didn't mention weather or pesticides/fertilizers at all. Statements such as "Current agriculture doesn’t need an artificial energy source" are plain wrong. Producing fertilizers takes quite a bit of artificial energy and the bulk of the corn and soybean farmers the author is pointing to are the ones heavily using them. And to completely ignore weather and climate is to ignore the single most important variable factor in farming.

It's also a very US-centric view. There is a ton of innovation happening in other world markets, especially with smallholder farmers. Especially around financing.

The author completely ignores financing (even saying there is no VC money in agriculture which is false), which with larger farmers is actually one of the biggest issues for farmers today. Given that farm equipment is getting bigger and more costlier, a lot of thought goes into financing that equipment. Insurance is also a huge deal, and there's certainly a lot of room for streamlining the process of insuring crops and obtaining payouts.

Also, no mention of drought and other extreme weather events. Additionally, no talk about how the Ogallala Aquifer (and others) is being decimated and continued trajectories would be catastrophic in just a few decades. The higher yield and minimal water and and getting rid of pesticides/fertilizers and removing transportation pollution are interesting things to research and see if we can do better. It is already a high bar of productivity but so where horses compared to walking but they were replaced with something better.

I'm very familiar with the Ogallalla aquifer. When the drought of 2012 hit, people were very worried about it never recovering. After several 'wet' years it appears to be fine. Mother Nature is stronger than we give her credit for

I'm not suggesting it won't ebb and flow. The issue with aquifers is that as parts go empty and stay empty for long periods of time the sand that stores the water compresses and then can never be re-expanded to be refilled. There is also the worry about what will happen if we have several "dry" years. All this to say, we don't have unlimited fresh water.

I would like to build on top of your point:

- Indoor farming would not have to worry about things like drought. As a water feeding system can be led all the way to the ocean and the salt removed using pure sunlight as power.

- Indoor farming has shown to yield crops with 96% less water in many cases, again solving the problem mentioned previously.

- Many areas don't have ready access to tons of water so these water conservation techniques will be absolutely necessary.

- The lack of need for pesticides and weed killers and other poisons will also have major advantages.

- The indoor operation can be significantly less emitting in terms of greenhouse gasses. Without the need for large gas powered machines for harvesting, these crops can be way more efficient.

- The indoor operations can be built vertically thus allowing cities to feed themselves without having to ship food across the globe, further providing exhaust benefits.

>Indoor farming would not have to worry about things like drought. As a water feeding system can be led all the way to the ocean and the salt removed using pure sunlight as power.

Are you aware of how much water it takes to produce the output of the Midwest or Central Valley? We'd be talking about the largest desalination project in human history by orders of magnitude.

As of 2013, Israel had a desalination capacity of 500 million cubic meters per year.


As of 2015, the US used ~450 million (edit: fixed from billion) cubic meters PER DAY for irrigation.


Obviously not exactly a fair comparison for numerous reasons, but it gives a sense of the scale we're talking about here.

> As of 2015, the US used ~450 billion cubic meters PER DAY for irrigation

I think you mean to say 450 billion liters, which would be 450 million cubic meters.

Your source says this: "For 2015, total irrigation withdrawals were 118,000 Mgal/d"

So roughly speaking Israel desalinates in a year how much the US uses for irrigation in one day. That doesn't sound so outrageous. Israel is a small country.

Ah you're right, I read the wrong line in the wolfram alpha output. That does make it seem less crazy. It would still be a crazy project but not "entire GDP of the US" crazy.

Israel coastline: 171 miles

US coastline (not including Alaska): 5,800 miles

So the US has 33 times the coastline. It sounds like we're only one order of magnitude off to meet those desal needs.

ps. If we include Alaska, which is another 6,600 miles of coastline, we could again halve that need.

Coastline distance is irrelevant, you could use a single mile of coastline to extract this much water. The issue is infrastructure and energy costs. Traditional irrigation is about 3 orders of magnitude cheaper than desalination.

All those rivers dumping into the ocean demonstrate how rarely it’s needed. Long term pumping water from the eastern US to the Midwest is vastly cheaper than the kind of massive and effectively pointless desalination effort required.

If your nos are correct and if desalination was done at the scale of Israel in US, it would mean it would be able to provide 10% of the irrigation needs via desalination.

This is off course not a big deal in USA and desalination and economics of certain agri crops if reassigned can lead to better outcomes. However i am sure India can do with the level of desalination Israel has (scaled up to its population size) as can other middle Eastern countries. If not today, then may be a decade or 2 in the future. This can enable habitation in many areas in land and water scarce countries.

You know if we are already going through the effort of transporting & stacking dirt vertically for these things we might as well go all the way and integrate them into the pylons of offshore wind turbines, which'll guarantee them a viable support structure, an infinite amount of available seawater & more than enough power to desalinate it locally. Heck, the harvest logistics means it'll give fishing boats something to do off-season too.

So with indoor farming let's say you'd only need 5-10% of that, right? That suddenly makes it basically the same magnitude.

We'd only need 5-10% of the water?

That might be possible, I don't know enough about it to say for sure. I could see us needing somewhat less just due to evaporation differences, but I'd think plants need basically the same amount of water to do what they do regardless. How do we get down to 5-10%?

I don't know how, but a different post claimed that indoor farming needs 97% less water. I imagine it has something to do with using a closed loop. (And the remaining percents are obviously due to imperfections, and the actual water content lost due to the removal of the produce from the loop.)

Most of these sound reasonable, but I've never bought into the "grow vertically" idea. It seems to ignore physics.

Sunlight is delivered as electromagnetic power (watts) proportional to surface area. Plants naturally grow on the surface of the earth, and therefore receive a small proportion of that power which they use to convert CO2 into sugars and eventually plant mass which we eat. Stacking a bunch of plants on top of each other cannot change that the lower plants must receive less power, and therefore cannot grow as much. And that's ignoring the added complexity and logistics (read: overhead) of maintaining a system that stacks plants on top of each other, which would surely obliterate whatever 2-digit% efficiency bonus you can eke out of stacking. The universe doesn't work like Minecraft.

Chemical and water use reduction seem to be a pretty good outcome, as well as being able to ignore seasonality.

I would like to see some numbers on farm equipment (in?)efficiency before throwing that out as a fact. Color me skeptical but it doesn't seem obvious at all that rebuilding a 10000-acre greenhouse every 20 years will necessarily produce less greenhouse emissions than running a few tractors. Or even that harvesting food in a greenhouse takes less energy than doing it with a tractor.

Every once in a while you see some high school science fair project where a kid has the brilliant idea of making 3D solar cells... maybe little pyramids or ridges instead if a flat plane, to capture light from all angles.

The kid gets patted on the head. Those who know better, immediately recognize there is no great increase in power obtained as the 'shadows' caused by the raises structure invariably decrease the efficiency down to that of a plane.

Anyway, vertical farming reminds me of this. You would defintyl need artifical lights.

Akyway, it's amusing watching amateur would-be tomatoe growers get excited about a technology that has been around as long as Cheech and Chong.

Using these systems for decorative purposes, on the other hand, is a cool idea. It's a fast and cheap way to make an 'instant' hedge. I have a 'wall' of pole beans planted in this manner which thrive and create a solid mass of greenery within a month of planting.

It's been a few years since I've followed vertical farming, but I have recalled an argument being made that an artificial light source can be made efficiently by using a single-wavelength, super-efficient LED with a specific color that stimulates photosynthesis.

On the other hand, I've also read (old, long-lost) sources that state that the energy cost per loaf of bread is about $10 for indoor farming, vs $5 for outdoor farming.

These specialized lights won't save vertical farming today, but I will keep following the progress. If nothing else brings value to vertical farming, the fact remains that local food independence is valuable; growing food in a dense apartment or a dense city will pay dividends in the event of large-scale famine or civil unrest.

Sure you can do that but now you have a much harder problem: Invent a novel energy source.

“Stacking a bunch of plants on top of each other cannot change that the lower plants must receive less power, and therefore cannot grow as much“

That assumes all light comes from straight up. That isn’t even true if the sun is straight overhead, and definitely not true close to the poles.

I don’t know whether it’s profitable, but I would think the economics of vertical farming on Iceland (sun lower in the sky, greenhouse heating cheap, imports expensive) are different from those in Equatorial Guinea.

No, it assumes that power is delivered based on surface area with respect to the sun. Vertical/3d farming can't work more than ~2 plants deep, where "depth" is measured as the number of plants between a given plant and the sun. Sure, build it vertically on the north pole, but it's still gonna be essentially "flat". You can't magically get power deep into a 3D farm when there 20 other plants on every side that would get the light first.

I have never heard anyone suggesting that vertical farms enabled 3d planting. It simply removes the linear relationship between square foot of land and number of plants.

Indeed, if the market is remote and energy rich (e.g. remote Canada with lot of surplus hydropower but lack of sufficient sun) the economics changes

> Stacking a bunch of plants on top of each other cannot change that the lower plants must receive less power, and therefore cannot grow as much.

I thought the common idea (and implementation) of indoor vertical farming used artificial lighting at each level. Possibly only using light in the wavelengths actually used by the plant, not "wasting" power at other wavelengths like the sun does.

Also, indoor and especially vertical farming can save precious land. Maybe the US has enough land for farming, other countries certainly do not. Rain forests burned to make space for soy or palm oil are proof of that.

Rain forests are being burned down because these areas are poor and farming is the simplest way to make money starting from scratch. Multi-million dollar vertical farms don't help.

Forget even about money, sometimes if you cannot participate in the 'official' economy and have no prospects, you need bare minimum capacity to cultivate land and feed yourself and your family. This is what happens in developing countries.

Save land? Most land in the US is doing NOTHING, or has little potential for anything other than carbon sequesteration

What is the calory output of vertical farming?

The only things I ever see grown in those vertical farms are low calory, short-shelf-life leafy greens, and the occasional bland tomato.

Greens are nice and all, but calories are what keep us alive. Until they can produce calories, I will continue seeing them as a pointless distraction.

Calory-dense foods are more economic to ship long distances though. From a 'power plant to plate' energy conversion viewpoint, leafy greens that don't ship well are probably the best things to grow in urban farms, assuming people will eat them anyway.

I had the impression indoor farming is heavily hoping to pivot to marijuana.

That's easier to secure vs. a field of marijuana so there might be other reasons besides growing.

Outside of North Korea, humanity in 2020 has no shortage of calories.

I believe that in 2020, we may have a shortage of calories if we were growing all of them in a sustainable manner.

I will also note that in 2020, we have no shortage of leafy greens.

I think this all hinges on your definition of "sustainable".

If you count using fossil fuels to make fertilizer as unsustainable, you're probably right. But that is very debatable.

Sure, we can quibble about all day.

But consider my second point. If there is no caloric crop sustainability crisis in our future, then there is definitely no leafy green sustainability crisis looming on the horizon. The problems that vertical farming of kale solves are... Not very impactful ones. If all the kale in the world disappeared tomorrow, most of us wouldn't even notice.

I think we can agree to be unconvinced about vertical farming.

That land is quickly abandoned. Rain forests are burnt because once you drain the soil of nutrients (takes less than 3 years) you can just move on to a different plot. Conventional farming reuses soil by adding fertilizer to artificially replace the missing nutrients.

If you look at satellite images of Indonesia, you'll see that much of it is now just palm oil. Most of that area was burned at least a decade ago. It would be nice if they'd abandon it, presumably a new rain forest could grow over the next century or so. But they continue to grow monocultures.

> water feeding system can be led all the way to the ocean

If you're thinking pipes, the water might become toxic after 500km or so.

Efficiency and land use is indeed an issue in India as with the struggles to feed an increasing population with a higher protein intake requirement as well.

Instead of natural gas -> Fertilizers route, a solar or renewable energy -> LED route can help for certain crops provided they do grow efficiently.

The article also doesn't mention all the diesel fuel used by tractors and harvesters during cultivation as well as trucks to move the harvested crop to the elevator and then to wherever it needs to go to be consumed.

Then again most of the crops people are talking about doing vertically are things that are planted and harvested mostly by hand, so maybe that's not such an oversight.

Fair points...I'll follow up in another post.

A really great non-profit focused on more sustainable agriculture is The Land Institute. Generally people also don’t also understand that some advancements in agriculture take decades or even centuries.

One example of an advancement from The Land Institute is their focus on domesticating a perennial cousin of Modern wheat. This is no small task given humans have been domesticating modern wheat for thousands or years. Although the cousin still yields relatively less grain, it has significantly deeper roots, is much more resistant to weeds and big in turn requiring less pesticide and can harvested with existing equipment. With time it’s not unreasonable to think it would have comparable yields to modern wheat.

They have a number of projects and been focusing on sustainability since 1976.


+1 to The Land Institute. To give others context, perennials require much less input than their annual cousins, both in terms of labor and also petrochemicals.

The other big benefit is carbon sequestration. Perennials typically root far deeper into the soil, giving prairies enormous amounts of (carbon sequestering) root mass. This also has benefits in terms of erosion control — soil loss is one of the biggest, not talked about threats to society.

Finally, perennials can help — again through extensive root systems — improve water capture, recharging aquifers.

They actually had a run with the grain in General Mills cereal! The commercial name of the Modern wheat cousin Is “Kernza”



Yes, Land Institute is excellent. I follow their work closely. They are on the right path.

Of course they don't understand it, their field is technology. Agriculture requires years of specialisation and most people here if they do have a degree are computer scientists, doctors, biologists, etc. But it's rare to find someone who has genuine passion and knowledge of agriculture. It is far removed from the city lifestyle and it is incredibly hard to break into, both for land reasons and because it's a hard job.

Moreover, agricultural sciences is probably just not a very commonly pursued degree for people in the city (citation needed).

So that brings me to my main point: disrupting an industry is usually done by people who want money when all the other good ideas have been taken. There is nothing wrong with this, but the cost with this fast paced approach is that the oldest and most complex industries like agriculture are going to put you in your place if you haven't done the work to understand them.

Agreed. I grew up in rural Australia, when I moved to the city it was funny to see people talk about farmers in a willy nilly fashion. Farms are extremely hard to build, and the knowledge to run them has to be built into tradition. I'd wager if a governments policies bankrupted a large amount of them, you could almost starve the nation with no remedy. It would take lifetimes before anyone learnt how to till the land again.

I agree that farmers are misunderstood and under-valued by people who are only able to live in stability because of the ability of farms to deliver food to them, but isn't farm complexity able to be documented and analyzed in a similar fashion as other "very complex" fields like law and finance?

It seems like farmers are still beholden to long "if-then" chains and risk analysis (what to plant, where to plant, how to plant, etc. based on predictive yield), just that the underlying mathematics hasn't been as accessibly documented because it's not as profitable.

So "generational knowledge and tradition" are important, but I don't see how that changes the fact that this sort of thing can be written down and analyzed.

(Edit: I should clarify that I am not in favor of "disrupting agriculture" and I also do not think that mathematicians can somehow usurp farmers and plan better farms than the ones that already exist. I'm just wondering what's stopping the logic and practices of the ones that already exist from being documented and reproduced without "lifetimes" passing, as you say.)

I think you're right to believe that because it is true and I skirted over why I believe it takes generations.

It is not impossible but difficult to document how to do effective farming because every farm has its own individual needs. And what may be true for one farm will likely not be true for another. Hence relying on first hand knowledge, albeit extremely fallible, is more reliable than reading a book and then destroying your crop for a year. (Obviously farmers read, study and improve)

The main reason would be location.

- How does water irrigate around your property? Where is the clay? Where does the water lock in when it sinks in different acres? What happens when there is a drought in this area? What happens when it floods? What should you do when there is extreme weather? - What makes the soil in this locale good? What is it naturally good at growing? How should you replenish the soil? What native wildlife contributes to the soil? What insects plague the area and do they have decade long life cycle bursts? What to do when a swarm of locust come? - When does your first frost generally occur? What plants can you grow through a frost? Maybe Kale will survive because although there is frost, you live in a valley where the humidity is higher so the Kale can live. You can't grow X crop because the wind is ever so slightly stronger every 5 years because of atmospheric shifts.

And just to make it more fun, sprinkle on the problem of economics(supply/demand) and logistics.

Also it would be hard not to meet a farmer who calls it a "way of life" because it absolutely is. They live far away from the spoils of civilisation, work incredible hours and live isolated lifestyles. They laugh at city folk because a city man "wouldn't last a week on the farm", which is probably true. Fun fact: Australian farmers have twice the national suicide rate than the average man.

That all makes complete sense, and the misunderstanding of "tech people" made obvious from your paragraph of questions that farmers have to answer. Computers always do what you tell them to. It sounds like farms do not, even when you give them the "right" instructions.

"No farmers, no food", after all, and yet for some reason the suicide rates stay high. It's the same among American farmers. Dairy farms are shutting down at high rates in Wisconsin, where I'm from.

What can the spoiled children of civilization to do help farmers? What can I do? I didn't even know about this plight until I was out of engineering school.

Suicide rates (in Australia, at least) increase with the remoteness of an area. This shows up as an industry phenomenon because farming is an industry that is only present in rural and remote areas. However men in unrelated industries who live in remote areas are also at significantly increased risk of suicide, and recent meta-studies have found a distinct lack of work actually differentiating between "farmers" and "non-farming rural residents", or even a standard definition of "farmers" (is the farmer's wife counted? What if she works off the farm as well?). (Making this differentiation is important because it influences the approaches taken to reduce suicide - e.g all rural people lack access to mental health resources, but if only farmers were committing suicide then that would appear not to be a significant factor)

A super short paper that points to lots of references - https://www.crrmh.com.au/content/uploads/Briefing-Paper_FINA...

This may not be a high quality interjection, but I just want to point out that agriculture has been continuously disrupted for ~250 years.

Back then 90-95% of people worked in food production. Now it's 2-3%, and they produce vastly more food per person.

So it's not like farmers are not used to change.

The term disrupted isn't really appropriate here IMO, it is more of a continual and rapid evolution, which actually even better captures the dynamism of technological change they have to follow. To stay as productive as possible, farmers have to keep up on and integrate new innovations on the fly, often without definitive singular signals one would describe as "disruptive" in tech.

Sure, "disrupted" is a bit hyperbolic.

I was actually thinking if an open source project to do this 'if then' predictive analytics that works across the world is available. It would be a great contribution to humanity if someone can work it , but like every complex problem i do not think it is that easy to distill all the information especially without sufficient profits.

This is what happened in Soviet Russia (productive farmers were deemed class traitors, shipped off to siberia, and obviously net farming productivity collapsed), Ukraine leading up to the Holodomor (knock on effects from russia’s actions), and in Mao’s china (government mandated agricultural actions forced farmers away from their evolved / cultural practices and caused food production collapse).

Systems like this are more complex than the foolish give them credit for being!

Mao's Four Pests Campaign was particularly disastrous and ended killing millions in famine.

He decided that sparrows, which ate some fruits and seeds, should be destroyed. He didn't realize that they also eat locust larva and other pests, which exploded in population without sparrows. Those pests ended up killing massive amounts of crops after the people were ordered to eliminate all sparrows and their eggs -- it ended in widespread starvation.

Specialization sure, but science is universal.

Unless agriculture is built on trade secrets or art, you can contribute.

This is one of my criticisms of Medical. It's not a science or the barrier to entry would be significantly lower, and as a result cost would be lower.

Degrees are good, but not necessary if you can do math and get experience.

So, in my grad program in ag, we had a Bay-area ML startup come in to give a seminar on how they were revolutionizing agriculture. The presented their findings on how to increase yields (which they claimed could only be understood from their algorithm).

The problem they were diving into was well understood, and has been researched to death for the last 100+ years. And they had the relationship backwards, not understanding their "input" to increase yields was actually a response to low yields. They were the opposite of helpful, but rather a waste of our time.

As with anything, it helps to know the current state of knowledge before you jump into contribute. An understanding of math doesn't get you there.

The difficulty with medicine lies in attempting to control the death count involved in gaining the experience. Some supervision required.

There are lots of not-fully-understood processes in the world that only work because we lucked into some way of doing them. If you come at these problems with a scientific mindset but with no real experience, you are going to have a bad time.

There is a lot more to medicine than just math and science. That is not a sound basis for criticizing the field. You'll have to do better than that.

>There is a lot more to medicine than just math and science

This is the problem.

That is not a problem. It's simply a fact that you'll have to accept.

I spent a few months at a consulting company working with a precision agriculture startup, and my mind was totally blown when I first learned how much technology goes into agriculture these days. I feel like a lot of tech people have a mental image of outdoor farming still being somewhat primitive (I certainly did!) which could cause the misconceptions mentioned in the article.

For a small peek watch this simple farmer dig a hole, put a seed in, and put dirt on top: https://www.youtube.com/watch?v=PqK5667B5As

Or think that every pistachio that goes to market has been visually inspected and individuals sorted, and has been for the last 25 years.


Did you mean to post a link to a YouTube video demonstrating that? Your link is a dupe of the parent.

Thanks for letting me know

That's basically a robot, eh? A farm mecha.

My graduate school advisor is a big name in satellite-based navigation (e.g. GPS), and I spent a lot of time learning about state-of-the-art advances in GPS techniques such as precise positioning. I was surprised to learn that many of the former students in our lab went to work for John Deere of all places. At the time, I also had an image of outdoor farming being fairly primitive, but this was an eye-opening revelation to me.

We have auto-steer on all large equipment, full stop. Planting is a science down to the square foot to optimize yields. Spraying is optimized to 2 square inch across every field. Soil checks for nutrients, compaction, and other factors are weekly in the fall and spring, and monthly in the summer. Moisture checks are twice weekly in the summer.

For livestock - they have routine blood screenings for disease and nutrient deficiencies. Rotation through pasture is decided via nutrient content and growth rate of pasture plants. Breeding and genetic lines are strictly controlled via artificial insemination. Animal growth rates, health, and any number of other factors are tracked long-term to decide lineages to keep, modify, or eliminate. All feed supplements are planned to absolutely optimize feed/meat conversion ratios.

The problem with farming isn't that the data doesn't exist, or that the technology isn't being used. It's that the data lives in 18 different places, some in my head, and that the technology is ungodly expensive.

The only way I can see to make SV and ag work well would be to focus on what would otherwise be mid-sized businesses. Large scale operations already have the tech and data. The farmers who run operations of <2000 acres can't afford the large scale purchases, and do much of what I talked about via 'inherent' and 'inherited' knowledge (i.e. they know the north pasture needs to be emptied for two months early spring, but don't know how to improve the plant growth there without messing everything up).

I met a startup where I live during an event. The develop a solution for famers to integrate all their data in one system instead of spreadsheets, think of a farming ERP.

One of the founders worked since his early teens driving large machines during harvest season. He said that agriculture is already now able to be fully automated, from GPS controlled tractors and such to milking and feeding robots. I had the same revelation, modern farming is way more tech heavy automated than I thought.

My PhD is in sustainable agriculture, and I have 18+ years experience in both field and greenhouse ag. Ironically, unlike many here, I went from agriculture to data science/programming.

One thing missed by a lot of the comments: Indoor systems tend to be incredibly fragile affairs. If you've ever been in a well managed commercial greenhouse, you will notice a ton of sanitation procedures. There are greenhouse pests and diseases which are never an issue in the field, in large part because there is an entire ecological system of checks and balances working out in a field. Even in modern intensive ag fields. The truth is an agricultural field is an amazingly complex system which we don't fully understand (we are only starting to explore soil ecosystems and plant roots). Vertical farms are disconnected from this, though the costs might not be obvious. As a consultant, I watched a "trendy" aquaponics startup crash and burn because they underestimated this.

With twenty years previous experience as an agronomist I can tell you that this article is right on.

You want to know cropping ag's biggest problem? Too much data. Farmers are collecting all sorts of data - soil samples, weather station data, aerial infrared photos and yield monitor data to name a few. But there are few tools that give actionable information from all that data. Actionable in prescribing something that results in a positive ROI.

Now as an agronomist who soil sampled, walked the field multiple times every year and sometimes even rode the combine with the farmer I was able to do that - sometimes.

Someday it will happen but it's my opinion that AI is a long, long way from performing that job. But I do hope I live to see it.

Thanks. Agree---lots of data, little 'actionable' solutions out there

Out of interest, what kind of prescriptions would you hope data could make? Are you talking about coming up with new ideas for production or further optimisation for what already exists?

Farmers were told to collect all this data back in the early nineties at the beginning of what was called the precision farming movement. They bought the sensors, collected the data and they're still waiting.

They're looking for the data to point at problems that could be solved to make them more money.

“Premature optimization is the root of all evil”

It should come as no surprise that programmers who spend all day thinking about the theoretical problems they might run into may be bad at understanding current limitations and bottlenecks in the real world. This doesn't just apply to agriculture. Think of how many startups you know aimed at addressing problems that seem imaginary outside of the bay area.

At the same time, I think we underrate the benefit of naive amateurs throwing themselves into industry. If Stripe actually fully understood the amount of work they had to do to get to the other side of a complex, messy, and competitive market, I'm going to guess they never would have done it in the first place.

If you haven't read about the Green Revolution [1], you probably should.

Basically, this is was a series of technological developments in the early 50s and 60s that completely revolutionized agriculture. High-yielding seed varieties, fossil-fuel fertilizers, chemical pesticides, etc. During the Green Revolution, the proportion of common feedstocks that are edible grew from 4-5% to 40-50%, and the number of humans that can be supported on earth by a typical 2000-2500 calorie diet grew from ~1-2B to 10+B. Most of the things we hate about modern agriculture - pesticides, GMOs, monocultures, Monsanto's dominance, the loss of small family farms, coupling between agriculture and fossil fuel extraction - came about because of the Green Revolution. But without it, 80% of the world population would be dead or never born.

Agriculture isn't really in need of Silicon Valley style disruption, because it happened in the 50s. We currently produce enough food on earth for everyone to have a 3000+ calorie daily diet, and we could increase the world population by 50% with current food output and still have enough to eat. The problems with agriculture today mostly concern distribution and tail risks - we produce plenty, but it's allocated inefficiently (wealthy people eat veal and foie gras, poor people struggle to get enough basic grains) and it could be wiped out by a blight or supply chain disruption. Silicon Valley doesn't really help with these problems, and if anything exacerbates them.

[1] https://en.wikipedia.org/wiki/Green_Revolution

Norman Borlaug was my hero. A man most Americans never heard of did more to help people in the twentieth century than any other person. He was more impactful than Elon Musk without all the publicity.

Traditional agriculture has "solved all of the scale problems" through the use of pesticides, destructive monocultures, and disruption of the natural water cycle. It's also built on the idea that diesel is cheap both for the tractors to farm in the midwest and the trucks to deliver goods to markets around the country. Should any of those fragile pillars collapse due to regulation (not likely), major environmental catastrophe (pretty likely), or disruptions in the global fossil fuel economy (possible), solutions like vertical farming start making a lot more sense.

This has the infuriatingly common fatal logical flaw of wrapping "farming" in one giant layer of abstraction and comparing indoor vs outdoor at the broadest scale.

Indoor farming, or greenhouse farming, or high-tunnel farming, or a zillion others are all incremental adaptations of particular plants and particular markets. You cannot compare the global corn and wheat markets to the nyc lunch salad market. "Farming" has always meant thousands of different things, and for some of those things there will be markets for indoor grow ops. This is not an assertion, we all know there's a very robust one right now.

Debating indoor vs outdoor farming at this broad a level is like debating cars vs bicycles as if we have to pick one.

If anyone would like to see an extremely deep dive into the exact scientific measurements at which certain plant markets become viable at certain energy prices you will find this half hour very well spent: https://www.youtube.com/watch?v=wsaufB5F8dk

I find this story odd because it seems like it's over-emphasizing the large scale/ low manual labor crops. In other words, the article is talking about inexpensive, long shelf life crops:

> The Midwest in the United States has close to 90M acres of corn, 85M acres of soybean, and 30M acres of wheat.

Maybe I've got this entirely wrong, but my understanding is most vertical farming focuses on producing highly perishable fruits and vegetables which often still require a fair amount of manual labor and where being close to market is a benefit.

I haven't yet seen large scale vertical gardens being commercially successful yet, but if they do, I'm certain they won't be producing corn, wheat, or soybeans.

There's only so much that can be grown and mechanically harvested, and the US surely excels at producing maize, wheat, or soybean, but vertical farms don't try to compete with those. The production of other crops does not happen in the US that much, but it also operates at a rather spectacular scale - instead relying on poorly paid laborers abroad.

There's a need for automated harvesting machines for a few remaining crops - apples, lettuce, etc. The big field crops - wheat, corn, etc. have been fully mechanized for decades if not centuries.

There are vision guided fruit picking machines. They're too slow, too fragile, and need too much supervision. But they mostly work. What they need now is good practical mechanical engineering. The 2016 version:[1] The 2019 version.[2] When they get about 2x faster, have half the parts count, and can be routinely pressure-washed, they'll be ready. The "AI" part is done.

One of the simpler automated systems is automatic weeding. Machines come in several forms, but the most successful seem to be wide implements towed behind a tractor. Deere has some of these. They recognize weeds with cameras and do something about them. Some stomp or pull, some zap with electricity or a flame, some squirt on an overdose of fertilizer. It's "organic", too; no pesticides. You can get this as a service in a few areas.[3]

[1] https://youtu.be/mS0coCmXiYU

[2] https://youtu.be/-PtqZA2enkQ

[3] https://www.robovator.com/

This is correct. But the crops they target (veggies mostly) are also grown outdoors. But the higher margins on these veggies give the indoor people and hydroponic people an opportunity to compete (also year round veggies command a premium as well)

I was blocked as a suspected bot by the wordpress site, so I'll post here...

I think this discussion requires a bit more nuance. Of course classic row crops like corn, wheat, soy, oats, etc., are unlikely to ever make sense for indoor crops. But that's not what any of these businesses are tackling. Instead they are focused on high value fruits and vegetables, herbs, and fresh greens. You only have to look at the agricultural success of the Netherlands to see that these crops can be grown for profit at scale, indoors. During the winter months they augment the greenhouses with light, but they are also taking advantage of the sun as much as possible. In greenhouses you can grow with far less water, and you can produce fresh, local food that doesn't have to be cooled and shipped nearly as far. I'd like to see an honest comparison that looked at a tomato and a handful of fresh cilantro being sold in NYC or SF from a local greenhouse with augmented light versus comparable produce shipped in from Mexico or somewhere else warm.

I think there are also many good arguments for shifting our diets away from the commodity crops and towards more fruits and vegetables, so as the world gets wealthier and more people seek diverse, healthy foods, we might see new models that make increasing sense.

I immediately thought of the Netherlands as well. As I understand it, they began vertical farming more crops thinking that they would be the R&D department for world agriculture, because they could get results faster, and then it would go into "production" in conventional farms, but were surprised to discover that they could (in some cases) compete for the production as well.

It makes total sense to me that this would not be the case for every crop (probably not corn, wheat, or soy for example). But I would also be surprised if vertical farming made sense for NO crops.

The problem with industrial outdoor farming isn't the efficiency, it's the toxic pesticide applications, the environmental pollution in air and soil getting in the crops, and lack of nutritious crop diversity leading to inefficient food supply chains. Growing plants with coal indoors isn't a solution either but hyper-efficient indoor and vertical farming is getting closer by the day and more funding needs to go into new evidence based controlled environment farming techniques. Check out https://youtu.be/VIrXQo00OWc for an example of what hyper efficient indoor farming looks like.

I really like this video series from Exa Cognition on Vertical Farming that goes over many of the issues the article posts.

https://www.youtube.com/watch?v=dnCQuwCtqJg https://www.youtube.com/watch?v=qGyAeqdkkbw https://www.youtube.com/watch?v=qGyAeqdkkbw

And here's a video from Techno Farm he mentioned


really good videos, although an honest breakdown of costs is what will help with the naysayers

As a consumer, I don't care a tiny bit about making my food less expensive (by making production more efficient). Food is already very cheap.

I care very very much about reducing the suffering of farm animals. I do not want to become vegan (for health reasons), but the guilt I feel because of my contribution to animal suffering is one of the worst parts of my life.

Please, please, smart young technologists out there: figure out some cool technology to make it possible to raise farm animals efficiently while also ensuring that they live comfortable, decent lives.

Not everything is a technical problem. No technical solution is going to make the cheapest way to get animal parts in our fridges involve the animals living fulfilling lives.

I've been thinking about this in the context of applied ecology (Permaculture, et. al.) It's undeniable that modern mechanized mass agriculture is incredibly efficient and already highly automated. It's kind of fantastic. (And very challenging to compete against.)

The two main downsides (IMO) are related: fragility and ecological ignorance.

The article touches on this: "soil is a natural resource that will become endangered if we do not mitigate the severe erosion problems that stem from single species field that are barren (re: nothing actively growing) for 30-40% of the calendar year (in North America)."

(Imagine installing millions of acres of solar panels and just switching them off for 1/3 of the year.)

Broadly speaking, if our agriculture destroys topsoil rather than creating it we're gonna have a bad time.

An interesting challenge would be to automate food forests. For concreteness, check out what these folks are doing: https://www.youtube.com/user/plantabundance

This is one family working on their home plot in a suburb who have converted it into a really cool food forest with chickens and lots and lots of different crops.

Imagine replicating this across millions of acres, without involving hundreds of thousands of people (which wouldn't be a bad thing, but it couldn't compete with mechanized agriculture.) What kind of automation could help with that?

Fallow fields are essential to maintaining soil health, a successful practice going back 10k years. Additionally, some of the "barren" fields are actually being used to alternate crop - it is a technique to grow crops with limited natural water. Modern best-practices avoid a lot of erosion by using no-till or low-till systems.

I read up and watched some videos in re: "syntropic agriculture" last night. Ernst Götsch's farm produces crops all year w/o any inputs and with continuous improvement of the soil quality.

Several weeks ago, I finally looked up what "permaculture" is all about, and it blew my mind away.

It works even better than modern agricultural processes. The ideas are inherently distributed and decentralized, and when implemented along the ethical principles, bypasses many of the wealth inequality. It builds up resiliency through diversity (something the tech world is only starting to explore with Kubernetes and containers). It goes beyond mere "sustainability" and into regenerative processes. These are very practical ideas that have had 50 years of implementation proving out those design patterns.

It requires a different way of thinking about how we grow and get our food.

My techie friends all love it when they see these ideas, and yet, the development of permaculture design tracks the development of the personal computer, internet, and smartphone. But it is also big blind spot. Many of the design patterns are low-tech or no-tech (which, not depending upon a supply chain, is much more resilient).

Dwarf Fortress is a lot of fun, but I have found that applying and implementing permaculture design is a lot more challenging and rewarding.

Weirdly, the thing I find missing most from this discussion is finance.

If we take the premise that information is valuable (decrease inputs, improve yields) and that equipment is valuable (automation), then there's a very real return to be gained by using these valuable products. What's nuts to me is that we would ever ask the farmers to bear the risk of these products. I mean, sure they could I suppose (better returns overall!), but it concentrates all the risk in the worst places.

I feel like there's a much better opportunity here for a targeted financial product. "Implement our methods with our data, and we'll skim a percentage of your profits." Imagine if the risk of buying a new tractor was gone, because it was provided by the company. The risk of data integration was also gone, because it's guaranteed to work with the provided tractor. And the risk that the data is crap is also gone, because the financing and return risk is borne by a diverse number of institutional investors. Almost like weather / crop insurance, but much much bigger.

For the farmer, the sale is simple: Do our thing, and you don't have to worry about paying for stuff that may not be valuable. You might make slightly less total profit in the good years as the price of offloading that risk.

For the investors, it's also a great story: look at these great startups! Wouldn't you like a piece of that productivity and return?

...and maybe the startups are wrong, and the equipment doesn't work, and everyone learns a lot while they go bankrupt. Everyone except the farmer, who offloaded the risk.

Surely this is a thing that someone is working on?!? I'm sure it's available in bits and pieces, but a unified financing and operational solution seems like it would be a slam dunk.

Why did the title get changed from the article's title ("What Silicon Valley Doesn’t Understand About Agriculture")?

Because Silicon Valley thinks nowhere else matters, and likes to pretend its parochial local monoculture is never a problem.

I would hope that the savviest of SV or startup folks recognize that the actual planting, growing and harvesting of crops isn't the thing that they are trying to innovate on.

Rather it's the disastrous logistics chain and resultant waste, leading to overproduction and augmentation of our food system, is the problem trying to be solved.

> actual planting, growing and harvesting of crops isn't the thing that they are trying to innovate on.

There are enormous problems there, such as a heavy reliance on human labour for picking and processing.

The problem with going after waste reduction is that the tab for that is picked up through subsidies. The inefficiency is in policy.

The problem with human labor in agriculture is that we've already spent a lot of effort to remove humans, the jobs that are left are really hard to automate.

Can you explain this "disastrous logistics chain"? I am a bit confused given all the effort that goes into logistics by farm companies.

It's the logistics chain from the farm to your plate that's the disaster. That's not for lack of trying. But having exposure to farming growing up, there's no good way to get a ripe berry from Washington State to Florida before it spoils without freezing it, genetically engineering it (which is fine, I'm pro GMO, it's just costly) or treating it. Even if you do, handling it will see large losses.

So the trick is, how do you reduce how far something needs to travel from the time it's ready to harvest until it's consumed.

Well, you can get it there, but berries aren't worth enough to make that journey by air. That is generally reserved for seafood and there is an equally amazing logistics system to do that. Calling one of the most amazing processes on the planet "disastrous logistics chain" is just disingenuous.

30% of all food produced - approx 1 Trillion dollars worth - is lost in the supply chain, and contributes the equivalent of the third largest CO2 producer if it were a country.

I'm not sure how to describe that other than a disaster.


There is room for improvement, and there is significant financial incentive to realize that improvement. That said, I'm not sure that measure is the whole picture. If we are losing 30% of our food to the supply chain, what is the alternative? Perhaps we could farm things more locally and shorten the supply chain? If we did that, would we still get 100% of the yield of the old approach? What I am saying is that if we try and fix the waste problem, it would very like be at the expense of reducing yield. The extreme example is the tomatoes I am growing in my back yard. None of them will be wasted, but I'm fairly certain that the yield per square acre is absolutely atrocious.

I call it an unavoidable cost of making sure people have food on the table. There is no way on this planet that you can design a supply chain that you produce the same amount of food that is consumed. This isn't parts that get put in some widget, people have different tastes at different times. I'm honestly surprised its only 30% given the fickle taste of people.

The CO2 production will reduce as we steadily change from diesel to electric. Ocean going vessels are just environmental problems that treaties seem to ignore.

There are two main options to avoid having to deliver food from (for example) Washington State to Florida.

One is to change consumer behavior to focus much more on in-season products that can be grown locally. This is a difficult social challenge.

The other is to change plant behavior so that they become products that are always in season and can be grown everywhere. This is a difficult technical challenge, but things like indoor vertical farming can potentially solve that. The problem described in this article is that it does not (yet?) work for effective farming, but making it possible to grow the appropriate berries or fruit locally throughout the year would fix the logistics chain by eliminating most of it.

About two years ago, I started with a much fancier AgTech with Hydroponics. Went all the way to the YCombinator interview in Mountain View (the last one for Indians), and rejected with something in the lines of "not advanced enough".

Spent time researching, talking, and more researching about the core problems of Agriculture in India. It is one sector where everyone loves to toss and play around, the most politically involved and abused, with huge numbers but contributing less than 20% of the total Indian GDP. Everyone seem to have a vested interest -- both good and bad.

At times, I'm shit-scared that I'm trying to help solve something so massive and gigantic that if I can make an iota of difference, it would be huge.

Of course, my hammer is Technology and I'm trying to find just the precise nail-heads to hit, one at a time.

You seem like the right person to ask about this.

My sense was that hydroponics were so expensive that they really only made sense for one crop: marijuana. If your plant is selling for $1000, then spending $50 per plant to increase its yield and 'baby' it so that it sells for $1200 makes sense. But for most plants, it doesn't make economic sense. (Tomatoes, maybe. But they're going to have to sell for a higher price point, in more expensive markets).

Is that accurate? I'm an amateur so if this is uninformed, feel free to correct me.

There are lots of really expensive products that can be cultured indoors, are perfectly legal, and aren't drugs

I'm not being dense here - I really thought there weren't? Like the core problem with ag as a business is that it's very low margin and it's extremely hard to make money.

I can only think of a handful: marijuana, fancy tomatoes at Whole Foods, maybe coffee or vanilla (?) beans... but even there, notice how there's either fierce competitive pressure, or it's possible to easily overwhelm the market (vanilla beans have to compete against synthetics). I mean there are surely some I'm forgetting, but it's a small number, and I thought that deterred VC investment.

For your staples - corn, soy - that's still much cheaper to grow outdoors, so hydroponics can't win that market.

Would be happy to connect to discuss more. There are plenty of opportunities in India for improvement, it is just that is a complicated market. The humanity improvement aspect here is bigger than the financial one for sure. The only way to start is with something that works low tech and cheap enough for average farmers to get into.

Sure. My profile has connection details.

Not strictly relevant to the article, but I'm not excited by the efficiencies as much by the idea that it might offer a space-efficient option for me to have my food grown locally.

Food production at the moment is very much out-of-sight, out-of-mind. I don't have a feel for what monoculture are developing in the food industry, I don't have a feel for what the supply chain risks are. If food ever stopped flowing in from wherever it comes from to my city, I'd be in trouble.

It isn't totally rational, but I dream of being able to invest in food grown a few blocks away from me. If it only cost double existing prices that'd be a solid win.

I'm not sure how prevalent this is around the world, but in my city we can register for produce baskets. In spring, I register on a website and I choose and pay in advance of the whole season a local farmer (less than 50 km away), who comes once every week to distribute his baskets a few streets away. The produced is freshly picked the same morning, it varies from week to week, it's a small family farm and I know it doesn't contain any pesticide or artificial fertilizer.

That's common in small town or exurban areas in the US.

Uh, for you. Some of us live near farms -- it's neither out of sight nor out of smell (heh).

The US is the most agriculturally productive nation in the world by a fair margin. Food and fuel are two of the things that the US is unlikely to run out of even under conditions of global nuclear war.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact