Top
Best
New

Posted by m-hodges 4 days ago

I vibed a proof of Conway's conjecture(overreacted.io)
https://github.com/gaearon/conway-refinement#why-i-think-its...
271 points | 294 commentspage 2
renyicircle 4 days ago|
The Claude output in the first one-shot counterexample attempt is hilarious. I hate its writing most of the time but this stuff is next level deep-fried slop.

> And the control column confirms the resonance-necessity conjecture empirically: break the skeleton alignment and the joint kernel dies at the constrained window, exactly as the transversality heuristic predicted.

> The den has air in it.

> Drift fuel exists.

1attice 4 days ago||
Discovering that poetry was cognitive compression is alone one of the latent findings LLMs unlocked.

You can so easily imagine this shit being read in a 90s slam poetry coffeehouse. Trust me I was there

nonameiguess 4 days ago|||
I had a manic friend in college who later became a Emmy-winning television writer. One night over a quarter century ago, he stayed up about 24 hours straight doing nothing but writing, completely free form, some of it prose, some structured rhyme, some of it dialogue with stage directions. He pinned it to his dorm walls like it was wallpaper for a week, then we took it out to the field and burned it. The surviving paper that didn't burn made up bizarre strings of words that sounded much like this, which he retyped and called it poetry. It totally worked.
fwip 4 days ago||||
This sort of thing is why STEM students need more humanities classes. The idea that poetry is semantically denser than prose is, like, obvious, to anybody who cares about poetry.
simianwords 4 days ago|||
I never got the point of poetry, even after giving it an honest shot many times. It felt far too compressed to express any thing. I could not separate it from how I would appreciate music. Maybe its because I don't start from the same base as others.

You are right, maybe some humanities classes would've helped or I'm just wired this way.

fwip 4 days ago||
It's pretty similar to music, yeah. Or perhaps music can be similar to poetry. Just this morning I was re-reading the lyrics for "Pale Green Things" & "Woke Up New" by The Mountain Goats. There's a sort of specificity in intention that the form tends to end up dictating, that can really carry a rich emotional context or metaphor more directly into the listener's brain.

(Of course, prose is sort of like a superset of poetry - it's not a hard dividing line. You may find especially poetic lines in your favorite longform works, especially in good fiction books.)

I do think that many English teachers do a poor job of encouraging a love, or even understanding, of poetry. One of my highschool teachers was in love almost with categorizing the forms of poetry - sonnets and sestets and villanelles, and seemed to think that if we learned the rules about what lines had to rhyme or which syllables to stress, we'd enjoy poetry. I hated it! But then the next year, even against this backdrop of learned avoidance, my teacher was so in love with the ability of poetry to communicate - the unfolding of a few words into meaning like a flower in bloom - something clicked. He focused almost not at all on the structure of poetry, but rather on finding works that spoke to you. I'm still not a lover of poetry, even now, but I have poems I appreciate.

1attice 4 days ago|||
Not what I was actually pointing to, but thanks; I'll pass the critique along to my old MA thesis advisor. ;)
djeastm 4 days ago||||
>one of the latent findings LLMs unlocked

Either I really misunderstood high school English classes or this is what poems have always been known to do.

renyicircle 4 days ago|||
That makes perfect sense. Using language in unorthodox ways to convey very specific concepts that only make sense to you and sound like bullshit to others.
1attice 4 days ago||
I never expected a machine to be better at articulating complex thoughts with a spare number of semantic coordinates, but then I never expected to find out that my poetry was golfing in the latent space
creamyhorror 4 days ago||
Absolute bad-metaphor-laden slop, in a dramatic writerly voice. These LLMs are trained on too much pretentious writing.
doctoboggan 4 days ago||
Nit, I know, but I don't think they are trained on too much pretentious writing so much as they are RLHF'd into it because that writing impresses people.
rlue 4 days ago||
> Take all the numbers you have so far. Then, “spawn” a new number in every gap between the numbers you already have (crucially, “to the left of all” and “to the right of all” also count as “gaps”). Apply this step forevermore, and you’ll get surreal numbers.

I'm not a mathematician. Can someone explain to me how this approach gets you beyond the rational numbers?

Also, this was formatted as a blockquote, but as far as I can see, this blog post is the only instance of this formulation online.

sebzim4500 4 days ago|
You can have an infinite number of numbers on both sides. E.g. you can ask for a number between 0 and 1,1/2,1/3,... (which gives you an infinitesimal)
doctoboggan 4 days ago||
> a sort of epistemic performance art project.

Agreed, and it's a wonderful piece of art. I look forward to seeing the actual publication and reaction from the math community.

patcon 4 days ago||
Deeplinked reply from Prof Vincenzo Mantova[1], who is reviewing results: https://news.ycombinator.com/item?id=49761718

[1] https://eps.leeds.ac.uk/maths/staff/4058/dr-vincenzo-l-manto...

deiptx 4 days ago||
Is the author suggesting that understanding has no value? because this is what I could gather from reading this.

Edit: Another depressing fact is that he generously paid to LLM megacorps while piggy backing on human help for free and in the end calls the proof his or LLM's.

danabramov 3 days ago|
Is a YouTuber speedrunning a game by exploiting an integer overflow bug suggesting that gaming has no value? No, they're just having fun with the medium.

To answer directly: I think understanding is the primary value. And I have reasons to hope that my "work" here can ultimately contribute to a better understanding: https://news.ycombinator.com/item?id=49761718. However, there are also other reasons to do things than producing value. You can do things for fun, to show they're possible, to ask meta questions about the field.

Regarding calling the proof "mine", I've replied to that here: https://news.ycombinator.com/item?id=49755885.

deiptx 3 days ago||
Following your analogy it is not a very fair game if it allows winning for those having access to latest exploits. Playing against skiddies who bought the hack kind of ruins it.

I read your comment and my edit was inspired by it. Your article only mentions nameless mathematicians, without really giving credit.

danabramov 3 days ago||
>Your article only mentions nameless mathematicians, without really giving credit.

I explicitly give credit to the papers the proof builds upon. Quoting from the article:

>The proof was only possible thanks to the many existing results from [References](https://gaearon.github.io/conway-refinement/#/references). In particular, [A factorisation theory for generalised power series and omnific integers](https://doi.org/10.1016/j.aim.2024.109513) by S. L’Innocente and V. Mantova has played a crucial role in the proof.

The project README (https://github.com/gaearon/conway-refinement) also states:

>Over the years, Berarducci, Pitteloud, Pommersheim and Shahriari, and L'Innocente and Mantova developed increasingly strong results about exactly that. This proof builds on all of that work and claims to finish the refinement conjecture.

It includes the same list of references: https://github.com/gaearon/conway-refinement#references

Regarding who I've talked to over email, I wasn't sure I want to "out" them explicitly in the article due to the article's likely polarizing nature, but, since he already engaged in this thread, I can confirm that this person was Prof. Mantova (who is a co-author of the primary paper that the result I published rests upon, and who I did explicitly credit in the text of the article and in the references).

You can read Prof. Mantova's own reflection here: https://news.ycombinator.com/item?id=49761718

I have not had an extended conversation with anyone else.

>Following your analogy it is not a very fair game if it allows winning for those having access to latest exploits. Playing against skiddies who bought the hack kind of ruins it.

You misunderstood the point of my analogy. The point was not that it is a game in which you "win" or "lose" and thus I have "won" or something like that.

The point was that "obtaining a true proof without understanding it" doesn't imply "understanding has no value" any more than "completing a game in an unintended way" implies "playing has no value". You stated a false implication. That false implication is what I am highlighting with the analogy. Understanding has a societal value; doing something weird for the fun of it has value to the person doing it; sometimes people do things with no value at all, and that's okay too. Each individual act is not a statement about what kinds of things have value.

Personally, I do not see mathematics as a game of "win" and "lose". If the broad value is primarily in understanding (which I think is correct), then my "contribution" doesn't make much dent in that. However, it might, as Prof. Mantova mentions, help show the way to an actually useful proof that improves collective understanding. That would please me beyond merely having had fun. In any case, I've enjoyed the experience of it in part due to the absurdity of it. Maybe it requires a certain kind of sense of humor.

Regarding the models, as far as I know, all of the models I have used are currently broadly available. Yes, they require subscriptions so access is not democratized yet. There's nothing I can do about that. I think it's unreasonable to require that someone playing around with LLM's must use the inferior free models.

deiptx 2 days ago||
I had not read the proof or any part of it because it is LLM generated. Most mortals won't for the same reason. It looked to me as if Prof. Mantova stepped from the crowd to get some credit which was not given. But maybe he did not care for it in the first place so you are right, it is not for me to judge.

And about the "value" I think we picked different meanings for this word:

For me for something to have value means it can be used for something else. In this sense you cannot pick what has value. There are people with deeper understanding of the subject than yours yet they were unable to use it to come up with a proof. You had no understanding but you had succeeded. Therefore understanding in this particular case had no value. You did not have to state it, it flows as a natural conclusion.

You can also say that being good at chess or catching all the pokemon has a value for you or for some society. This is true but it has to do with self-actualization rather than dealing with circumstances.

I also do not see how you can say you did not win anything. This proof will be a great addition to your portfolio which in turn will serve as a leverage while searching for your next job, even if it never comes up explicitly. It is a dream for some folks to be talked about on hacker-news. Good for you that you can treat is as if it was nothing, and I do not mean it as an irony.

bastawhiz 4 days ago||
These are indisputably good results all things considered. But I have to wonder whether a more scientific and hands-on approach to working on the material would have been better. When I vibe code, I don't just hype the LLM up and tell it to keep going. I interrogate it, I ask it to back up and replace its jargon, and I force it to be accountable. It smells to me like a lot of the circling could have been avoided (even without domain expertise) by just enforcing processes. Even just keeping the Lean more up to date would have likely saved tokens: it doesn't matter if it took longer each week, since the total runtime mostly wasn't the bottleneck.
danabramov 3 days ago|
>I interrogate it, I ask it to back up and replace its jargon, and I force it to be accountable

I do that too when working on software. Here, I did a little bit of that, but it is much harder when I have almost no domain knowledge (aside from understanding the statement of the conjecture), since at each point the LLM might trick me anyway, and it would probably take me a year to understand the concepts enough to tell when what it's saying doesn't make sense.

>Even just keeping the Lean more up to date would have likely saved tokens

That's the conclusion I came to by the final week! But don't underestimate how much Lean needed to be written in the first place to even "catch up" with the reference papers. I had no option to keep it up to date when I started.

cubefox 4 days ago||
It's quite the irony that in the end he says

> Although the current generation of models is trained to complete tasks rather than to enrich our understanding, and today’s AI companies are misaligned with the goals of the mathematical community, I hope that with time we’ll find ways to use these tools in harmony with human research.

while citing "A Severe Misalignment of AI in Mathematics" [1], which condemns exactly the thing he is doing himself: Mindlessly producing theorems without a corresponding human understanding of the underlying proofs.

1: https://mathandai.org/

danabramov 4 days ago|
Perhaps you and I read this document differently. This is not a “famous problem”, I am not a “major AI company”, and I am genuinely interested in increasing mathematical understanding. To the last point, here is a comment from a mathematician who co-authored the paper that my proof is largely built upon: https://news.ycombinator.com/item?id=49761718

The happy case here is that my obtuse proof leads to a concise and illuminating mathematical proof, which is exactly my hope for the endeavour. I think this could then be a positive example of AI/human and amateur/professional collaboration.

What would a positive example look like to you? What do you think the manifesto argues for?

cubefox 3 days ago|||
Okay, at least the comment from the mathematician sounds positive.

Positive example: Someone is 1) independently interested in a particular conjecture, 2) he lets an LLM prove and explain it, 3) he ends up fully understanding the LLM proof and is then able to phrase it in his own words.

GPerson 4 days ago|||
It is a famous problem!!!!!!!!!! You literally understand nothing about the culture you’re stomping all over.
danabramov 3 days ago||
I may be wrong here but my impression is that surreal numbers in general are kind of a niche area that hasn’t enjoyed a ton of interest. Additionally, this specific conjecture did not have any “prizes” attached to it and was not on any list of famous problems I could find. It’s even difficult to Google. What precisely do you mean by it being a famous problem?
GPerson 3 days ago||
It’s a problem that someone could describe in a talk and attribute to a prominent mathematician and which would have resulted in significant career opportunities for having solved in 2021. You don’t understand this because you don’t care that you’re degrading the tiny chance people had to actually learn mathematics.

Edit I apologize to Dan Abramov for venting my frustrations about things outside of either of our control and unfairly using him as a punching bag. He seems to be an intelligent person and I hope he continues learning mathematics using whatever tools he sees fit, including AI. It was wrong of me to do this and I will take a break from this website for one week.

tuesdaynight 4 days ago||
Damn, I was not expecting Dan Abramov when I read the title.
FiatLuxDave 4 days ago|
This year, LLMs have been involved in a number of interesting proofs of conjectures. But that is not even half of mathematics. Has anyone tried to use an LLM to generate a mathematically interesting conjecture, on the level of Conway's refinement conjecture? If so, what happened?

With all the talk of mathematicians possibly being obsolete, I'm wondering where the future conjectures that future LLMs would prove might come from.

nphardon 4 days ago|
I would not be surprised if an LLM could have generated this conjecture, since it's such a natural extension of the integers, does it hold for surreal integers is not a big leap. It's totally predictable / probabilistic and thats what they do at a high level.

My read on the current sitch is that the drama is more around the humans behind the LLMs, like the OpenAi *people*, who are dishonest, thieving, technocrats, stealing human mathematicians work and claiming credit for it. Since the LLMs are only trained on finite amounts of data on the internet I don't think anyone expects them to replace mathematicians in a serious way. But surely the field is changing dramatically, people will adapt.

More comments...