Yeah it's not like the same corporations working on AI are working on humanoid robots to put the AI inside or anything! You've probably got 3 or 4 good years of hammering shingles or unclogging toilets for $10 an hour before an Optimus Robot replaces you and you're back to unemployed!
10-15 years. Assuming a majority brown America can even produce those robots
They will be made in China, the infrastructure to produce that tech has long been gutted and exported out of the US.
The good news is that those robots should be able to rebuild an industrial base in relatively short order.
Doubt it, from what I have seen of the technology it lacks the higher thinking necessary for true autonomy.
What we call "A.I" is just a large language model trained on immense amounts of human generated data. It doesn't think, it creates composite responses based off of available data sets.
You could it to pilot pilot a robot that picks crops or runs an assembly line but it always breaks down the second it requires proper thought and reasoning to do something.
The researchers developing this stuff have explicitly stated that we still haven't figured out how intelligence works, let alone how to replicate it.
What we call "A.I" is just a large language model trained on immense amounts of human generated data. It doesn't think, it creates composite responses based off of available data sets.
You could it to pilot pilot a robot that picks crops or runs an assembly line but it always breaks down the second it requires proper thought and reasoning to do something.
The researchers developing this stuff have explicitly stated that we still haven't figured out how intelligence works, let alone how to replicate it.
That's the state of the art from about 18 months ago, things are moving rapidly. I would not count on the idea that the only method for AI being LLM based forever. The use of current tech will accelerate the arrival of new tech that will be able to surpass these limitations, and this has been happening for the past half decade.
Every time skeptics think we have hit a wall, some new method or approach comes out and completely upends the board. The exponential gain in function has accelerated, not slowed.
Every time skeptics think we have hit a wall, some new method or approach comes out and completely upends the board. The exponential gain in function has accelerated, not slowed.
I have heard this before, they used to say the same thing about the space race back in the 70's and that we would have a functional moon base and be terraforming mars by now.
None of it came true and everyone who understood the limits of the technology saw it coming decades in advance.
Technology doesn't perpetually improve, history is rife with examples of periods of technological regression. And we are overdue for one.
None of it came true and everyone who understood the limits of the technology saw it coming decades in advance.
Technology doesn't perpetually improve, history is rife with examples of periods of technological regression. And we are overdue for one.
Shitty example because they cancelled the space program so there'd be more money to give to niggers.
Have you actually had to work with these things?
Because I have, both on the end user and dev side. They are not smart enough, they are not even close to smart.
Nvidia is leading the charge on this shit and yet they still can't get their own A.I models to produce stable drivers over a year after their flagship products release, and that's with the models being trained on data from the best programmers money can buy.
Because I have, both on the end user and dev side. They are not smart enough, they are not even close to smart.
Nvidia is leading the charge on this shit and yet they still can't get their own A.I models to produce stable drivers over a year after their flagship products release, and that's with the models being trained on data from the best programmers money can buy.
People who are incapable of prompting AI to produce results as good as or better than what most entry level employees can produce are gonna be the first to be laid off. Of course, frontier models don't even need prompting anymore.
I will repeat the question:
"Have you actually had to work with these things?"
So far everyone I have met who says that these things are good enough has never had to work with them or just uses them to generate porn pics.
I have direct experience working with these things and they find ways to fuck things up even in very narrow and well defined environments. They code about as well as jeets do.
"Have you actually had to work with these things?"
So far everyone I have met who says that these things are good enough has never had to work with them or just uses them to generate porn pics.
I have direct experience working with these things and they find ways to fuck things up even in very narrow and well defined environments. They code about as well as jeets do.
To answer your question: Of course I work with AI daily. Everyone who is still employed in a job indoors not involving a hairnet in America works with AI daily. It's good and it gets better every week. Does it occasionally do stupid shit? Of course, just like humans do stupid shit. Does it ever phone in sick or try to start a union? No. That's why it's gonna replace everything.
EDIT: Made my tweet less annoying
>Is your go-to argument always pretending that you're a domain expert at everything? You've done it three times just in this thread and it's annoying.
I'm not pretending to be an expert on everything, I'm stating that my IRL experience goes directly against what you are saying and asking what IRL experience you have had that proves your assertions.
I asked multiple times because wouldn't answer the question.
>To answer your question: Of course I work with AI daily.
In what way? Have you had to use AI to write code?
>Does it ever phone in sick
More often than you think, we had severs go down all the time to the point that management got scared and restarted old programs due to be replaced with AI.
>or try to start a union?
No, but the companies who run the services change terms all the time and once enough of the industry is dependent on it the prices will skyrocket across the board.
I am not trying to insult you. I just start having doubts when my lived experiences go directly against what someone is telling me.
I'm not pretending to be an expert on everything, I'm stating that my IRL experience goes directly against what you are saying and asking what IRL experience you have had that proves your assertions.
I asked multiple times because wouldn't answer the question.
>To answer your question: Of course I work with AI daily.
In what way? Have you had to use AI to write code?
>Does it ever phone in sick
More often than you think, we had severs go down all the time to the point that management got scared and restarted old programs due to be replaced with AI.
>or try to start a union?
No, but the companies who run the services change terms all the time and once enough of the industry is dependent on it the prices will skyrocket across the board.
I am not trying to insult you. I just start having doubts when my lived experiences go directly against what someone is telling me.
Follow
@GoonPatrol @caekislove @brokenshakles @Kalogerosstilitis2RevengeoftheJunta
>once enough of the industry is dependent on it the prices will skyrocket across the board.
Not to mention that AI model collapse will happen once AI generated data massively exceeds human generated data to the point that it will start feeding off of it's own data, leading to worse output.
@xianc78 @GoonPatrol @caekislove @brokenshakles @Kalogerosstilitis2RevengeoftheJunta AI is just another trust the plan two more weeks story as far as I'm concerned
So much this.
People don't understand just how much of a problem degeneration is in technical fields. Just copying something without a high level understanding of it can cause irreparable problems.
People have had to deal with this long before A.I.
This is why so much software gets scrapped and has to be remade from scratch, which is only possible when there's enough competent people to do that, something A.I will reduce the supply of.
I know for a fact that there are high level organizations running systems that handle critical functions where there is no one on the team who knows how the system fully works.
People don't understand just how much of a problem degeneration is in technical fields. Just copying something without a high level understanding of it can cause irreparable problems.
People have had to deal with this long before A.I.
This is why so much software gets scrapped and has to be remade from scratch, which is only possible when there's enough competent people to do that, something A.I will reduce the supply of.
I know for a fact that there are high level organizations running systems that handle critical functions where there is no one on the team who knows how the system fully works.
