OpenAI fires 2 researchers over leaks of potentially new AI breakthrough called Q*

OpenAI has terminated two of its researchers, Leopold Aschenbrenner and Pavel Izmailov over alleged information leaks, according to a report from The Information. The saga further underscores ongoing tensions within the company.

Aschenbrenner, known for his role in the AI safety team, was notably associated with OpenAI’s chief scientist, Ilya Sutskever, who had been embroiled in previous conflicts within the company, notably during the upheaval involving OpenAI CEO Sam Altman last year.

The specifics of the leaked information remain undisclosed, but speculation surrounding an AI breakthrough dubbed Q* emerged during a period of boardroom unrest last November. As you may recall when Sam Altman was from the company, there were speculations that Altman was also fired by OpenAI’s board over ‘Q-Star’ breakthrough seen as a threat to humanity.

“It’s not clear what information the two fired staffers leaked. The other staffer—Pavel Izmailov, a researcher who worked on reasoning—had also spent time on the safety team. The ouster of the two men is among the first staffing changes that have surfaced publicly since OpenAI CEO Sam Altman resumed his board seat in March. That followed an investigation led by OpenAI’s nonprofit board, which exonerated him for actions leading up to his short-lived firing last November,” The Information reported.

The other researcher, Pavel Izmailov, known for his work on reasoning, also had ties to the AI safety team. These terminations mark some of the first visible changes in personnel since Altman resumed his position on the board in March, following a thorough investigation by OpenAI’s nonprofit board that absolved him of prior allegations.

Image source: The Information

Aschenbrenner’s involvement with OpenAI’s “superalignment” team, coupled with his association with Sutskever, adds layers to the intrigue, especially considering Sutskever’s pivotal role in the events leading to Altman’s temporary removal in 2023.

While tensions within OpenAI may have simmered down in 2024, these dismissals suggest that there’s still plenty of drama unfolding behind the scenes. The decision to part ways with a reported ally of Sutskever only serves to stoke the flames of speculation, raising questions about the dynamics within the company’s leadership.