[teiferer]

Kids hardly know what a multivariate equation is. Unless you use "kid" to denote 20-year old college students enrolled in a math program which some people do.

The other claim is doubtful too:

> while it require an experienced expert hours of work in Lean 4.

No, it doesn't. If you have an actual expert, it only takes a few minutes.

And besides, isn't this exactly what an artificial intelligence would solve? Take some complex system and derive something from it. That's exactly what intelligence is about. But LLMs can't deal with the complex but very logical (by definition) and unambiguous system like Lean so we need to dumb it down.

Turns out, LLMs are not actually intelligent! We should stop calling them that. Unfortunately, there are too many folks in our industry following this hyped-up terminology. It's delusional.

Note that I'm not saying LLMs are useless. They are very useful for many applications. But they are not intelligent.

[gus_massa]

I studied 2x2 linear equation system in high school at 14 (13?) y.o. It was a technical school with more math and physics, and later specialization (like chemistry, electronics, building, ...). I think in a normal school they study that at 16 y.o.

We also teach 2x2 systems to 18 y.o. in the fists year of the university for architects, medics and other degree that don't need a huge amount on math. (Other degrees like engineering or physics get 4x4 or bigger systems that definitively need the Gauss method.)

[teiferer]

And if you ask one of those medics 5 years later to solve one, the response might make you depressed. (Or just 4 weeks after their maths exam.)

[exe34]

Unfortunately, whenever you try to exclude LLMs from the holy church of intelligence based on what they can't do, you end up excluding a whole lot of humans too.

[teiferer]

That's underestimating human intelligence.

Even low-IQ humans can in principle learn how to use Lean to represent a multivariate system. It might take a while, but in principle their brain is capable of that feat. In contrast, no matter how long I sit down with ChatGPT or Gemini or whatnot, it won't be able to. Because they are not intelligent.

It's a great achievement of the AI hype that the burden of proof has been reversed. Here I am, having to defend my claim that they are not intelligent. But the burden of proof should be on those claiming intelligence. The claim that earth is a sphere was extraordinary and needed convincing evidence. The claim that species have evolved through evolution was. But the claim that LLMs are intelligent is so self-evident that rejecting the idea needs evidence? That's upside-down!

[1dom]

> Here I am, having to defend my claim that they are not intelligent.

It's because all you've done is made a claim without any evidence. Someone pointed out a challenge that most claims about them not being intelligent can't submit any evidence that can't also be met by an LLM.

But instead of submitting any evidence to support your claim, you descended into hyperbole about how hard done by you are being expected to support your claims.

In science, it's okay to say we don't know. The amount of disagreement - even amongst smart people - about if LLMs are intelligent or not, suggest to me that we just don't have universally accepted research and definitions that are tight enough to decisively say.

But you're talking not only like you _know_ the answer for sure, so much that you don't need to support it with evidence or credentials, because those who disagree are obviously just poor victims of the AI hype machine.

Please make sure you pass your knowledge of your LLM discoveries onto the scientific community, you could change the world!

[teiferer]

Did you read my post? The claim is "LLMs are intelligent". And instead if requiring evidence for that, apparently most folks (including you) are fine with just accepting that claim and require evidence if somebody questions this. That's what I'm doing.

It's like a religion. "God exists" is the claim. Nobody needs to provide evidence that this is not the case. LLMs are intelligent" is the claim. Nobody needs to provide evidence against that. In either case, the burden of proof is with the one making the claim.

> In science, it's okay to say we don't know

But that's not what's happening. LLMs are called "AI". You know what the I stands for, right? It's not "artificial we-don't-know-if-intelligent".

[1dom]

No you're not questioning it, you're making statements against it, without evidence, which is just as useless as the statement without evidence.

Like I suggested, like the person you responded to suggested: when science tries to prove or disprove LLM intelligence it generally descends into disagreements about definition or evidence (neither of which you provided).

The reason why no evidence was provided for the original LLMs are intelligent claim is because - if you read through this thread - you were the first to make that claim, and the counter claim.

> But that's not what's happening. LLMs are called "AI". You know what the I stands for, right? It's not "artificial we-don't-know-if-intelligent".

I don't know what you want me to say here - if you want to continue acting like there's some widely accepted and agreed on definition of intelligence that everyone who isn't you is an idiot for not knowing, then carry on.

I don't have a reliable definition for intelligence. Like I said, if you do, please share your finding with science, you could settle some fairly big debates and change the world in a meaningful way.

[teiferer]

My claims:

1. LLMs are widely called "intelligent". Evidence for my claim: The term "artificial intelligence" that is used everywhere. It has its own TLD.

2. There is no evidence that this terminology is applicable. Questioning it faces some variant of "well do you have evidence to the contrary?". Evidence for my claim: This thread.

You are welcome to disprove my claims, as in the scientific spirit that you say you uphold.

[card_zero]

IDK, they look intelligent, like the world looks flat.

[exe34]

As opposed to most humans? Have you tried reasoning with somebody "just trying to do my job sir"?

[exe34]

> Even low-IQ humans can in principle learn how to use Lean to represent a multivariate system.

That's an article of faith. In principle, elephants can fly at least once.

[card_zero]

Only if you're pedantic about it. I find I can arrive at all sorts of absurd conclusions like that by being extremely pedantic.

[exe34]

It's almost as if thinking carefully about words leads to the realisation that words are approximations that are meant to describe, not prescribe.

[card_zero]

If "intelligence" describes LLMs then it isn't doing a very good job.

[exe34]

try the bottom half of the population!