Ex-Chief Scientist of OpenAI Starts New Company to Develop ‘Superintelligence’

Former Chief Scientist Ilya Sutskever, one of OpenAI’s co-founders, has initiated his own AI company called Safe Superintelligence Inc. along with ex-OpenAI employees Daniel Levy and Daniel Gross. The focus of this startup, as implied by its name, is centered on creating a robust AI system in a secure manner.

“We tackle safety and capabilities simultaneously, viewing them as technical challenges to be addressed through groundbreaking engineering and scientific advancements,” the company stated in a recent post. “Our goal is to advance capabilities rapidly while ensuring that safety is always prioritized. By doing so, we can scale peacefully.”

Sutskever was reportedly involved in the events leading to OpenAI CEO Sam Altman’s temporary resignation and subsequent return. In a blog post last year, he discussed the concept of “superintelligence” alongside then-OpenAI executive Jan Leike, now at Anthropic. They suggested that superintelligence could become a reality within this decade.

With a singular focus on realizing this vision, Sutskever expressed, “This company is unique in that its primary product will be safe superintelligence, with no diversions until then.” He emphasized the company’s insulation from external pressures and competition.

This Tweet is currently unavailable. It might be loading or has been removed.

The startup’s website features a minimalist design with basic styling and echoes the statement shared on Twitter. Safe Superintelligence is currently hiring and aims to form a select team of engineers.

Levy also expressed his excitement for the new venture online, stating, “I can’t imagine working on anything else at this point in human history.” He envisioned Safe Superintelligence as a “high-trust team that will achieve extraordinary results.”

Defining safe AI on a grand scale, Sutskever remarked, “By safe, we mean safe like nuclear safety as opposed to trust and safety.”