The former scientific director and co-founder of OpenAI is starting his own artificial intelligence (AI) company.
Ilya Sutskever, who retired from ChatGPT's founder last month, announced the launch of Safe Superintelligence (SSI) in a post on Wednesday (June 19).
Starting a new company: https://t.co/BG3K3SI3A1
— Ilya Sutskever (@ilyasut) June 19, 2024
“We approach safety and functionality in parallel, as technical problems to be solved through innovative engineering and scientific advances,” the company said in a social media post.
Superintelligence is within reach.
Building secure superintelligence (SSI) is the most important technological challenge of our time.
We launched the world's first Straight Shot SSI Lab with one goal and one product: safe superintelligence.
It's called “Safe Superintelligence”…
— SSI Inc. (@ssi) June 19, 2024
“We plan to evolve our capabilities as quickly as possible while always keeping safety top of mind, allowing us to scale with peace of mind,” the company wrote. “Our focus means we're not obsessed with administrative costs or product cycles, and our business model ensures that safety, security, and progress are all insulated from short-term commercial pressures.”
SSI has offices in Tel Aviv, Israel, and Palo Alto, Calif. Sutskever's co-founders are former OpenAI researcher Daniel Levy and former Apple AI leader Daniel Gross.
Sutskever left OpenAI in May after playing a role in the ouster of CEO Sam Altman last fall, and was removed from the company's board of directors days later when Altman returned to the helm.
But the two were on good terms, at least publicly, at the time Sutskever announced her departure from the company.
“The company's trajectory has been nothing short of miraculous, and I am confident that OpenAI will build AGI. [artificial general intelligence] “Under the leadership of Altman, President Greg Brockman and Chief Technology Officer Mira Murati, the company will be safe and profitable,” he wrote.
Sutskever was part of the company's Super Alignment team, which focused on AI safety.
Another member of the team, Jan Reike, soon resigned, citing the company's loss of focus on safety, and then took a job at rival AI company Anthropic.
The launch of Sutskever's new company comes shortly after AI companies pledged last month to implement “kill switch” policies that could halt the development of cutting-edge AI models if they exceed a certain risk threshold.
As PYMNTS wrote at the time, some experts have questioned the feasibility and effectiveness of such a policy, as well as its impact on innovation, competition and the global economy.
“The term 'kill switch' is odd here because it sounds like the organization has agreed to stop research and development of a particular model if it crosses a red line related to risk to humanity. It's not a kill switch, it's just a loose agreement to adhere to ethical standards in model development,” Camden Swita, head of AI and machine learning innovation at AI company New Relic, told PYMNTS.
“Tech companies have made these kinds of deals before (for AI, social media, etc.) so this is not new,” Swita added.
