An unsuccessful attempt to remove OpenAI CEO Sam Altman was spearheaded by Ilya Sutskever, one of the company’s founders, who said he was launching a safety-focused AI startup.
The well-known AI researcher Sutskever, who quit the company that made ChatGPT last month, announced on social media on Wednesday that he and two co-founders had founded Safe Superintelligence Inc. Development of “superintelligence,” a term used to describe artificial intelligence systems that are more intelligent than people, is the company’s sole objective.
Work on safety and security will be “insulated from short-term commercial pressures,” according to the company’s business model, which Sutskever and his co-founders Daniel Gross and Daniel Levy said in a prepared statement. The company also pledged not to get sidetracked by “management overhead or product cycles.”
According to the three, Safe Superintelligence is an American business with headquarters in Tel Aviv and Palo Alto, California, “where we have deep roots and the ability to recruit top technical talent.”
Sutskever was part of a group that made an unsuccessful attempt last year to oust Altman. The boardroom shakeup, which Sutskever later said he regretted, also led to a period of internal turmoil centered on whether leaders at OpenAI were prioritizing business opportunities over AI safety.
Known as artificial general intelligence, or AGI, Sutskever co-led a team at OpenAI dedicated to securely creating AI that is superior to humans. He stated that he had intentions for a “very personally meaningful” project when he left OpenAI, but he gave no further information.
Sutskever declared that he made the decision to leave OpenAI
Days after his departure, his team co-leader Jan Leike also resigned and leveled at OpenAI for letting safety “take a backseat to shiny products.” OpenAI later announced the formation of a safety and security committee, but it’s been filled mainly with company insiders.