'Shit in, shit out', AI is coming for agriculture, but farmers aren’t convinced

I am a backlog, hear me score.

Sounds like they aren't actually market rate if the labor market is rejecting the prices at those levels.

Could you elaborate on "subsidies don't scale"? In the US, farm subsidies are a huge chunk of our budget and, to the best of my understanding, help keep food prices low. I'm not informed enough to know if it's an inefficient solution though.

You can't anything your way to miracle profits without destroying society in the process.

I suppose destroying society for profit is top of the barrel for people like you.

Would I be correct in assuming that your name comes from 'The Point'? What a classic!

I also appreciate your point about the reality of agriculture, I think too often people miss just how narrow the margins there are and why farm life has always been so financially unstable. It used to be the weather that could make or break you, and it still is, but not world commodity markets, the price of fuel and fertilizer, and trends in a number of areas can all do it as well. Add in the top-down pressure to keep prices low and bottom-up pressure to target migrants and... what a rough mix.

Polyculture has two very significant advantages.

- Because you're pulling multiple crops off a field instead of one, the value of those crops is significantly higher.

- The crop is much more resilient. The different crops will have different tolerances to drought, disease, insects, wind and other disasters that commonly affect farming.

When farmland was cheap you could increase your profits per employee by buying more land and bigger machinery. But these days farmland is expensive and farmers are concentrating more on increasing yields per acre. In my area farmers spend > 10x as much per acre on inputs like fertilizers and pesticides than they used to in the 90's.

Polyculture will be hard and expensive, but if it lets you double profit per acre, farmers will do it.

> combine harvester not requiring a driver

Already pretty much the case. They may have a driver as a monitor/troubleshooter but for planting and harvesting it's all optimized down to the inch with GPS guidance steering the machinery.

There is a solid equilibrium in a commercial monoculture and artisanal and subsidised polycultures for insurance. The transition will never be easy. But at least the groundwork is there.

Two of my friends who farm already have self driving planters and harvesters, controlled by GPS, that lay down different kinds of seeds and fertilizers depending on location within their fields. The farmer is there to fix things.

> The economy of the scale favors monoculture

I mean it does, until it doesn't and you loose all your crops/soil/water.

Monocultures degrade the soil, and you need to put more and more input in to make sure grow. Because there is no ground cover between crops, wind takes away your soil. If you mess up/mistime your pesticide/herbicide input, you'll loose all your crop. The more you grow, the more the soil turns to sand, requiring more fertiliser input, which mean more costs.

If you take the yield per m2 of an allotment, and compare it to a really high quality agricultural field (ie Lincolnshire fens, or some prime land in the USA.) the allotment will have a much higher yield with less input. I don't have the figures on hand, but you'll need to turn to the bearded hippies at the henry doubleday to get precise numbers.

If, through automation, you could have three or four crops in the same field, providing cover all year round. you could then start to keep your soil, because the wind cant get at it (crop dependant). you can grow sacrificial crops for pest control, meaning less pesticide input. This means more worm, which will stop flash flooding (more water permeability without ploughing)

But all that requires high speed "pick and place" weeding/harvesting machines. Those are going to be complex and expensive until scale kicks in (think how complex tractors were, or threshing machines)

The future we’re already in should be characterized as cheap complexity. As an example, in manufacturing one offs were cost prohibitive, now it’s close enough to mass manufacturing cost that in many cases there is no point taking that extra step. Also people are increasingly concerned with food provenance, I would love to have a local automated greenhouse growing hairloom vegetables.

They'll do it if it's profitable. More output per acre is a pretty big motivation.

Grad students will work on the problem for the ecological benefits. Then big ag will scoop it up for the profit motive.

This is effectively my "dream" job for semi-retirement at the moment. I don't need to make top dollar any longer, but I crave doing something "real" that actually matters and I work for/talk to people who actually make useful things for humanity. I still need a decent income to pay for toys, hobbies, and time off work until I hit retirement age.

A contracting IT job trekking between a few farms and monitoring them remotely sounds pretty great to me! And right up my alley in terms of skillsets and interests.

The way they are treating cattle as pure machines to optimize production I can only hope synthetic meat will come soon.

Thanks for the definitions. Doesn't make for a better existential generalization from the more specific to the more general term.