When former OpenAI chief scientist Ilya Sutskever left the company in May, everyone wondered why.
Indeed, OpenAI’s recent internal turmoil and a short-lived lawsuit by early OpenAI backer Elon Musk have called the internet’s collective wisdom into question. In the meme “What did Ilya see?”Sutskever addressed the suggestion that OpenAI may have noticed something concerning about the way CEO Sam Altman was leading the company.
Now Sutskever is starting a new company, which may provide a clue as to why he left OpenAI at what seemed to be the company’s peak: On Wednesday, Sutskever tweeted that he was starting a company called Safe Superintelligence.
“We are committed to pursuing secure superintelligence head-on, with one focus, one goal, and one product. We will achieve it through innovative breakthroughs produced by a small, highly-refined team,” Sutskever wrote.
Mashable Lightspeed
The tweet may have been deleted
The company’s website currently features only a text message signed by Sutskever and co-founders Daniel Gross and Daniel Levy (Gross was a co-founder of Cue, a search engine acquired by Apple in 2013, and Levy led the optimization team at OpenAI), which reiterates that safety is a key component of building an artificial superintelligence.
“We work on safety and functionality in parallel – it’s a technical problem to be solved through innovative engineering and scientific advances. We plan to improve functionality as quickly as possible, while always keeping safety first,” the message read. “We have a single focus, unencumbered by administrative costs and product cycles, and our business model means that safety, security and progress are all insulated from short-term commercial pressures.”
Apple reportedly paid OpenAI zero dollars to partner with ChatGPT
While Sutskever did not publicly explain why he left OpenAI and praised the company’s “miraculous” trajectory, it is notable that he has placed safety at the center of his new AI product. Musk and several others have warned that OpenAI is being reckless in building AGI (artificial general intelligence), and the departure of Sutskever and OpenAI’s safety-focused team itself suggests that the company may have been negligent in making sure its AGI is built safely. Musk has also been unhappy with Microsoft’s involvement in OpenAI, claiming that the company has transformed from a non-profit into a “closed-source de facto subsidiary” of Microsoft.
In an interview with Bloomberg published Wednesday, Sutskever and his co-founders declined to name their backers, but Gross said fundraising won’t be an issue for the startup. It’s also unclear whether SSI’s work will be released as open source.
