Hacker News new | past | comments | ask | show | jobs | submit login
What vertical farming and ag startups don't understand about agriculture, part 2 (thinkingagriculture.io)
121 points by kickout 14 days ago | hide | past | favorite | 122 comments

> Auto-steer systems that leverage machinery that already exists.

That links to a company selling a retrofit system for 7500 Euro and no annual fees. That's super cheap compared to buying a new tractor and I hope they are successful. But...

My brother-in-law hobby farms a few hundred acres of corn and soybeans and built an autosteering system for his 1950s tractor. The only thing he bought new was an Arduino board to tie it all together. Everything else was used farming equipment and his total cost was less than $1000 (USD). He's not even a programmer, just an engineering degree from Flyover State U. He cobbled together open source software to make it work.

The traditional farmer is very resourceful. They aren't going to allow your company much margin if they can do even part of it themselves. Good luck to VC-backed firms.

> an engineering degree from Flyover State U

Looking at the US News and World Report rankings, 5 of the top 12 schools might (depends how you count Pittsburgh and Austin and) qualify as "Flyover State U." There's a huge range of schools in flyover states, and there are some very well-regarded engineering schools there.

I think parent meant that the schools that folks who work in ag go to tend to be "ag and mining" type schools, most of which are not state flagships.

IME that's true at least int he midwest. Most people I've met who work in ag in the midwest did not attend Madison/Ann Arbor/UIUC/etc. They attended places like SIU-E, UW-Whitewater, Minnesota @ Rochester, Missouri @ Rolla, etc. Usually because those schools were like 1/2 of the price and closer to home.

And what parent said was also true -- those schools do teach lots of useful skills; if you're selling machines to farmers, your clients probably know more about mech.e than even most of you software engineers.

"They aren't going to allow your company much margin if they can do even part of it themselves." Interesting observation.

It reminds me of the myriad of products and services targeting bootstrapped/self-funded entrepreneurs, who also tend to be resourceful by nature.

Or programming tool companies selling to programmers who notoriously will question why your tools aren't open sourced?

This is true for a large fraction of the programming world, but not all of it. As just two examples I happen to know:

  1) graphics / games / vfx
  2) energy / mining
It is the norm in these markets to pay for industry leading software tools, and I know multiple companies making profit-per-employee in ballpark FAANG range.

Cool to see that the same open source dynamics that make it possible for three engineers to ship a great web app in 3 months (Linux, Node, React, etc) also have made it much easier for people like your brother to use off the shelf tooling like Arduino to solve their own problems.

Do you mean to tell me that such a small operation just happened to have lying around some old lidar systems?

Or does the "steer" in "auto-steer" refer to a bovine?


You program your desired path into the system and it uses GPS to follow the path, using a device that turns the steering wheel as necessary.

The farmer is still required to avoid unexpected obstacles, usually by stopping and moving the obstacle.

Farmers have been using GPS since the 1980s, long before the rest of the general public.

You certainly don't need lidar to do auto-steering on a farm that you own yourself. You can get decent straight-line results using a PID controller on a compass alone, and tossing in a GPS module would make the whole thing even more accurate and robust. As as the OP said, there's already a bunch of open-source stuff out there to help make that happen.

It's doubtful that advanced computer vision is required to steer a tractor around a field. Adding it might improve yield somewhat, but consider the law of diminishing returns.


Many people would be amazed by what I have lying around in my basement. However, the LIDAR is gone: I only needed for a single project and I shipped it to the client.

That was a pretty dumb farm joke. But it got funnier when so many techies got distracted by the lidar reference.

If you assume there isn't anything in the field for the tractor to run over, you can just use GPS.

While you are making convincing case for that particular farm, let's take it into the next decade or two:

The difference between a U graduate and a non-U software developer is in the experience (and mindset) of fixing a problem vs scaling the solution. I.e., fixing the meta-problem.

A scalable solution is the one that becomes permanent, what moves the needle of the market, and what gets written up in history books.

I've never been able to understand the excitement around vertical farming from VC. I assume the primary reason for VC interest in this boondoggle is just that sleeping at the Comfort Inn in Sedalia, MO is not nearly as fun as a champagne black tie at the media lab.

But, during COVID, I now think there actually is a strong case for large-scale vertical farming: robustness. It's a backstop against famine in case of long-tail events: "locusts", plagues that wipe out key crops, once-in-a-millenia systemic flooding, a super-volcano eruption, etc.

But it's not going to be profitable. It's going to be horrendously unprofitable. And it's not something we should be doing now; it's just something that we should be able to quickly start doing immediately in case of a catastrophic event, and at a scale large enough to get 1200-1500 calories to everyone each day.

In other words, the primary customer for vertical farming is government, not the private sector. Unless you can mass produce morels, as the author notes :)

No one wants to start a career in Agriculture, it only makes sense when you are payed extremely well and currently that's the case when you can make the bare minimum but that minimum is a lot somewhere else.

In West Europe, Agriculture is done by Eastern Europeans who can work for few months and make some money that is uninspiring the West but life-changing in the East. Western youth would't touch it.

In Turkey it is done by people who escaped a war or tyranny : Syrians and Afghans. It makes sense for them because the alternative is worse but the Turkish youth would not touch it unless it's some kind of organic avocados hipster thing.

The Agriculture production is now cheap but unless everything gets robotised at the same pace(or faster) with the people leaving the business or the inequality in the world vanishes it can get real expensive real quick.

The vertical farming looks like an option before automatic drones produces us all the produce in a sense that no one ever seen a tomato field in life.

But as the author states, farming is massively automated and the efforts are still going on. So, maybe it's about lifestyle? Having a vertical farm in each garden could be desirable even if not efficient.

It’s interesting to think about it this way. Thanks for the new lens.

Poster makes a great point about what SV doesn't get about farmers, but there is a huge range of things farmers do get about SV that are not obvious to many people who don't know farms or farmers very well.

I used to be CEO of a venture backed agriculture startup, and we learned pretty quickly that the hacker mentality applies very well to supporting farmers. Internally, we talked about providing farmers something like an SDK for the automation tools we were selling, where they could mod it to their own situation. In most cases, skills from soldering to metal work to carpentry were not only realistic to expect, but complete shops with air tools, etc. were usually present and often complete machine shops with CNC and 3d printers.

OP would be glad to connect if you see this note.



Thanks, perhaps I will reach out later today (going to fields ironically enough!)

I think that the biggest problem with agriculture today is that it's utterly unsustainable. It is ran on fossil fuels and we need to mine minerals for the fertilizers. But still the most worrying issue is that the modern farming is consuming topsoil and we will run out of it in 60 years if we don't change our methods [1].

There are interesting things going on in regenerative agriculture scene where the main idea is to produce new topsoil and through that make farming sustainable. Here is a nice lecture by Richard Perkins - who is one of the most famous regenartive agriculture advocate - explaining the principles of it in a lecture he kept in a food hackaton. https://www.youtube.com/watch?v=3Knn7ZH4Tiw

(1) https://www.theguardian.com/us-news/2019/may/30/topsoil-farm...

The kicker for me was reading the very consumable Dirt to Soil [1], written by a farmer who's applying regenerative ag techniques on a large scale, and very profitably. Transitioning from input-dependent, technology-sucking financially-overextended traditional monocropping IS POSSIBLE. This isn't just your backyard garden anymore. Something needs to change in mainstream ag. I'm moving onto a small farm to do something about it, in my small corner of the world.

You've managed to describe exactly what annoyed me about these two posts. OP's argument is that whatever is cheapest this year is the only thing that should ever be done. We definitely need something better than current farming practices.

The amount of economy of scale in agriculture is disconcerting to me. Surely finely-tuned, monoculture mega-farms in rigid supply chains have disadvantages?

I was hoping with improving technology, the minimum viable size of a farm can be brought down to something more culturally sensible. Even if existing technology in agriculture is as good as the author claims, access remains an issue.

Perhaps the real innovation we need right now is to make technology more affordable.

Monocultures are vulnerable to catastrophic phenomena that affect the monoculture. See Irish Potato Famine for one example.

Additionally, monocultures tend to deplete soil nutrients, so need land management in the form of fertilizer, crop rotation, or lying fallow. The latter two are often combined.

These are (obviously, I hope) not insurmountable.

I'm wondering this myself - and I think it's worth doing the math to sort out. To me, the essential question is "Is it possible to grow produce in a decentralized fashion?" i.e solar panels on the roof of a large apartment complex generating enough power to sustain the building.

Could a hypothetical FarmPod3000™ exist at a reasonable scale? I mean, people built greenhouses before right? Could such a system be economically viable at the size of a shipping container? A family home? Staten Island?

To be clear, I don't think the FarmPod3000™ would disrupt 90M acres of corn farms - we actually eat <10% of that anyways[1].

We should factor in the +/- environmental & public health impacts of such a system to the viability calculation as well.

[1] https://www.washingtonpost.com/news/wonk/wp/2015/07/14/how-c...

One inherent inefficiency about mega farms is that most of the soy (not sure how much of the corn, but a lot) goes to feed animals. Eventually, we "harvest" 10% of the input energy as calories. So these systems will never be more than 10% efficient in terms of human sustenance, no matter what technology is thrown at them.

Soy use in Brazil: https://globalforestatlas.yale.edu/amazon/land-use/soy

Trophic levels: https://en.wikipedia.org/wiki/Trophic_level#Biomass_transfer...

Can you think of products where mass-production makes things less efficient?

The apparent efficiency may depend on whether the manufacturers and the immediate customers pick up all of the costs, or whether negative externalities or other costs are picked up by third parties. E.g mass agriculture may make food cheaper at the cost of environmental damage due to use of pesticides, soil erosion, etc. These costs may be picked up by wider society or just ignored.

Yeah, in the big picture that's true - lots of industries suffer from this - but I meant from a "disrupt the market, make lots of money" frame of reference, which is where all these things seem to start from.

After reading Part I yesterday I got to thinking about SpaceX.

In his biography, Ashlee Vance describes how Musk built a knowledge base about rocket construction after his exit from Paypal. Both the "how to" and a first principles approach to the economics of it.

Musk knew the calculus that he could attempt to build rockets for far less than his competitors before he ever put a dime into SpaceX.

In contrast, vertical farming sounds more like the computer engineer equivalent of "I built a thing, isn't it shiny and cool" without understanding the industry, economics, or competitive landscape.

Vertical farming seems to be taking an ahistorical approach. With both both vertical farmers and VCs ignoring industry knowledge [0,1]

Sticking with Musk, he took an ahistorical appraoch to Tesla's Fremont factory with automation. He learned a very expensive and wasteful lesson that other auto manufacturers had known for decades.

[0] https://news.ycombinator.com/item?id=23633298 [1] https://news.ycombinator.com/item?id=23634030

> Sticking with Musk, he took an ahistorical appraoch to Tesla's Fremont factory with automation. He learned a very expensive and wasteful lesson that other auto manufacturers had known for decades.

Sometimes things change and the lesson everyone learned is no longer valid. Fools sometimes get lucky.

And sometimes they don't. In case of automation, Tesla really ignored decades of experience just because. And it wasn't even slightly related with anything tesla really inovated on. Not sure these "maybe we are lucky" initiatives are a wise thing to do.

The allure of vertical farming also seems to lie in the "urban vs. rural" culture war, in that farming in cities claims to obviate rural existence

Thank you for pointing this out and making the implicit explicit for me.

There is a lot of economies of scale to be achieved by more growing cycles, better customization of growing conditions, improvements in LED in the future.

The bottleneck will always be cheap energy. However the industry will progress from non calorific high value foods like lettuce and leafy greens (they may be low calorific but have a lot of value in terms of nutrients etc. and are expensive) to stuff like cherry tomatoes, cucumber as renewable energy starts getting cheaper.

The fundamental difference between SpaceX and Farming is that the latter actually provides direct utility to humans.

I still don't understand why we need to go to space. I know I'm going to get roasted here for saying this. And yes, I know that "space tech improves earth tech" -- cool, so let's just make better earth tech in the first place. Better farms, schools, transportation. We only have one earth.

"I built a thing, isn't it shiny and cool" -- sweet rocket Musk, what are you going to do with it? Study space worms? Let rich people travel in space? Build a mars colony (lolol). That's not going to happen on any meaningful scale. Put down the sci-fi and open your eyes. We need help, here, now.

You've probably already read it, but I'll link it anyway for those who haven't:

## Letters of Note: Why Explore Space?

[+] https://lettersofnote.com/2012/08/06/why-explore-space/

Why go to other countries? Europe+Africa has all the space we need.

1. There is an unimaginable number of resources in space. Asteroid mining, if we crack it, would make space a far better place for manufacturing.

2. It was originally rich people flying on planes and sailing on boats. If it's technically possible and valuable, economics will optimise the cost away.

People were traveling on boats already and traveling on boats wasn't as hazardous as sending people on a controlled exploding vehicle. Plus on earth, existing tech in europe/africa was going to work in most places of discovery on earth. Everything in space will have to be re-invented, without the thousands of years of iteration (like ag). Have people really even figured out how to farm on Mars? No because no one has farmed on mars. And farming on mars, that tech may likely not be useable on earth, because we've spent the last thousands of years optimizing for our planet. Utilizing the soil the sun and existing manufacturing. We'll build tech for mars, and then what next? What is the next uninhabitable planet we'll visit and rebuild entire infrastructure, while abbandoning the 1 good planet we have. I don't disagree that space travel is important and finding planets is important, but i think those things pale in comparison to the problems of our only guaranteed hospitable planet.

edit: I think pushing limits of human achievement is important, but understand that we haven't even solved stupid obvious problems like feeding every child in this world. Think about that. There are hundreds of millions of children who don't eat. You want to talk about scale and talk about solving problems, think about all the minds that could be the geniuses that get us to mars. There is so much lost potential just because of malnourishment. Yes going to space will get us better manufacturing, but the potential of having entire generations that are well fed and thinking optimally, we could be squandering countless einsteins, simply because we are spending money trying to get to mars. I think that's what gets me about the way we spend money. Human output is random, and the people who change the world, it is random where they are born and so is their upbringing. We should be taking care of every single new born that enters this world because one of them, could change the world in unimaginable ways. Sorry I don't htink you asked for this comment, but your comment about why people traveled to new countries just sparked a thought dump. Sorry :)

Provide low latency broadband internet to the entire surface of the Earth (Starlink). To me that's a good enough benefit to justify it.

Musk is serious about building a Mars colony. Of course if you dismiss the possibility offhand with a "lolol" then you won't agree with what he's doing. But personally I think it would be beneficial to have space colonies, and we eventually will have them in the far future if we don't blow ourselves up first, and accelerating the timeline for that is a fine thing to work on. SpaceX is not zero-sum with farming.

Space exploration gets such a small fraction of our resources as it is.

I agree we can invest and function better in other domains also. It's just not a choice we have to make; we easily have the bandwidth to do all of it.

And the practical benefit to space exploration or readiness is non-zero right now. Satellites, defense (from others but also celestial objects), climate science. The future hypotheticals, your "Sci fi", may one day be added to this list.

The GPS that your automated combine uses to guide itself is space tech, not earth tech.

Modern farming is heavily dependent on positioning and weather satellites. The quality of these services scale with density and SpaceX lowers launch costs which allows more satellites which in turn will create more efficient harvesting and help prevent crop losses due to unexpected weather events.

> And yes, I know that "space tech improves earth tech" -- cool, so let's just make better earth tech in the first place.

If you threw a bunch of money at just "try to replace common pieces of earth technology" I don't think you'd get much.

Having both specific problems to solve and new and difficult constraints spurs creativity in a way that just "how would I make a better fastener" wouldn't. It's too open-ended otherwise. What does "better" even mean there?

> I still don't understand why we need to go to space.

I'm sure you don't need GPS. Or care about researching the Climate. Or care fundamental research on physics. Or want to use the internet on planes. Or care about defense. Or ever use earth imaging made from satellites. Or care about people in disaster situations find a route to flee the region. ...

We should probably get to space one day. Doesn't mean we can't solve earth problems too.

No practical reason. Same with understanding cosmology or studying dinosaurs - all of these things have zero effect on our day to day lives.

https://www.planttape.com/ mentioned in the article is so cool. I wonder if they were inspired by pick and place machines with electronic parts on reels.

Machines like these exists in all scales you can imagine.

I had new vineyard planted two years ago. Big tractor with automatic planting station arrived and with GPS base station they were able to plan in around 2cm accuracy. There was no automatical steering, but whole machine and tractor shutted down when driver move out of mission plan. You can decide what is row spacing going to be, what space between plants are going to be and you are done.

What would take me week of manual labor took machine just 40 minutes (well kinda: one servo broke down and they had to call mechanic from Germany which took off two days of very expensive machine time in prime season).

Here you can see simmilar machine: https://www.youtube.com/watch?v=cC32JgiDxo4

There have been similar systems for seed-sowing for decades. This is the first system I've seen anything like it for seedling transplanting, but the concept is not so far different from the seed-tapes that horticulture has used for quite a long time, so I'd be more inclined to think the inspiration came from closer by than electronics.

Correct, its a horticulture driven innovation (makes it more cool IMO).

https://paperpot.co/ is the same idea as a manual labor version. I never used it since you need more machinery to prepare the beds for it, but from what I read they are very popular with small scale market gardeners.

Might be coolest engineering solution I've seen in my short time on this planet.

Presumably the submitter of this link is also the author?

Do you have any suggestions on what areas are good for innovation?

I also find myself at a complete loss for what technologies are actually out there, how they're used and what a typical farm looks like and how it operates. Even the data of "90M acres of corn, 85M acres of soybean, and 30M acres of wheat" is complete news to me.

What are some sources to get at this data, to get a feel for what this system looks like?

I know that the "global village construction set" [1] was marketed as trying trying to solve some of these issues but I'm not sure it really developed into anything actionable.

Keep up the good work, I'd like to see more writing. Do you have any suggestions on other blogs that talk about similar subject matter?

[1] https://www.opensourceecology.org/gvcs/

Umm...start with USDA NASS for agriculture statistics in the US. agfundernews.com for the startup scene. Just reseach in general. Deep Learning, Nanotechnology, Drones (UAVs), genetic engineering all have a home in agriculture. I'd be happy to PM you more resources.

Yeah USDA NASS is the spot for your ag statistics. even then NASS is not 100% accurate but pretty solid. Drones, meh. nobody wants pictures of their farm bud. What do you do with pictures of your farm? Nothing. I've been working with farmers for 6+ years and only a handful of them like the drones mostly because it's a fun toy. Sure tell me that you can get AI imagery and machine learning non-sense from your farm pictures.

The only thing i agree with you on is genetic engineering. And that is not startup space really. That is like giant bio-tech companies breeding corn seeds in a lab. You need a phD for that shit you dont just build a startup that 'disrupts' seed genetics.

Free idea: There is MASSIVE distrust in the NASS right now in the farming community to the point of the data becoming questionable. It'd be nice if place like Indigo Ag (which acquired TellusLabs I believe) could assist/takeover some of the things remote sensing or satellites can do (like planted acres) so that USDA NASS doesn't need to rely on unpopular and time consuming surveys

Where can I find the communities that are upset/distrust NASS? Is this on twitter/forums/fb? I'd like to take a deeper look into the problems they're discussing but I'm an outsider, not sure where to begin looking.

One problem is that farmers often consider it a secret they don't want to share. For good reason, if you know how much crops will produce next fall you can go to commody markets and make a lot of money. That money comes out of the farmers potential profit in some cases as they are playing the other side of the trade and so the less you know compared to them, the better the deal they can get.

Farmers know what is in their fields, but prices are affected by fields on other continents which the farmer doesn't know. Any information the farmer can get but you cannot is to his advantage. (this applies to all farmers, they need to know what happens in Ukraine, Brazil, Australia... But they need you to not know)

Very late reply...But if you want to hear an admitedly biased (towards negative) source, go read the crop comments on agweb.com

Not affiliated with them, but there's a startup in New England that sounds like they've gotten pretty good results and feedback using drones. I think they use hyperspectral sensors find things like areas that are under/over-watered or that are running out of nitrogen, etc. The whole things is supposedly automatic, so you don't even need someone directly piloting the thing.

I know that there's been some success in remote sensing of aquaculture at the academic level that's also demonstrated some promise with drones.

[1] https://www.american-robotics.com/

A note on genetic engineering: a relatively new tool called CRISPR is changing the field. A grad student with a few hundred dollars is now capable of inserting, removing, or modifying specific genes in the entire genome. It's becoming relatively trivial to make Frankenstein plants.

Check my Bio. I'm familiar with big bio-tech companies that breed corn, :)

VERY familiar, lol.

Saw a drone they were using at a farm in North Dakota. It was definitely big enough to need an FAA license. They were damn enthusiastic about it, I get the feeling a couple of them chipped in for it.

If the incumbents keep acting like the backside of a donkey, a new generation of repairable farm machines would probably work but need a lot of capital and damn good relationships with local dealers.

To your first point, a fair number of drone "service companies" have gotten going in the state, so owning your own isn't completely necessary. If you want to sell to those companies (and the farmers they serve), developing new sensors and/or analytics to back them up is one way to get into that space.

To add, if you are a tech person study the combines following the harvest from south to north. That will lead you to so many aspects of what modern farming in the great plains is and probably give you some ideas for tech services that farmers would be willing to pay for.

I want to see the end of skim the top businesses that come out of SV. They pretty much set up a database, a front end and expect everyone to pay them a percentage of a business that already works. But because they capture a young crowd, it eventually takes over. I hope Ag is resistant to that.

I'd rather see SV help the industry keep its jobs and just create new machines. Like the drone that picks weeds (or just makes a map of all weeds in the field for someone to go pick). Perhaps a drone that scoops up bugs via vacuum attachment.

Whatever it is, I'm kinda just sick of SV starts that want to skim the top of some existing process.

https://farmersfriendllc.com/ good example of a thriving company in the space, though smaller scale and no VC funding

Anyone have an alternative link? My company relies on Cisco's reputation system[0] to decide if a domain should be blocked. Oddly, thinkingagriculture.io is "untrusted | poor".

[0] https://talosintelligence.com/reputation_center/lookup?searc...


hopefully this works for you

Sorry, brand new blog. Probably some security issue I lack knowledge of

This will happen anytime you put something new online. It's basically a whitelist and they wait until they see a few hits before you get added.

Modern mechanized agriculture is incredibly efficient and productive, but it degrades soil health over time and loses topsoil.

There are methods that increase soil and fertility over time (Permaculture, Syntropic farming, regenerative agriculture) but they are typically relatively labor intensive. China and India have made great strides in recent times.

India's Water Revolution #2: The Biggest Permaculture Project on Earth! with the Paani Foundation https://www.youtube.com/watch?v=jDMnbeW3F8A

India's Water Revolution #3: From Poverty to Permaculture with DRCSC https://www.youtube.com/watch?v=KtHuIlfyJao

Regreening the desert with John D. Liu https://youtu.be/IDgDWbQtlKI?t=82

> It all started in 1995 when Liu filmed the Loess-plateau in China. He witnessed a local population who turned an area of almost the same size as The Netherlands from a dry, exhausted wasteland into one green oasis. This experience changed his life.

To me it seems that the two main areas to focus on to make a real contribution are:

1) Make modern agriculture ecologically harmonious.

2) Make ecologically harmonious agriculture more automated.

Obviously these converge.

(My feeling is that we would be happier with lots of little farms rather than a few large ones, but I'm not interested in debating that particular point. I care about ecological harmony. Future generations can work out the economic forms provided we don't crash the system before then.)

That plant tape tech he links to at the end is incredible.

That idea has been around for a while. I'm aware of similar tools for small farms.

(0) https://www.johnnyseeds.com/tools-supplies/transplanters/pap... (1) https://paperpot.co/

Your statement that 'Vertical farming for non-vegetables (or fruits) is likely dead-on-pitch.' is true with current technology. But there have been significant developments in this space in the last few years. See this work done by Zach Lippman at CSHL. https://pubmed.ncbi.nlm.nih.gov/31873217/ Early flowering, dwarf varieties such as those described along with renewable energy (thinking of electricity rates going negative in Germany recently) may significantly shift the economic viability of growing fruits indoors.

Really nothing to say about this article... but the "ag" abbreviation is strangely disturbing. Do people pronounce that in speech? If so, how?

What makes you say it’s disturbing? Having grown up near several Ag colleges, I never questioned it.

I cannot avoid reading it as silver (the chemical element), especially when the first letter is uppercase. I had a hard time understanding some sentences of the article due to that. My mind just refuses to parse "ag" as "agriculture". It's just that I'm not used to it.

EDIT: In your case, I cannot help to think about your colleges as very shiny, kitsch buildings.

> Do people pronounce that in speech?

I've heard it frequently in politics. (It's also an acceptable Scrabble word.) Pronounced the same way it is in the word agriculture.

Chapter in a James Bond book "The Man From Ag And Fish", that is the Agriculture and Fisheries Ministry.

It is pronounced the same way as the beginning of agriculture.


Yes farmers say it in speech. Pronounced ag. Like the first syllable of agriculture.

It is a weird abbreviation. Laziness precludes me from typing agriculture too much.

Pronounced "Aaahggg"

I believe there's python documentation about this.


I'm going to go out on a limb and say that SV tends not to understand most of the non-tech markets it aims to disrupt, and instead gets occasional success by either completely ignoring/sidestepping regulations and protections (ex. Airbnb for short-term rentals) or by throwing so much investor money at undercutting incumbents that the natural market dynamics are temporarily skewed (ex. WeWork for office leasing). Innovation in most industries tends to come from people with deep expertise and team skillsets that extend far beyond software.

Your comment helped me realize my alternate opinion - what I think is healthy about the SV ecosystem.

Changing the term, "natural market dynamics" to "regulatory captured markets", we get:

> or by throwing so much investor money at undercutting incumbents that the [regulatory captured markets] are temporarily skewed

And, taking my sibling comment from @simonkafan, let's change his term "well-functioning industry" to "regulatory captured industry":

> I feel that SC startups are like a horde of locusts flying over a [regulatory captured industry], pretending to improve it positively but all they do is destroy the competition

Consider, for a second, Uber and Lyft. Was the taxi industry well functioning? Did it have natural market dynamics? Definitely, assuming you acknowledge the predicate that the industry players are effectively cartels who control regulations.

Also consider that Uber is unprofitable, has no real path to profitability, pays its drivers poorly and overcharges users. How is that healthy? It seemingly has failed in every possible way as a business yet people continuously prop it up as something to be envied or respected.

One thing that is so often left out of this conversation is that Uber was making a profit in rides when they only had the black car service and Lyft hadn't entered the market yet with regular people driving their private cars. It's hard to tell how much of this is inherently unprofitable as opposed to different companies trying to price each other out of the market.

This makes me wonder what even would be a natural or desired state of any market like this. I don't want companies to have agreements on bottom prices they aren't gonna go under. At the same time this dynamic clearly is fueled by the desire and promise to turn a profit once the competitor is gone. We don't want to set a monopoly in any market. On the other hand why aren't we seeing this in older industries? Why aren't airlines trying to drive each other out of business on price?

If Lyft can enter the market easily, then the market is a commodity market. A commodity market has very few returns. In fact, you can predict that it is a commodity market because of the existence of taxi monopolies. There is no barrier to entry to the taxi market: just get a car. Since there's no natural barrier to entry, they had to create a regulatory barrier to get profitability. Uber making a profit pre-Lyft is indicative of nothing; they are in a commodity market, so their returns will be low single-digits.

Airlines are notoriously bad investments over long-ish time scales, so that may not be the best example.

I think for as much as people talk about businesses undercutting a market to capture it and then profiting from a monopoly position, very few businesses ever make it past the undercutting stage.

As wrong as the efficient market hypothesis is about certain things, a great many industries actually do exist in a sort of perpetual state of trying to balance capturing market share with low prices with staying profitable.

Your point makes sense. But, it largely chooses to be unprofitable (outside of corona times).

I’m not sure that it pays its drivers more poorly than the taxi industry, and I’m not sure that users overpay. Do note, though, that both of those points imply “healthy economics / markets” (dynamic pricing on both sides).

I remember taxi drivers when I was a kid rarely complaining about getting underpaid; they seemed to feel that they were entrepreneurs or entrepreneur-adjacent and generally made enough to get by. Almost as if they were real contractors with the ability to negotiate for wages.

But that doesn't forgive using regulatory capture to prevent anyone else from benefiting from the same things that helped you, that's just pulling the ladder up after you.

Uber did fix the problem of people who would happily make money driving a taxi being unable to because of regulatory capture. They made much worse the problem of drivers making a living wage, and when the sugar rush finally ends and they ultimately have to raise prices you know that extra margin isn't going to go to make the drivers whole and bring their wages back up to what a taxi driver used to make. It's going to go into Uber's investors' pockets.

Although personally I think Uber will just go the way of WeWork once people realize they own (close to) zero vehicles, have (close to) zero employed drivers, and lose money on every ride. What happens after that? Well, public transit has been gutted already so I guess... um... scooters?

The profitability claim doesn't have a lot of merit for me. Uber is already priced at a level that many people don't accept, to become profitable they will likely have to charge prices that would drive massive sections of their userbase away. The assumption that somehow that they will be able to charge much more and lose no users makes no sense to me.

Where would you go? You don't have a car, don't really like public transit. Taxi fairs are higher. People will pay more and some will leave the net effect would be higher profits short term.

Uber most definitely doesn't overcharge users, the amount of hidden expenses from driving a vehicle are absurd (taxes fees repairs tickets insurance gas loan)

Uber service is profitable, they're just spending a lot of money on R&D for new stuff.

> Was the taxi industry well functioning?

Uber and Lyft do not seem to be in favor of a well-functioning industry either. The main strategy of Uber seems to be spending money to get a monopoly in the market and then increase prices and become profitable. How is that a well-functioning industry?

We forget that SV business are business that want to earn money. Regulatory captured industries and markets are not the problem, the problem is that those business aren't in them.

I don't want to sound like an old man yelling at playing kids but sometimes I feel that SC startups are like a horde of locusts flying over a well-functioning industry, pretending to improve it positively but all they do is destroy the competition - with the help of VC funds that allow them to survive for years without a positive cashflow.

Not sure I'd call the pre-Uber taxi system well-functioning, and in a lot of cases, Uber and Lyft have increased competition.

Because for some time now SV has been more about money and business than innovation.

VCs for the most part want a massive return on investment to fit their business model and won't settle for niche businesses that promise only reasonable profits, no matter how technically innovative they may be or what they might lead to down the road. On the other side housing prices are through the roof, so good luck affording a garage/basement to bootstrap something in unless you're already indoctrinated into the VC/FAANG culture and successful within it, in which case you're naturally bias towards perpetuating more of the same.

For this reason SV is a software-to-software-integration (not software in general IMO) innovation hub. Software scales easily, requires no physical property to build things on and can return massive profits. SV excels at building software to accelerate other software. Building software to accelerate physical situations or specific communities that don't exist in the valley requires expertise that must be acquired outside of the valley and doesn't thrive in the valley.

In ag terms SV is a monoculture of software-to-software integration solutions that generally doesn't play well with other crops, with all the efficiencies and drawbacks therein.

I'd add that these successes only happen in markets that are so stagnant and moribund that they're easy to disrupt. Hotel and especially taxis strike me as those. Every trip in a regular taxi I've had in the past decade has been followed by a thought of "never again."

Agriculture is a hyper-competitive low-margin market, and I have never gotten the impression that this industry is stagnant or incompetent. Those are the areas where disruption by outsiders isn't going to work.

I don't disagree with your comment in general. But, just like everywhere else, ag is a very wide umbrella. There are niches which are ripe.

For instance, it is telling that you can drill deep into the earth, pump a liquid to the surface, process it, transport it a significant distance to an expensive port, pump it into giant floating tankers, sail those around the world, suck the payload back out and pipe it to additional processing plants, pipe it across a continent, refine it into relatively finished product, pipe it hundreds of more miles to a depot, pump it into tankers, transport it to underground tanks, then pump it back up to sell to customers. You can imagine how much a product would cost per gallon. Then, as a customer, walk inside to pay for the product you just pumped and snag a gallon of milk. Guess which gallon is more expensive?

Put the milk back on the shelf and drive to your neighbour's house. Walk around to the back porch where they sell their milk out of an old fridge. Um, no, they don't sell it. That would be illegal. They just happen to have a fridge full of milk that people keep stealing. But the thieves have a conscience. They keep stuffing money into the tin can next to the fridge.

Because in the U.S. it is legal to sell a product with a big label that says, "WARNING: THIS PRODUCT WILL KILL YOU." Unless that product is milk.

Plenty of people have figured out how to sell milk cheaper than gasoline. But they are restricted by regulation.

You may well agree with the protections those regulations provide. But it might also look to you like the same problem as taxi cartels.

Right, but that isn't a conspiracy but rather because raw milk is risky: research says 1 in 6 people who drink it get sick, and of those who do 20% had serious complications.


As opposed to, say, cigarettes? Which are legal.

Your link does not actually support your claim that 1 in 6 people who drink raw milk get sick from drinking the milk. If anything, it indicates that only 1 in 6 drinkers get sick at all. There doesn't seem to be a control group (or any real kind of science here), but if 1 in 3 people who don't drink raw milk get sick this might merit some follow-up studies. Perhaps raw milk lowers your risk of illness.

It would be silly to make a statement either way from the information contained in your link.

right. we all know how "reliable" those studies are...

Guess which gallon is more expensive?

Around here, they're about the same. So I'm not sure what point you're trying to make.

Really? His point is that milking cows is easier than oil extraction and refining.

Cows are not without their own costs. But it does seem that it might just be possible to find some room for disruption here.

Correct. Taxi logistics had been stagnant for a long time. Human dispatchers and operators.

Farming is constantly improving yields and optimizing. It might look stagnant to outsiders but it's a meat grinder.

In one of my college classes, all of the "business" majors looked down on the farm kid who'd show up to class with blood on his boots during hunting season.

Then he'd give presentation after presentation / example of how his farm is using GPS coordinated UAVs to survey the crops and drive the tractors, and all sorts of other technology deployments that the kids in boat shoes had zero comprehension of.

Taxi logistics had been stagnant for a long time. Human dispatchers and operators.

Taxis in Chicago had an app even before Uber and Lyft. I think it was called Halo, or something like that.

You forgot "money", you know how we are going to solve this problem, by throwing a lot of money at it. Case in point - Softbank

Sometimes, being a total outsider will see things other people cannot.

But I think the effect is pronounced in agriculture because of the stereotype that rural = Trump voter = poor moron. I think that few people in SV have a true grasp of the amount of intelligence, electrical and mechanical domain knowledge, and vast wealth in agriculture. There are gobs of farmers worth $10m+, it’s just that in their communities, farmland is higher status than Ferraris or Arcteryx, so they stick with their F-250s and overalls and invest their money into more farmland, growing ever wealthier and more efficient.

Not to mention that a big tractor or combine harvester costs more than an average Ferrari or Lamborghini (unless you happen to buy a Lamborghini tractor I guess ;)).

Looking now I could find a used New Holland combine from 2005 for $71000 USD. That money gets you a used 2000-ish Ferrari 360.

And the combine makes you more money after you buy it.

If you're not leveraged to the hilt paying for that combine!

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