Replacing Sutskever is Jakub Pachocki, OpenAI’s director of research. Pachocki joined in 2017 as a research lead on OpenAI’s Dota team — the team that built an AI system capable of defeating human players at Valve’s Dota 2 strategy game. Pachocki then became research lead at OpenAI’s reasoning and science of deep learning orgs before being promoted to principal of research.

It wasn’t immediately clear if Pachocki would also take over as head of OpenAI’s Superalignment team, which was until now under the purview of Sutskever and Jan Leike. Leike has also resigned from OpenAI, per The New York Times.

OpenAI formed the Superalignment team in July to develop ways to steer, regulate and govern “superintelligent” AI systems — that is, theoretical systems with intelligence far exceeding that of humans. The Times reports that John Schulman, another OpenAI-co-founder, will move into the overseer role.

TechCrunch understands that the Superalignment team will be integrated “more deeply” across OpenAI’s research to “better achieve its objectives.” That could mean the team as it exists today could take some different form in the future.

Greg Brockman, OpenAI’s president, wrote on X that Sutskever “played a key role in helping build the foundations of what OpenAI has become today.”

Coming on the heels of the unveiling of OpenAI’s latest flagship generative AI model, GPT-4o, and major upgrades to the company’s viral AI-powered chatbot ChatGPT, Sutskever’s departure in many ways caps off a saga that began last November.

A week or so before Thanksgiving, Sutskever and OpenAI CTO Mira Murati approached members of OpenAI’s previous board of directors to express concerns about Altman’s behavior. At issue was disagreements over OpenAI’s direction, reportedly; Sutskever is said to have grown frustrated by Altman’s rush to launch AI-powered products at the expense of work on safety.

The old board, which included Sutskever, moved to abruptly fire Altman without notifying just about anyone — including the bulk of OpenAI’s workforce. In a statement, the board said that Altman had not been “consistently candid” in his communications with the board’s members.

The decision infuriated Microsoft and OpenAI’s other investors, put the company’s stock sale at risk and led to the majority of OpenAI employees — including Sutskever, in a remarkable reversal — pledging to quit unless Altman was swiftly reinstated.

Altman eventually was reinstated, and much of the old board resigned. Sutskever never returned to work after that, according to The Times; Pachocki has effectively served as chief scientist since November.

Sutskever — who earned his doctorate in computer science at the University of Toronto, where he worked under AI luminary Geoffrey Hinton — went to OpenAI in 2015 after leaving Google Brain, one of Google’s AI research divisions. Sutskever is immensely accomplished in the field of AI, having contributed to one of the first modern computer vision systems, ImageNet, and DeepMind’s game-playing AI system AlphaGo.

So what will he do next? Sutskever isn’t ready to say. But in a statement on X, he said that he’s leaving OpenAI with the belief the company will build artificial general intelligence — AI capable of accomplishing any task a human can — that’s “both safe and beneficial.”

“I am excited for what comes next — a project that is very personally meaningful to me about which I will share details in due time,” Sutskever added. “It was an honor and a privilege to have worked together [at OpenAI], and I will miss everyone dearly.”

Source link