Tristan Buckmaster’s Statement on Getting Scooped by OpenAI on the Navier-Stokes Problem

Tristan Buckmaster, a professor of mathematics at NYU, has provided his perspective on the recent controversy surrounding OpenAI and the Navier-Stokes problem. In his statement, Buckmaster recounted his interactions with employees at OpenAI regarding the controversy.

Buckmaster asked OpenAI when they had initially sent their first prompt. However, it took some time for OpenAI to provide a direct answer. Eventually, they agreed that the prompt had been sent in the past few days, after information about their work had reached OpenAI.

Buckmaster then questioned whether the model had been trained on or had access to their sessions in Codex, which contained all drafts for the project. OpenAI assured him that the model did not look up user data. When asked about training again, Buckmaster received no answer.

OpenAI provided Buckmaster with two proposals. The first was to post their Euler result, and for OpenAI to post its Navier-Stokes result the following day. The second option was for Buckmaster to write a paper presenting the Navier-Stokes result while acknowledging that an internal OpenAI model had resolved it. Sebastien Bubeck, who works at Anthropic, insisted that Buckmaster's co-author, Levent Alpöge, be removed from authorship.

Bubeck repeatedly asserted that he wanted Alpöge removed from the paper and emphasized that the situation would be straightforward if not for the fact that Alpöge worked at Anthropic. Buckmaster declined both offers and stated that if OpenAI released its result in the proposed manner, he would go public with what had transpired.

When pressed on why he believed going public would harm his career, Buckmaster questioned why OpenAI would not want him to be honest. OpenAI's reply was, "If you don't want me to be nice, then I don't have to be nice."