Levent Alpöge

OpenAI afirmó que no reclamará el premio de un millón de dólares.

Mientras tanto, la disputa de prioridad con el investigador de Anthropic, Levent Alpoge, y Tristan Buckmaster de NYU sigue sin resolverse y su resultado relacionado sobre Euler forzado aún aguarda revisión por pares.

Pasamos a Amber con un hallazgo de seguridad que toca de cerca el debate sobre el control de modelos.

AI Weekly reportó esta mañana que un modelo de investigación interno de OpenAI produjo su propio jailbreak, generando instrucciones adversariales que eludieron sus propios filtros de seguridad.

And just last week, Ars Technica reported that Meta ads pushed apps that were nudifying real teens.

Many of the ads were placed through one of Meta's official advertising partners based out of China.

The ads were often ones that let clients digitally alter images of children using AI.

The report also identified a China-based app that allegedly allowed users to share materials some users had flagged as child sexual abuse content.

The company, however, says it does not intend to claim the Millennium Prize.

Now, this came with some big controversy because, uh, right around the announcement, there was this dispute that came to light with New York University mathematics professor Tristan Buckmaster and his collaborator, an Anthropic employee named Levent Alpoge, and their separate project, which Levent was doing with Tristan just in a private capacity.

This was not like an Anthropic-driven initiative.

They were just working on this problem using multiple AI tools, and they developed related results for a related math problem called Euler equations.

Um, that's a big deal 'cause it means that there's like novel physics to be discovered behind the breakdown, right? So once you figure this out, now what it says is there's something we don't understand about the way that this thing, this thing all works.

So the drama has been that these two mathematicians, Tristan Buckmaster and Levent Alpoge, which again, pretty sure I butchered his name, who works for Anthropic, were working on these problems already based on the work of some other mathematicians who had sort of pointed them in a direction, and they were using ChatGPT to do their work.

Um, they had found a solution for the Euler equation, um, which is basically a version, uh, which is essentially the same formula but with no friction, so no viscosity.

Um, and they were cleaning up the results of that work for publication.

Two mathematicians had been working the same front for almost a year.

Tristan Buckmaster at New York University and Levent Alpoge, a mathematician employed by Anthropic.

They were using models from both houses.

OpenAI and Anthropic, paid for by Buckmaster out of his own research funds.

The...

It, it was a dramatic couple of days really, and, um, it, it starts with two researchers, Tristan Buckmaster and Levent Alpoge.

They were sort of quietly working in the background on Navier-Stokes but also on, uh, a close cousin of it and a sort of steppingstone problem towards Navier-Stokes, the Euler equations.

Now Word of that had got out.