Live data from Hacker News

Erdos 281 solved with ChatGPT 5.2 Pro

twitter.com

111–120 of 310 posts

Re: Erdos 281 solved with ChatGPT 5.2 Pro

#111

> no prior solutions found. This is no longer true, a prior solution has just been found[1], so the LLM proof has been moved to the Section 2 of Terence Tao's wiki[2]. [1] - https://www.erdosproblems.com/forum/thread/281#post-3325 [2] - https://github.com/teorth/erdosproblems/wiki/AI-contribution...

[flagged]

I firmly believe @threethirtytwo’s reply was not produced by an LLM

Re: Erdos 281 solved with ChatGPT 5.2 Pro

#112

> no prior solutions found. This is no longer true, a prior solution has just been found[1], so the LLM proof has been moved to the Section 2 of Terence Tao's wiki[2]. [1] - https://www.erdosproblems.com/forum/thread/281#post-3325 [2] - https://github.com/teorth/erdosproblems/wiki/AI-contribution...

[flagged]

Are you expecting people who can't detect self-dellusions to be able to detect sarcasm, or are you just being cruel?

Re: Erdos 281 solved with ChatGPT 5.2 Pro

#114
post #94

Earlier quoted context omitted.

yea, i was suspicious by the second paragraph but was sure once i got to "that’s not engineering, it’s cosplay"

It's also the wording. The weird phrases "Glorified Google search with worse footnotes" what on earth does that mean? AI has a distinct feel to it

And with enough motivated reasoning, you can find AI vibes in almost every comment you don’t agree with.

For better or worse, I think we might have to settle on “human-written until proven otherwise”, if we don’t want to throw “assume positive intent” out the window entirely on this site.

Re: Erdos 281 solved with ChatGPT 5.2 Pro

#115
post #47

Earlier quoted context omitted.

I suspect this is AI generated, but it’s quite high quality, and doesn’t have any of the telltale signs that most AI generated content does. How did you generate this? It’s great.

Their comments are full of "it's not x, it's y" over and over. Short pithy sentences. I'm quite confident it's AI written, maybe with a more detailed prompt than the average I guess this is the end of the human internet

To give them the benefit of the doubt, people who talk to AI too much probably start mimicking its style.

Re: Erdos 281 solved with ChatGPT 5.2 Pro

#117

> no prior solutions found. This is no longer true, a prior solution has just been found[1], so the LLM proof has been moved to the Section 2 of Terence Tao's wiki[2]. [1] - https://www.erdosproblems.com/forum/thread/281#post-3325 [2] - https://github.com/teorth/erdosproblems/wiki/AI-contribution...

This illustrates how unimportant this problem is. A prior solution did exist, but apparently nobody knew because people didn't really care about it. If progress can be had by simply searching for old solutions in the literature, then that's good evidence the supposed progress is imaginary. And this is not the first time this has happened with an Erdős problem.

A lot of pure mathematics seems to consist in solving neat logic puzzles without any intrinsic importance. Recreational puzzles for very intelligent people. Or LLMs.

Re: Erdos 281 solved with ChatGPT 5.2 Pro

#118
post #102

Earlier quoted context omitted.

I think that was Tao's point, that the new proof was not just read out of the training set.

The model has multiple layers of mechanisms to prevent carbon copy output of the training data.

forgive the skepticism, but this translates directly to "we asked the model pretty please not to do it in the system prompt"

Re: Erdos 281 solved with ChatGPT 5.2 Pro

#119
post #94

Earlier quoted context omitted.

yea, i was suspicious by the second paragraph but was sure once i got to "that’s not engineering, it’s cosplay"

It's also the wording. The weird phrases "Glorified Google search with worse footnotes" what on earth does that mean? AI has a distinct feel to it

Dude is swearing up and down that they came up with the text on their own. I agree with you though, it reeks of LLMs. The only alternative explanation is that they use LLMs so much that they’ve copied the writing style.

Re: Erdos 281 solved with ChatGPT 5.2 Pro

#120

Earlier quoted context omitted.

Your intuition on AI is out of date by about 6 months. Those telltale signs no longer exist. It wasn't AI generated. But if it was, there is currently no way for anyone to tell the difference.

> It wasn't AI generated. You're lying: https://www.pangram.com/history/94678f26-4898-496f-9559-8c4c... Not that I needed pangram to tell me that, it's obvious slop.

Good thing you had a stochastic model backing up (with “low confidence”, no less) your vague intuition of a comment you didn’t like being AI-written.
Post reply on HN