Ilya Sutskever, the co-founder and chief scientist of OpenAI’s for-profit arm who helped briefly oust CEO Sam Altman last year, is leaving the artificial intelligence firm.
“After almost a decade, I have made the decision to leave OpenAI,” Sutskever wrote in a May 14 X post saying he was going to work on a “personally meaningful” project. He added he was confident the firm would build an artificial general intelligence (AGI) that is “safe and beneficial.”
Altman posted on X that Sutskever and OpenAI would “part ways” and the former chief scientist “has something personally meaningful he is going to go work on.”
OpenAI announced Jakub Pachocki as its new chief scientist. Pachocki has been the firm’s director of research since 2017 and led the development of its GPT-4 large language model.
Sutskever was part of an effort that successfully saw OpenAI’s board briefly push out Altman as CEO in November last year before he was later re-hired to the role after backlash from employees.
The Information reported at the time that Sutskever — who was one of six board members for the nonprofit OpenAI, Inc. — told employees Altman’s ouster was the board doing its duty to ensure OpenAI “builds AGI that benefits all of humanity.”
Sutskever later wrote in a November X post: “I deeply regret my participation in the board’s actions.“
After Altman was rehired, Sutskever stepped down from OpenAI’s board and it’s been unclear what role he had at the firm which sparked the meme template “Where is Ilya?” that speculated what position he had at OpenAI.
Related: Elon Musk sues Sam Altman, OpenAI over agreement breach
AGI is a term for a hypothetical artificial intelligence that can perform the same as or better than humans on a range of tasks.
Altman said last month he doesn’t care how much it costs to make an AGI, even if he spends $50 billion a year.
Others, such as Ethereum co-founder Vitalik Buterin, have called for a measured approach to developing a possible intelligent AI saying it poses unknown risks.
AI Eye: How to get better crypto predictions from ChatGPT, Humane AI pin slammed
Update (May 15, 5:20 am UTC): This article has been updated to add information throughout.