Not really, they’re just lying to try to fleece more investors

I mostly agree with you, but AI cannot “act autonomously”. You prompt it, and it acts based off of that prompt.

Think of it like doing the laundry: I load my clothes in and “prompt” the washing machine to do my laundry.

When the washing machine switches from the rinse cycle to the spin cycle, it’s not acting autonomously, it’s continuing to execute my prompt.

Is AI capable of doing something you didn’t expect it to do? Yes. Does that mean it’s acting autonomously? No.

I would also argue that the “power” of LLMs is vastly overstated. A large part of that is for the exact reason you say tho, people with zero knowledge or skill are playing with fire; And when they get burned they blame the machinery, not the operator.

I mean, I guess if you wanted your documentation and reports to be worthless slop bullshit that could work.

Companies do those things for a reason tho, and they’d want them to be accurate.

Every story you read about how their AI is “so powerful it’s scary” or where they characterize the AI as “going rogue” or some other dumb shit like that is manufactured in an attempt to impress investors

I can’t believe I have to keep explaining this to people, lmao

First of all, this is a math equation. Math is what computers do, they are essentially really fucking good calculators.

That math problem hadn’t been solved before, that’s true, however, if you actually read the excerpt you just shared, you’ll see that the equation was solved using preexisting theories and formulas. Ergo, humans had already manufactured every individual piece of the puzzle, and had used them in other contexts that the AI could be fed and then reuse.

Saying that AI came up with the solution isn’t the full story. That’s like saying that AI created a shitty piece of slop art. It did no such thing, it plagiarized a bunch of work from other people and essentially made it into a collage.

When MC Hammer (rather, his producer) sampled Rick James for “Can’t touch this”, did he become the bass player? No, that was still Rick James, they just took his song and altered it to make theirs.

Gotcha, well that’s a different issue (one I believe expands beyond self-driving cars to all new cars that are way more computer integrated than they should be).

Getting hacked is a good reason to be skeptical tho, it’s probably the slickest way to kill someone and make it seem like an accident.

I’m a good driver, and AI is fucking stupid.

The only way you could possibly convince me to even get into a self driving car would be if my only other option was to get into a self flying helicopter.

Also, just to nitpick, the self driving car can’t “go rogue”, that would imply that it has consciousness, sentience, an understanding of what it’s actually doing, judgement, etc. a bunch of shit it doesn’t have. Self driving cars can and do “malfunction” frequently, but they can’t “go rogue”.

Don’t let the techno fascists’ constant anthropomorphic descriptions of their fuckware chat bots confuse you.

What tedious work?

In order to answer that question, you would first need to know how to make AGI (this is not known to anyone).

In order to know how to emulate human intelligence, you would first need to understand how human intelligence is constructed (this is not known to anyone).

In order to understand how human intelligence is constructed, you would have to understand how human cognition works (this is not known to anyone).

In order to understand how human cognition works, you would need to understand how the exact mechanisms upon which human brain works in the first place function (this is not known to anyone).

I could continue, but you should get it by now.

An LLM cannot do that either.

An LLM can only ever rearrange what is put into it and shit it back out. You’re never going to get any knowledge out of it that wasn’t previously fed into it (and even then it hallucinates and doesn’t know shit because it’s literally just code).

We do not know how to make an AGI, ergo LLMs do not “know” how to make an AGI (and in fact, don’t actually know shit at all).

They’re not escaping users control, they’re fucking computer programs.

People are using them irresponsibly.

If you accidentally run someone over in your car, the car didn’t “escape your control”, you just did a shitty job of controlling the car.

It’s not the same thing.

(Fuck “AI” btw)

The limit for what?

The bottom line is that AI is both more expensive and less productive than humans for at least 99% of tasks

Wouldn’t something like that be machinery the saw company already owns?

Does slop not involve expensive machinery?

I agree, and the people at the top of the industry pushing this 100% want a slave.

They don’t know how to build the slave they want though, so they’re attempting to leverage manufactured fear about their fuckware in order to lower the value of real workers in order to push us closer to slavery.

We can’t stop them from telling these lies, but we can certainly help people know not to buy their lies.

That’s not conservative my man, it’s delusional

What information are you extrapolating this from? Just vibes?

Being below a certain employee count reduces worker protections

It does not change the cost of the slop (which is currently artificially tiny)

They can’t do AGI tho. They’re nowhere near being capable of even conceptually thinking about AGI.

They talk like they are to scare their investors into pouring more capital into their furnace, but AGI ain’t happening.

I don’t think we need to concern ourselves with that, we’re nowhere near an AI that can think or feel.

There is no path from LLM to sentience.

The concept of a computer thinking or feeling is an incredibly distant ‘in the future plausibility’ at best.

Yeah, the point of my explanation was to be simplified.

Both of your links ultimately back up what I’m saying tho.

The computer isn’t thinking, it’s doing a math equation.

Coding is like insanely well suited to LLMs by nature tho, that’s why it can actually do that.

Code is a language with stricter definitions and less verbosity than a spoken human language. Things only have one definition in code, one spelling, one syntax, etc.

There’s an element of artistry to human coding, but it’s more of a science than an art (compared to spoken human language anyhow). It’s easier to predict.

There are other promising “ai” technologies, like radiological image recognition, aren’t the same technology as LLMs, but the companies who make the LLMs constantly conflate their technology with that of more useful technologies in order to obfuscate their relative uselessness.

Even the coding models are being localized. There’s absolutely no reason to have a massive AI datacenter unless you’re making a shit LLM, and even then, the only reason to have that is to commit financial fraud essentially.

EDIT - I learned a bit more about “agentic” AI since you had mentioned it earlier, and what’s actually going on behind the scenes is wild.

So basically, an “agent” is a computer program that you give a task to. The agent then attempts to accomplish the task by engaging in a loop until the task is accomplished.

The structure of the loop is 3 steps. 1) Prompt an LLM - The agent asks an LLM what the first step in the task you asked it to accomplish would be. 2) The LLM provides a plausible answer, and the agent then attempts to carry out these instructions. 3) If the given instructions don’t work, the LLM appends the prompt to say “I tried step #1, it didn’t work, what should I try next?”. If the instructions work, the agent appends the prompt to say “I took this step and it worked, what’s the next step?”. Then the loop repeats.

So, this would work exceptionally well for code because the computer can Test and verify the code before continuing… but that’s just about the only use case for the technology

A good way to visualize this is to imagine yourself as an agent. Let’s say you need to do your laundry, that’s your task. So you ask the LLM what the first step is, and you do what it says no matter what. Then you go back, append the prompt, and prompt the AI again. This cycle repeats until either 1) you finish the laundry, or 2) you destroy the laundry, or burn down the house, or injure yourself blindly following the LLMs directions (this second outcome is the statistically probable one).

Would you do that? No, probably not. You can think, so you would know which instructions to attempt, and which to ignore. The Agent cannot think. The LLM cannot think. The agent strictly follows directions like any other computer program. The LLM is attempting to construct an answer that appears plausible, token by token. It doesn’t matter if it is plausible, or possible even, the LLM isn’t trying to give you the correct answer, it’s trying to give you an answer that appears correct.

Well your main assertion was a wildly incorrect definition of sentience.

You go on to say that (and I’m paraphrasing) “sentience is in the eye of the beholder”, which is also wrong.

There’s also the implication that you do think computers are sentient, but you haven’t typed that specifically; I know you said you didn’t want assumptions, but this is a fair inference. Someone who wasn’t emotionally invested I the idea of chatbot sentience wouldn’t push back so hard with basically 0 facts to back them up.

Your final assertion is that the Wikipedia article doesn’t immediately contradict you, which again, it does in the first sentence. You can see my previous comment for that information tho, I’m not retyping it.