The AI flop and the end of the Silicon Valley tech orgy

[Dolan]

That's basically the topic

[Goodspeed]

It'll be at least a decade before we can say AI "flopped". Of course, investors don't do long-term thinking, so they were all expecting immediate payouts..

Reminds me of some guy's idea about restructuring company shares so that there are short-term shares and long-term shares. Long-term shareholders would be committed to holding the shares for at least 5 years or so and they would have disproportionately more votes than short-term shareholders. Sounded good to me, considering the lack of incentive to think long-term is such a problem in corporations rn. Idk if it would hold up to scrutiny though, seems like there would be some issues.

[Dolan]

I'm still hoping for my AI-powered vegs and fruit-picking tractors if I get to have my own farm one day. I mean, I'm not saying it won't have any use. AI seems to excel at averaging down tasks, so it will probably be useful for tasks that any dumbshit could do. Like making a robotic hand reach out and pick a fruit.
Or maybe one day it could even be used for human-assisted surgery in cases that don't require a complicated procedure. And if the AI has a sudden software spasm, the human could take over and just get it finished.
But for tasks that require actual cognition at +100 IQ level, I wouldn't bet much on it ever becoming capable of. Under 100, it's all quite random anyway.

For now it just looks like it's gonna be mostly used as a jacked up search engine, digital blender that can output averaged down visual or textual slop and for some task automation jobs (visual recognition, etc) where a machine can do pixel peeping a lot better than any non-autistic human.

[BrookG]

Not arguing that it is impossible, rather not too simple. Picking fruits varies from low height plants, bushes, trees. Gathering olives requires a different technique compared to gathering walnuts or tomatoes. Also, the robot/AI would have to be able to visually recognise the ripeness and the timing of when to gather. I can't imagine farmers from Crete for example buying expensive AI to gather their olives while they drink raki in the local tavern. Indeed, AI can in theory help with many tasks, however, it will take many years to infiltrate in all sectors. In my city in Greece, this year a handful of auto-checkout machines were just installed in supermarkets. AI market atm feeds within the same corporate circles, exactly where the huge money is.

[Dolan]

Not arguing that it is impossible, rather not too simple. Picking fruits varies from low height plants, bushes, trees. Gathering olives requires a different technique compared to gathering walnuts or tomatoes. Also, the robot/AI would have to be able to visually recognise the ripeness and the timing of when to gather. I can't imagine farmers from Crete for example buying expensive AI to gather their olives while they drink raki in the local tavern.

[BrookG]

Not arguing that it is impossible, rather not too simple. Picking fruits varies from low height plants, bushes, trees. Gathering olives requires a different technique compared to gathering walnuts or tomatoes. Also, the robot/AI would have to be able to visually recognise the ripeness and the timing of when to gather. I can't imagine farmers from Crete for example buying expensive AI to gather their olives while they drink raki in the local tavern.

The video neither proves or disproves my point. It is unclear if the machine can operate equally with any kind of fruit/vegetable/seed. Also, the farms shown are immense, and probably the equipment they have already used is expensive, so it might be not an issue to invest in such a machine, which is not the case in smaller farms. Besides, it is also unclear if this is some pilot experiment or part of a research program.

[iNcog]

I mean AI has the potential of super intelligence but I guess everyone blew their load too eagerly when it came to AI in its current state changing the world.

it's another tool in the toolbox of humanity for now. would be nice to harness that technology for good obviously.

[Dolan]

Not arguing that it is impossible, rather not too simple. Picking fruits varies from low height plants, bushes, trees. Gathering olives requires a different technique compared to gathering walnuts or tomatoes. Also, the robot/AI would have to be able to visually recognise the ripeness and the timing of when to gather. I can't imagine farmers from Crete for example buying expensive AI to gather their olives while they drink raki in the local tavern.

video

The video neither proves or disproves my point. It is unclear if the machine can operate equally with any kind of fruit/vegetable/seed. Also, the farms shown are immense, and probably the equipment they have already used is expensive, so it might be not an issue to invest in such a machine, which is not the case in smaller farms. Besides, it is also unclear if this is some pilot experiment or part of a research program.

The vid was just a reaction to your post, wasn't meant to prove or disprove anything.
It's from a pilot project done in Australia, that started several years ago, so that was before the new and much hyped versions of commercial AI came out. It's probably from about 5-7 years ago, so back then we were seeing AI used in visual recognition software (autopilot feature in some cars). They took that and hooked it up to a robot vehicle with hands that can extend to a visually recognised apple and use a vacuum pump to suck the apple into a pipe that collects them all in a tank.
It probably got better in the meanwhile.

This is one example where I think AI has potential. And it could also provide a solution to any future population shrinkage and unsustainable pension systems. And a counterargument to those neolibs who claim our economies couldn't survive without immigration. I think automation will be the answer to that. Populations could keep shrinking as long as we get increasingly better at replacing the disappearing workforce with AI-powered machines.