Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The rise of the robot farmer

theguardian.com

Post reply on HN
Print view View original

81–90 of 187 posts

Re: The rise of the robot farmer

#81

Post by PhasmaFelis »

knieveltech wrote: post #24
clarkmoody wrote: post #12

Earlier quoted context omitted.

The immense productivity gains in agriculture have driven down food costs, which improves the standard of living of everyone. It also frees up those who would have worked farm jobs to do higher-valued labor in the economy.

In theory. In practice, this has lead to collapse of rural economies, rampant poverty, and epidemic drug usage amongst rural populations. As to driving food costs down, that's trivially disprovable. Loaf of bread in 1950: 12 cents. Adjusted for inflation that's $1.26 currently. A cheap loaf of white bread costs substantially more than $1.26. Edit: of course this gets downvoted, because econ gibberish is certainly mor…

A loaf of good-quality sandwich bread can easily go for $5, but cheap white can be had for around a dollar in most places.

Re: The rise of the robot farmer

#82

Post by HeyLaughingBoy »

otakucode wrote: post #79
nradov wrote: post #54

Earlier quoted context omitted.

There's nothing stopping you from planting a garden today. It's not a huge amount of work, and any robots that become available in the next few years will only marginally reduce that labor.

It's not a huge amount of work, but it is a significant time investment. I'm thinking more of a nearly-sealed box that is entirely automated, a family could purchase for a couple thousand dollars, and then it produces food for a few years. It's not so much robots in the sense of things which drive around the landscape that will improve that sort of things, but the work in the 'vertical farming' market. Isolated syste…

You're very unlikely to get a couple thousand dollars of food over a couple of years out of even a large "box," where the box is approximately the size of a bedroom. This is one of those cases where it's cheaper to simply buy the food from someone who has economies of scale on his side.

It costs even less if you cut out the middlemen (supermarkets, distribution, etc) and buy directly from the farmers where possible.

Re: The rise of the robot farmer

#83

Post by PhasmaFelis »

stevenwoo wrote: post #68
zackmorris wrote: post #49

Earlier quoted context omitted.

Hah thanks for that! https://xkcd.com/1425/ Still, image recognition is effectively a solved problem at this point for AI. Maybe there are still problems with notation and terminology that prevent widespread adoption?

Is there an update to this story or is it not true or some other possibility? https://www.wired.com/story/when-it-comes-to-gorillas-google... If you prefer a story not on wired (sometimes it shows a subscription nag for me) - https://www.theverge.com/2018/1/12/16882408/google-racist-go...

I don't think that's changed, no.

My hypothesis, incidentally, is that it's really hard for an AI to tell the difference between human and ape facial features at all; they're too similar. It seems easy to us because a big chunk of the human brain is devoted to face processing. The AI pretty much just knew "primate face," and knew that pale primate faces were usually human, but with a dark primate face it could only guess.

(Which doesn't excuse Google for releasing it in that state, of course.)

Re: The rise of the robot farmer

#84

Post by HeyLaughingBoy »

uxhack wrote: post #4
ihuman wrote: post #2

OP's link goes to a comment. Here is a fixed link: https://www.theguardian.com/environment/2018/oct/20/space-ro...

I think op wanted to point out the comment that was saying how impossible picking grapes would be by ai.

Why does it need to be AI? One of my neighbors has an Italian tractor that harvests grapes. Just takes one person to drive it along the vines.

Re: The rise of the robot farmer

#85

Post by PhasmaFelis »

otakucode wrote: post #44

I eagerly await the introduction of fully automated farming. Not, however, on farms. In my own garage. I see no reason I should simply let a "farmer" purchase the automation gear and then collect profit after wasting tons of energy in transportation. There's dirt, water, and sun at my house. There's no reason a robotic farming situation couldn't be shrunk down and made into something private individuals or small fami…

I think you're underestimating how much space it takes to grow food, and overestimating how much bulk shipping costs.

Re: The rise of the robot farmer

#86

Post by bluGill »

mr_overalls wrote: post #64
bluGill wrote: post #38

Earlier quoted context omitted.

Unlike what you would expect, small robots do more damage to the soil not less. Damage to soil is not linear with mass, A massive robot will do twice as much damage where it touches the ground (wheel tracks most likely, but there are other choices), but it will touch 10 times less ground to do the same work so the total averaged over all the land is much better. For the same reason, tracks generally end up doing more…

This. . . doesn't sound convincing to me. Soil displacement is a function of machine weight AND surface area of tires/treads. Additionally, I'd expect the sharp edges of steel treads to disturb soil much worse than a similarly-weighted tractor with rubber tires. Without doing extensive calculations, my anecdotal evidence is that I've seen bulldozer track "scars" from a single path that are decades old, while the dama…

You say the bicycle healed in a year, but you didn't measure that (this is a guess, but is seems reasonable). It isn't enough to say the grass appears green, is it as thick and growing as fast as untouched grass nearby.

A single track every 30 meters where nothing grows is less impact on the overall field's growth than 30 tracks every meter where things still grow but not as well.

Re: The rise of the robot farmer

#87

Post by PhasmaFelis »

dbcurtis wrote: post #61
the-pigeon wrote: post #19

Earlier quoted context omitted.

Bah. Difficult is the word. When someone says something is impossible is just means they don't know how to solve the problem.

Or maybe the guy that "knows how to solve the problem" is solving some other problem, not the actual problem. I count three wine makers among my friends. None of them machine harvest their own grapes, nor will source machine harvested grapes from other growers. It would be very bad for their respective brand images.

Does that mean they're convinced the results would be unsatisfactory, or just that their rustic image is more valuable than increased efficiency?

Re: The rise of the robot farmer

#88

Post by dougmwne »

nostromo wrote: post #73
rmason wrote: post #9

In Michigan we've already got family farms with a man, two sons and a single employee handling 10,000 acres. Robots driving tractors, trucks and combines would allow that to expand 5-6X. In turn that would lead to a reduction in farmers of up to 80%. This is a trend that has been running for close to a hundred years. Remember at one time a quarter of America's population were farmers and now it's under 2%.

But is the juice worth the squeeze? A single employee for 10,000 acres is nothing. (I'm assuming that the three owners would stay on with or without robots.) That person's salary is a rounding error compared to land, seeds, fertilizer, water, equipment, etc. I think the benefit of automation for commodity crops (rice, wheat, etc.) is almost nil. Labor isn't driving the cost for these crops. However, automation for la…

To get to those labor efficiency numbers we've made a lot of food quality and environmental sacrifices. TFA specifically mentions automated laser weeding instead of herbicide/GMO. Fertilizer microdosing is also mentioned which would greatly help fertilizer runoff pollution. I'm sure you've heard of other problems with factory farming, many of which are labor saving practices.

The cross section of farming, public health and environmental impact can't be reduced to a commodity spot price, at least not in our current economic system.

Re: The rise of the robot farmer

#89

Post by PeterisP »

bluGill wrote: post #38
tda wrote: post #27

Earlier quoted context omitted.

managing giant monocultures with a few giant machines (that may or may not be human controlled) is not what this article is about; it is about the huge potential that smallish robots have for farming. More diverse crops, less pesticides, less damage to soil, less energy use, mixed instead of mono cultures. Really exciting if this takes off, such a win for the environment, the landscape, and ultimately for everyone of…

Unlike what you would expect, small robots do more damage to the soil not less. Damage to soil is not linear with mass, A massive robot will do twice as much damage where it touches the ground (wheel tracks most likely, but there are other choices), but it will touch 10 times less ground to do the same work so the total averaged over all the land is much better. For the same reason, tracks generally end up doing more…

This seems a bit confusing - in the context of agriculture, when I hear "damage to the soil", then that implies (to me) that we're talking about the difference between the long-term soil erosion caused by plowing/tilling/cultivation versus growing different cultures for which the process requires less of such invasive activities and so doesn't erode the soil as much.

The effect of large tires vs small tires vs tracks vs feet is insignificant and irrelevant, it's a rounding error compared to the effect of whether those tires/tracks will be pulling a plow/till or not.

Re: The rise of the robot farmer

#90

Post by ben_w »

dsfyu404ed wrote: post #46
knieveltech wrote: post #24

Earlier quoted context omitted.

In theory. In practice, this has lead to collapse of rural economies, rampant poverty, and epidemic drug usage amongst rural populations. As to driving food costs down, that's trivially disprovable. Loaf of bread in 1950: 12 cents. Adjusted for inflation that's $1.26 currently. A cheap loaf of white bread costs substantially more than $1.26. Edit: of course this gets downvoted, because econ gibberish is certainly mor…

A cheap loaf of white bread costs substantially more than $1.26. Actually that's about exactly what it costs at Walmart. I wanna say it's $1.28 but I'm not 100% sure. Granted that's literally the cheapest loaf of bread you can buy not an average nothing special loaf of bread.

Food prices are confusing. Back in the UK I could get cheap loaves of bread for 55p ($0.71 at current exchange rate), while here in Berlin I see them for €0.59 in mid-range brands for a slightly smaller loaf that works out as $1.08 when adjusted for size as well as current exchange rate.
Post reply on HN