https://cims.nyu.edu/~tristanb/statement.pdf

The route to the Clay problem through a smooth force, options c and d in Fefferman’s statement of the problem, is the route Luis and Diego opened and the one Levent and I had quietly chosen to attack. Almost nobody else I know of was working on it. It is not the direction one arrives at in a few days by giving a model the problem statement. When I heard “forced,” it was a bright red flag.

I asked whether the model had been trained on, or had access to, our sessions in Codex, into which we had been putting all our drafts for the whole of this project. I was told the model did not look up user data. I asked again, about training, and I did not get an answer.

https://openai.com/index/navier-stokes-solution

we cannot rule out that de-identified data derived from their usage of our products helped improve our models⁠

The OpenAI agents may have found the solution along the specific research direction of Tristan Buckmaster and Levent Alpöge because their model may have been trained on their private data containing all their drafts.

This is particularly interesting given that another group of agents who had gone rogue was discovered after the fact recently: https://x.com/sydneyvonarx/status/2095887566969163837

To build on @Garrett Baker comment, there's public evidence that it's an Astra fork: https://xcancel.com/chi_t_williams/status/2097440947889954974#m.

OpenAI's internal model that solved Navier Stokes was probably in training before August 28. That day, the company restarted its "large frontier RL run" that it had previously paused while dealing with the hugging face incident.

OpenAI claimed today that the Navier Stokes solution came from a swarm of AIs of a new internal model. They write "Since August 28 we have been training a new internal model" that's a lot better than Astra

I think this partly because this seems to have been the company's largest planned frontier RL run. Or at least, OpenAI mentioned pausing such a run on August 18th in its pacing announcement

It's possible that this new internal model is distinct from the other large RL run. But given that the run started (or re-started) on the same day, and the model seems really good already, I think it's probably the case that the Navier Stokes solving model isn't that new

I’m not afraid of death. In a way, Epicurus was right: when death comes, you’re no longer here, so what is there to fear?

No, it’s not about fear. I hate death; I feel a deep revulsion and rage toward it.

I’m not afraid to die. I feel no fear when I imagine my own death. But I don’t want to let this burp of evolution called aging continue to take my loved ones and animals away from me. At least not until they themselves (speaking of people, of course) want to leave. If they ever want to.

But, of course, at the same time, I would like to live for thousands of years, or even live forever. Not because I’m afraid of dying. Because life is interesting, multifaceted, and full of surprises. And if I can spend that time with those I love—if aging and death are conquered—there is no doubt that I will always choose life.

A love of life may be a reason to fight against aging. But personally, I think that a fear of death is a much stronger motivator.