If AI’s primary function isn’t curing cancer like they said it was going to do, then it should not be utilized.
For jobs like that, human life is cheaper than AI.
You cant build a stable civilization on the foundation of capitalism. All of us are on the same path, its just some are further down the road.
Today us, tomorrow you.
Because AI is built by billionaire fuck-o’s who can only think of problems in their direct vicinity and what do billionaire fuck-o’s hate the most… doing their jobs.
Thus AI writes emails, hires people, fires people, etc.
The idea that there are people picking up their trash, mowing their lawns, or putting their lives at risk to make said billionaire fuck-o more money, doesn’t even register to them.
Us “normal” folks only occur to them when we become a problem.
I think it’s time we all accept that a terrifying amount of people have no perception of accurate reality
Because those jobs actually involve thinking and problem solving, which AI cannot do.
Its called “killing the entire middle class” and its way more profitable that all that stuff you mentioned.
Ooh i think I’ve heard it called “killing the lower class” too. You have to get past the “scapegoat immigrants” part though.
Because business is the most important thing in the world.
They’re too busy making CSAM for the rich pedos.
The same reason that self-driving cars don’t work very well. Shut us way more complicated than people realize because they do it all the time and don’t think about the nuances.
Significant part of my work is building robots. This is the correct answer. And amazingly, the thing that’s making robots actually capable of folding your laundry, which exists by the way, is machine learning and LLMs. But most folks will not be able to afford this, but they could afford to make their own.
You know the idea guy, who doesn’t want to do the actual work? But takes full credit when it’s done? AI is that guy’s best friend.
Is that kinda like the guy in the group project whos decided they wont do any work but will gladly accept the A when the sucker who chose to be group leader stayed up all night doing everyone else’s part?
This week’s Tom Scott episode on Nebula* is interesting on this subject.
Long story short: folk from a university in Essex are working on creating a robot that can pick strawberries, but it (obviously) turns out that it’s waaaay more complex than people think. There are untold different variables that humans just work around without having to think about them, like whether there are fruit inside the plant, obscured from view. A robot has to be programmed to be able to figure shit like that out.
Sure, it could easily run along the line just grabbing everything, but the plants in enormous growhouses can yield fruit for six months a year, so they need to be treated carefully. _ *it’ll be on YouTube next week
Simple answer: Because we’ve written AIs that work in the digital world, and mines, oil fields, toxic chemicals, and ocean plastic all exist in the real world. If you build a robotic body or arm or some way to do what the people in those places are doing, the AI could control it and do the job.
This is the one potential upside with things like Tesla Optimus. With a sufficiently general purpose bipedal humanoid robot under AI control, that robot COULD work oil fields and mines and toxic chemicals and the like.
However I suspect the miners, oil workers, waste handlers, and ocean cleanup technicians would rather keep their jobs, hazardous or not.
Machinery already replaced nearly all the jobs in places like coal mines (in the west). It did increase the safety for the remaining workers, who are mostly machine operators now.
Turns out that a huge-ass loader does more useful work than a LOT of humanoid forms. Optimus is about dreaming of never having to deal with servent classes. Its the modern dumb waiter.
Yup, and to be more exact the reason that this is the case is that training large models requires a LOT of training data.
We have a repository of text and video, the Internet, but we do not have a similar repository of training data for robotics.
That data is being created and robotics models are being trained but it is significantly more expensive when your training data isn’t mass produced for free* by people on the Internet.
*ish
Because we don’t have AI. We have algorithms putting words after another or pixels next to each other.
Better question is why we don’t use robots with specific algorithms to do those kinds of tasks.
We could if we wanted to. Not necessarily right now, but we’re pretty damn close. If all the damn billions we pour into the data centers putting words and pixels into order went into robotics, we’d have a world where no one had to get their hands dirty or risk an injury at work.
That would also make billionaires billionaires. But that would not stomp the regular folk down, and that’s the catch.
… It’s not “putting pixels next to eachother”. It’s figuring out how to filter noise to get it closer to what the prompt describes. Literally starts as noise, like an old TV that isn’t tuned to a channel. And then it applies a filter that it calculates will best get that noise closer to the tokens of the prompt. And then does it again and again and again
Regardless! Even if that investment went into healthcare, or a universal basic income…
Maybe we invest it into robotics to automate stuff. Or maybe we invest it into people.
Investing in people would be neatBut that will never happen until it is required to prevent collapse
Better question is why we don’t use robots with specific algorithms to do those kinds of tasks.
but we need a million different robots with the same algorithm or the data they send back to the company will be useless
Tbh, I figure there are robots for all that, they just aren’t LLMs and usually cost more than what folks are asking for the same job in some cases.
Unfortunately there kinda… aren’t. Robots to handle those tasks, that is. There absolutely can be, we just haven’t built them because the C-Suite has been terrified since 2011, that the rest of us would realize they are as useless as Forbes reported that year.
The main reason they want to take over automation of mental tasks is that we already can, fairly easily, automate single manual tasks. They have been able to automate away 99% of middle management and the C-Suite since 2011, when an algorithm that had been running for two years proved to be correct 99% of the time, when the human executives turned out to be correct in their conclusions only 68% of the time. They want the mental tasks automated so that no one realizes that they are just stealing everything they can from us, because they realized they are replaceable.
Can you share the source? I missed it at the time and would love to read it
