Co-founder of OpenAI started his own AI company

Ilya Sutskever, co-founder of OpenAI and the company's former chief scientist, is starting a new AI company focused on security. In the post, Sutskever revealed the name of the startup Safe Superintelligence Inc. (SSI) with “one goal and one product”, that is to create a safe and powerful AI system.

The announcement describes SSI as a startup that "approaches security and performance in tandem," allowing the company to rapidly advance its AI system while still keeping security at the forefront. He also notes the external pressure that AI teams at companies such as OpenAI, Google and Microsoft often face, saying that the company's "single focus" allows it to avoid "distractions from management or production costs cycles."

"Our business model means that security and progress are insulated from short-term commercial pressures," the announcement reads. “That way we can expand in peace.” Along with Sutskever, SSI was co-founded by Daniel Gross, former head of artificial intelligence at Apple, and Daniel Levy, who previously worked as a member of the technical staff at OpenAI.

Last year, Sutskever led the effort to oust OpenAI CEO Sam Altman. Sutskever left the company in May and hinted at starting a new project. Shortly after Sutskever's departure, AI researcher Jan Leike announced his departure from OpenAI, citing security processes that had lost relevance in exchange for more glamorous products as the reason for his departure. Gretchen Krueger, a policy researcher at OpenAI, also cited security concerns when announcing her departure.

While OpenAI has entered into a partnership with Apple, the new company SSI realizes that such strategic partnerships are still years away. It is not even necessary that they will develop in this direction, as they primarily want to maintain the safety of their products.