Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Amateur armed with ChatGPT solves an Erdős problem

scientificamerican.com

Post reply on HN
Print view View original

601–607 of 607 posts

Re: Amateur armed with ChatGPT solves an Erdős problem

#601

Post by themafia »

pylua wrote: post #549
themafia wrote: post #542

Earlier quoted context omitted.

Yes. A lot of brute force methods are the most inefficient means of solving a problem. Language Models are a sign that our current human information infrastructure and access methods are completely wrong.

Arguably, since everything resolves to an axiom, isn’t the solution in any training set ? The order and combination is what makes it special. Is current human information access methods wrong, or do we just synthesize data in a way that is inefficient for this sort of problem solving ?

> since everything resolves to an axiom

This isn't true. There are solutions that are beyond apparent reason and logic. This is what a "breakthrough" is.

> The order and combination is what makes it special.

Given an infinite amount of time a team of monkeys will produce Shakespeare. Is that "special?" Perhaps we should leave some room for _how_ those combinations happen and how efficient they are.

> Is current human information access methods wrong

They are wrong. The largest search company is also the largest advertiser. I'm surprised that anyone either fails to apprehend this or pretends not to.

Re: Amateur armed with ChatGPT solves an Erdős problem

#603

Post by gxt »

Anybody with access to these models can challenge it to test the hypothesis that spacetime is a 4d viscous fluid with the speed of light being spacetime's sound barrier, mass relating to viscosity, blackholes being cavitation bubbles, Hawking radiation our perception of surface tension, and gravity just being a pressure differencial? Thanks

Re: Amateur armed with ChatGPT solves an Erdős problem

#604

Post by bsza »

simianwords wrote: post #580
bsza wrote: post #508

Earlier quoted context omitted.

It is not “absurdist” to call out a baseless claim that doesn’t take into account over half of humanity, a percentage that will grow even further once investor money inevitably runs out. If your response to that is to wave away more than 4 billion people, then you’re not even trying to look like you care about reality, you’re just trying to make yourself feel better with some made-up nonsense. You seem to be under th…

> It is not “absurdist” to call out a baseless claim that doesn’t take into account over half of humanity, a percentage that will grow even further once investor money inevitably runs out. I love the confidence that comes from this claim. You can run open models in your laptop today compare to the best models from 2 years ago. But sure spread your FUd about investor money running out

I have nothing against open weight models, my issue is more with these mega-corporations posing as saviors of humanity. That said, how is your consumer hardware going to out-compete a datacenter when it has more mouths to feed per token than a datacenter? Who is going to give you money to run anything when a machine can do everything you can do?

No matter how you spin it, we humans are now becoming thermodynamically less efficient versions of LLMs. We contribute nothing of value to the system, so economics dictates we have no place in it except as investors. Skill is nothing now, and ownership is everything. So yeah, I'm afraid of the future. Call it FUD or whatever, I don't care.

Re: Amateur armed with ChatGPT solves an Erdős problem

#605

Post by famouswaffles »

imiric wrote: post #175
CSMastermind wrote: post #141

Earlier quoted context omitted.

Worth mentioning, though, that people have already tried running all of them through LLMs at this point. So this is proof of the models actually getting stronger (previous generations of LLMs were unable to solve this one).

> So this is proof of the models actually getting stronger (previous generations of LLMs were unable to solve this one). No, it's not. While I don't dispute that new models may perform better at certain tasks, the fact that someone was able to use them to solve a novel problem is not proof of this. LLM output is nondeterministic. Given the same prompt, the same LLM will generate different output, especially when it i…

This is one of a number of such results achieved only in the last few months with only the last crop of models. They have undoubtedly gotten better in this domain. Saying anything else is just denial. You can run these same problems on GPT-4 or 5 all you want, you'll get nowhere. In fact people did, and you're hearing about it now because it's these crop of models that are getting meaningful results.

Re: Amateur armed with ChatGPT solves an Erdős problem

#606

Post by Eufrat »

anthonyrstevens wrote: post #597
Eufrat wrote: post #172

Earlier quoted context omitted.

> You realize that sounds incredibly naive in the history of mathematics, right? This is after the fact justification. You are arguing that because a thing (number theory) showed practical applications we should have dumped a lot more effort into it. There is no basis for this argument whatsoever; it also seems to involve inventing a time machine. Number theory had no practical applications until the development of p…

>> Number theory had no practical applications until the development of public-key cryptography This is so wrong I don't even know where to begin. Modular arithmetic, numerical integration, pseudorandom number generation, error-correcting codes, predicting planetary orbits (!), etc.

At the risk of asserting the claim was obviously a bit facetious, number theory had the reputation of having very little practical applications and I don’t mean that silly quote by G. H. Hardy.

A lot of applications just required a lot more computing power to be practical. This all starts to happen around the same time (unsurprisingly) and if you’re going to make hay that Reed-Solomon coding was invented in 1960, I think it’s worth pointing the first big use of was on Voyager because the computing power was finally able to make these work. It’s not like people hadn’t started to notice some of this decades earlier.

Re: Amateur armed with ChatGPT solves an Erdős problem

#607

Post by alxndr »

pohl wrote: post #422
WarmWash wrote: post #401

Earlier quoted context omitted.

Language really only exists at the input and output surfaces of the models. In the middle it's all numerical values. Which you might be quick in relating to just being a numeric cypher of the words, which while not totally false, it misses that it is also a numeric cypher of anything . You can train a transformer on anything that you can assign tokens to.

Similarly, none of our comments actually exist as language on Hacker News—just numerical values from the ASCII table. We're deluding each other into thinking we're using language.

"The great enemy of communication, we find, is the illusion of it" —William H. Whyte
Post reply on HN