kzrdude

I wonder what Demis Hassabis thinks about this. I thought he cared a lot about mathematics.

You should read his comment from wednesday here: https://terrytao.wordpress.com/2026/09/07/finite-time-blowup...

And yes, at one point it gives you right about them taking advantage of him. He sat down for an interview and found later that they just used the "best clips" out of it for an OpenAI ad.

Well OpenAI took one step on the back foot at least, withdrawing from sponsoring this math hackathon event https://xcancel.com/danintheory/status/2098125701782372640

My university has an agreement with Microsoft copilot. We can log into copilot in many ways, and it's only if you log in the correct way that you get the "Enterprise Data Protection" copilot version, with a green shield symbol. There are many ways to go wrong here!

It seems like we're not supposed to care about the code quality then? I guess that's the compilers argument. But I'm not ready to give up the code just yet.. These LLMs don't even have a stable interface, they change every few months in how they interpret our prompts and tasks.

They have plausible deniability on that one: not making any profit.

Is that something we have credible evidence for? Do they serve a better model when artificialanalysis (the benchmark site) is making the requests, and so on?

Interesting, and if you don't mind, where do we put humans (and superhumans like Magnus Carlsen)? I think they have a heavy evaluation function and do shallower search.

There's this whole discussion going on about agents being more independent now. They don't follow instructions so well, they continue until the problem is done (sometimes too long), they don't ask the user for feedback.

That is a kind of benchmaxing: they are made to complete benchmarks tasks and one-offs well, and no longer work well in tandem with the user.

Regardless what you call it, it's a divergence between what the power user wants and what the model developers want, I think.

But the very fact that you go to "chatgpt.com" and write to them; "Dear Diary, today I thought.."; there is no reason they would not receive and process your data, unless explicitly promising not to (which also requires us to trust them).

The fundamental rule in this case is that if we offload our data to a cloud provider we can assume they read it, if they can, unless they promised very clearly they will not.

And just a day later we have the tech report available for V4.1: https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash/blob/...

Isn't one problem that it's hard to determine what equivalent compute is, for CPU search vs a neural net based engine like AZ or Leela?

V4 Flash was one of the big events of this year, and its already retired and replaced by V4.1 Flash.

The main claim, that "they do not seem to have any mathematicians capable of understanding what they put out", was also corroborated by Sebastien (OpenAI) who explained they don't have any experts on Navier-Stokes.

The Chinese labs have released interesting papers to accompany their releases too, especially DeepSeek and Kimi. This improves their standing among a few of us, who really like to see and read the papers with details about what they have changed and how their models work.

Now you are using the legislators' rationale that money is being made so we need to protect that, forgetting to weigh the cost for the community.

It's got 'max' and 'ultra' but I don't see a setting for level-headed.

On the contrary, a level-headed summary that gathers information from all the different sources is necessary.

Depends on how much risk they are willing to accept. What is strange here is that it's clear that OpenAI is both a service provider and a competitor to mathematicians. It almost reminds me of Amazon which both hosts external merchants and competes with them, sometimes copying their stuff. Similar but not the same.

We come back to the rule: "The cloud is just someone else's computer".

The way for people or companies or universities to control their data and information is to keep it on their own computers.

Let's see what happens. In contrast to this whirlwind of math that's going on right now, the millennium prize rules require publishing in a reputable journal and 2 years of waiting time to establish that the proof has been accepted by the community. So nothing happens in the short term.

The construction is that there is one file you need read and verify, the challenge file. If you've verified that file and trust that your lean compiler works correctly, the proof will be correct.

That file should be https://github.com/openai/NavierStokesAndEuler/blob/main/Com... in this case (286 lines).

I don't see how it would be negative PR. If anything, the love these breakthroughs and use it in their PR campaigns.

This "fefferman options c and d" thing sounds damning but that's nothing. Let's assume the forelaid proof is correct. Then option C or D is the only way to win the prize, those options are the only ones that solve it. The whole thing is just "prove well behaved" or "prove singularity", where the latter is the case that turns out to be the case.

If you don't trust the other party, then it doesn't matter how the checkbox is set. The fundamental rule, IMO, is don't send precious or secret data to a third party.

Some of the involved people are dramatically naive if they believe they can simultaneously hide themselves from OpenAI while sending their arguments to a cloud service owned by OpenAI.

While I support their argument - push for stronger data and privacy protections from OpenAI and similar - it is naive to believe we can have privacy while sending our data to third parties. It's clearly better to be safe than to be sorry here. Well, clearly better in terms of privacy. In terms of the maths gold rush, who can say what's better, that probably favours those taking more risk.

What's the clarification? It seems like they were aware of each other's work, eventually, but Stadlmann published (a preprint) first.

There are various forces at play here, academic honesty requires them to disclose any inputs regardless of license or ToS circumstances.

While common sense reminds us here that if you send your data to an external entity’s computer, you are no longer in control of said data. The lines have blurred here clearly over the last decade, but that should have made the theory yet more clear to everyone involved: your data will be vacuumed up unless you keep it sealed. Use your own computer if you want to be in control.

Using EUPL is mostly a way to signal that you want to use a license grounded in EU law, not in US law. That's at least the way I've looked at it.

I had a different strategy, starting with lots of connections, so I found Russia and China first. Albeit playing with a lot of mistakes.

I just did my first second-hand inference and it was satisfying. Turkey borders Bulgaria and Greece, don't know which is which, but those two both border North Macedonia, which was a correct guess.