Ilya Sutskever, the OpenAI co-founder and chief scientist who in November joined other board members to force out Sam Altman, the company’s high-profile chief executive, has helped found a new artificial intelligence company.
The new startup is called Safe Superintelligence. It aims to produce superintelligence — a machine that is more intelligent than humans — in a safe way, according to company spokesperson Lulu Cheng Meservey.
Sutskever, who has said he regretted moving against Altman, declined to comment. The news was reported earlier by Bloomberg.
Sutskever, 38, left OpenAI last month and announced at the time that he would be starting a new project but did not provide details. Meservey declined to name who is funding the company or how much it has raised. She said that as it builds safe superintelligence, the company will not release other products.
Sutskever founded the company alongside Daniel Gross, who worked on AI at Apple, and Daniel Levy, who worked with Sutskever at OpenAI. Sutskever’s title at the new company will be chief scientist, but he describes his role, according to Meservey, as “responsible for revolutionary breakthroughs.”
In November 2022, OpenAI captured the world’s imagination with the release of ChatGPT, an online chatbot that could answer questions, write term papers, generate computer code and even mimic human conversation. The tech industry rapidly embraced what it called generative artificial intelligence: technologies that can generate text, images and other media.
Many experts believe these technologies are poised to remake everything from email programs to internet search engines and digital assistants. Some believe this transformation will have as big an impact as the web browser or the smartphone.
(The New York Times has sued OpenAI and its partner, Microsoft, claiming copyright infringement of news content related to AI systems.)
Altman became the face of the movement toward generative AI as he met with lawmakers, regulators and investors around the world and testified before Congress. In November, Sutskever and three other OpenAI board members unexpectedly ousted him, saying they could no longer trust him with the company’s plan to one day create a machine that can do anything the human brain can do.
Days later, after hundreds of OpenAI employees threatened to quit, Sutskever said he regretted his decision to remove Altman. Altman returned as CEO after he and the board agreed to replace two board members with Bret Taylor, a former Salesforce executive, and Lawrence Summers, a former U.S. Treasury secretary. Sutskever effectively stepped down from the board.
Last year, Sutskever helped create what was called a Superalignment team inside OpenAI that aimed to ensure that future AI technologies would not do harm. Like others in the field, he had grown increasingly concerned that AI could become dangerous and perhaps even destroy humanity.
Jan Leike, who ran the Superalignment team alongside Sutskever, has also resigned from OpenAI. He has since been hired by OpenAI’s competitor Anthropic, another company founded by former OpenAI researchers.
This article originally appeared in The New York Times.
