thereitgoes456

My mistake, thank you for catching.

It's interesting that the Stirrup scores are somewhat close to the specialized harness scores. Either Stirrup is quite excellent or is it that harness doesn't matter so much after a point?

Yes, that is true, you might be right. But I think such, because the moat in their wrapper layer feels very thin and does not make sense in a world full of 20+ wrappers around similar models. It feels better to "only" have to deal with Ant/OAI (and other pure wrapper cos). Competing vs. Google, Meta, and every Chinese lab on earth, every one using their own bespoke wrappers, is much scarier.

Additionally: competitive frontier = more commodified = high valuations under more scrutiny.

(EDIT: A child comment has refuted the AA score comments, AA runs using their own harness.)

One of the new benchmarks is FrontierCode 2.0 by Cognition. AA seems to have accepted Cognition's benchmark scores without questioning, but they should have looked more closely.

On this benchmark, Anthropic's models are run using Claude Code; OpenAI's using Codex, and Gemini is run using... Cognition's own Devin (not Antigravity). Thus its score is much lower than it should be, knocking ~3 points off its AA score. All open-source models are also inexplicably run using Devin, damaging their scores as well.

This may even be intentional, if Cognition has beef with Google, since Antigravity was built using Windsurf's technology. But aside from that, it is to their advantage to make Anthropic/OAI look like leaders in the AI market. They are a company known to lie and exaggerate again and again, so I have trouble giving them benefit of the doubt.

Author seems to be a serial liar who takes AI work and “publishes” it. His website says he’s an AI architect, speaker, blockchain expert, author (he self-inserted into his book, with AI images of himself) and experienced in dozens of disparate disciplines. A section on “Students” at Princeton despite never being a professor there, and none of the linked papers mention him and their years don’t match.

> But I feel like I am asking the obvious question that has not been addressed: What if I do not want to spend my professional life doing these activities?

Then don’t.

Unfortunately, you don’t have the right to your profession not being changed from under your feet. Nobody does.

Right now it’s mathematicians’ time. Once, it was pilots, accountants, phone operators, weavers, farmers, and many others.

That’s true but not relevant. I have absolutely no sympathy for the rubes who thought FTX was a good place to put their money, but that doesn’t make the fraud any more OK.

Great work. I’m bored of Krillion because the AI judge sucks and the animations are WAAAAYYYYY too long. This feels like an upgrade in every way. Will add to my daily rotation of snacks, thank you :)

What an insulting article.

> There is a popular idea that substantial wealth belongs mainly to entrepreneurs.

This is probably because 24yo college dropouts are achieving billion dollar valuations after 12 months of work. Their secondary sales are worth more than following all the advice in this article will be for your entire life.

(To say nothing of NVDA, SpaceX, or other big tech acquiring them and making their billions liquid, despite no moat or profitability, just because they have so much money and need to spend it on something.)

No matter how much “grit” you have, you’re still at the mercy of such people, as e.g. the engineers at Windsurf were, who worked super hard but their founders sold to Google, walked away with hundreds of millions, and gave the employees nothing.

Of course, it depends on your meaning of “substantial”. Successful engineers have great wealth too, enough that they can be very happy and buy anything they want. But pretending the two levels are comparable is silly.

Why would the tip be from someone within OpenAI? If they knew it they would have surely omitted that result, there’s no way this is a desirable outcome for the capitalists either.

Presumably the tip would be from someone who’s familiar with the area but doesn’t want attention. Which is unlikely to be someone in OAI.

Tao didn’t write that and isn’t part of the group that did, he’s just reposting their letter.

Can you be more specific on how you expect AI to do this?

If you’re thinking by an independent, motivated do-gooder, TurboTax costs around $100 for most people, and few people are going to use some untested solution to save that small amount of money, vs. the scare of being fully liable for one’s own tax mistakes. FreeTaxUSA is even cheaper.

AI cannot explain chess moves it comes up with in an elegant way. What makes you think it will be able to do so for math?

I am asking about approaches, not results.

Nobody serious would deny this is incredible progress, but GP is making an unmotivated leap to RSI, so I respond to that framing. It’s an interesting argument to be had but I suspect few of us have standing to say one way or the other.

(Gesturing at the number of problems solved, or the number of years the problem was open for, isn’t an argument.)