I'm a bit worried about a YES resolution that just constructs a solution to some diophantine equation or something.

Is the $100k for the inference budget in today's currency value? The currency should also be stated, e.g., USD.

@TomBurns nominal dollars

It is very entertaining to see the opinions of all the people who think this won't be possible by 2030

Is it correct that YES resolution also requires the AI to do novel research, not just write a paper that could pass as research?

@zsig Yes

80% is completely absurd. I am sure that AI will be capable of writing parts of such papers by then, but currently models cannot even write correct but essentially trivial papers.

@LocalGlobal 10K at 70%?

@LocalGlobal This feels to me a bit like "Those sea lions that were trained to build motorcycles a few years back can source parts and draft designs, sure, but their welding technique is still terrible!"

Without a deeper model of what drives AI progress, the phrase "currently models cannot" is just not predictive of anything.

@LocalGlobal https://sakana.ai/ai-scientist-first-publication/ Sakana

AI recently were able to produce a paper that passed peer review for an ICLR workshop

Sakana AI

The AI Scientist Generates its First Peer-Reviewed Scientific Publication

This particular example is an indictment of the state of peer-reviews and scholarship in ML more than anything else

@jgyou I agree with this. The paper should not have passed the first round of peer review. I was most impressed by the fact that it could autonomously construct such a paper, graphics and all. This is another case where the exaggerated claim serves as a good prediction of near-future capabilities.

5 years is a very long time, even if the exponential improvement we have consistently seen for the past 6 years doesn't hold the whole way through.

@Haiku I think these responses might just be overestimating the standards that are expected of a workshop paper.

Wow the original bet was made at a 25% implied probability.

@MaxMorehead Check your biases folks, Daniel Litt is neither dumb nor an AI skeptic.

FWIW I think his implied odds are too low too.
But the gap is wild!

@jgyou to be clear, I'm saying the market is too high, since even Tamay thinks this is (slightly) less than 50%.

the ppl that have historically made mana on ai on manifold are those that think agi is coming soon, for some reason

@Bayesian This is a bit like saying in 1999 that people who have historically made money in the stock market in the age of the world wide web think that Internet is going to change everything. The Internet did change everything, but a lot of the stocks were overvalued.

Manifold hasn't existed for that long. I think strong prediction performance across many decorrelated domains (or over a long period of time) is a greater signal to me than doing well on Manifold. And while people who predicted the current state of AI deserve a lot of additional weight in their opinions, but I don't see any particularly strong meta-level reason to be convinced that the miscalibration of the past will continue in the future. So we are left to think mainly at the object-level at the state of AI science and the arguments for different AI timelines.

I do think that world-changing AI is quite possibly coming soonish, but probably not by 2030 (though I also think this market is a lot weaker than some "AGI" definitions).

@MaxMorehead Yeah I think also there's a general factor of "how fast is AI progress" where the answer for the last few yeras has been "faster than expected". There may be a big vibe shift where progress stalls. For instance @SemioticRivalry predicted that something like this would happen. AI optimists have had a consistently good track record for the last few years. Definitely. But this could flip. It's hard to know.

I think a "yes" resolution could come but could be unsatisfying. For instance perhaps the AI just comes up with some counterexamples to some famous conjectures. The Annals publishes that sort of thing.

People who bet YES, do you mind sharing your logic? and math background?

@mathvc One of my looser bets, basically I expect math to kind of just be solved (to >human level) by reasoning models scaling up + modestly impactful new techniques by 2030. It's something that requires 0 real world interaction and can be verified post-facto as true/false.

No knowledge of Annals or modern frontier math research.

@DavidHiggs I am not so sure a paper that brings a sufficiently new contribution can be easily verified post-factum, unless the AI model manages to translate its proofs to formal language that can be checked automatically. Checking Wiles' proof of Fermat's last theorem took years

@mariopasquato I’d half expect formal proof to be the default, and human readable to be the required translation, but haven’t really thought about it deeply.

@DavidHiggs I think this us a misunderstanding of what mathematicians do. Proof is not the main product of a math paper, rather it is human understanding of math (with proof as important tool/constraint). Historically, proofs that were valid, but didn't increase human understanding (e.g. 4 color theorem) were viewed as something between underwhelming and almost useless.

@AIBear nonsense. Completely disagree.

There are tons of absolutely central results in mathematics that are understood by nobody: Classification of finite simple groups etc