Prime Gaps at Most 186 | Hacker News Reader

Prime Gaps at Most 186

github.com

51 pointssimonpure11 comments

Open articleView on HN

Context - according to Wikipedia [0], just a few days ago someone posted a proof of a bound of 240 [1].

[0] https://en.wikipedia.org/wiki/Twin_prime [1] https://arxiv.org/abs/2608.31126

Related ongoing thread:

GPT-6 Astra - https://news.ycombinator.com/item?id=49554643

(see also https://news.ycombinator.com/item?id=49555621 from there)

blegh, both the code comments and readme reek of verbose claude puke.

> The registered Lean build passed without errors or warnings.

thanks. i like it when my proofs do not error.

Edit: just saw this was published under openai, yikes.

What is this AI-generated gobbledygook actually trying to say? That there are infinitely many pairs of primes p,q with q=p+186?

The first equation under "The result" seems to be saying they found an infinite sequence of primes whose density is forever greater than 1/186, which doesn't match my understanding of how prime density works.

it's saying that there is an infinite list of pairs of primes p,q such that |p-q|<=186. the syntax they're using is arguably bad, but standard

I think it is talking about the existance of close primes, rather than the densitiy of close primes.

??? this is unreadable lol. i feel like i understand the math behind it but the writing is so unintelligible that i dont even care about the results.

> ??? this is unreadable lol. i feel like i understand the math behind it but the writing is so unintelligible that i dont even care about the results.

Who cares, partially no one understands advanced math, and even fewer will one we replace mathematicians with AI.

This just proves that OpenAI is awesome, the value of brains is dropping precipitously, and Sam Altman needs to be made into a trillionare.

who knew the primary utility of pure math would be to juice pre-ipo valuations ¯\_(ツ)_/¯

Keep going

Reply on news.ycombinator.com