OpenAI's former chief scientist has raised $1 billion for his new firm to develop safe artificial intelligence systems.
Ilya Sutskever co-founded Safe Superintelligence (SSI) in June after departing OpenAI in May following a failed attempt to oust CEO Sam Altman in November 2023, which was initially backed by Sutskever.
Reports suggest that the investment values SSI at $5 billion. The investors include Andreessen Horowitz, Sequoia Capital, DST Global, and SV Angel, SSI said, as well as NFDG, which is co-run by Daniel Gross, an SSI co-founder and the CEO.
So far, SSI has ten employees in Palo Alto, California and Tel Aviv, Israel. According to Reuters, SSI plans to spend the investment hiring top AI engineers and researchers, as well as on the necessary processing power. Both staff and computing are costly when it comes to developing AI.
Sutskever initially backed the efforts to oust Altman, which appeared to largely focus on the tension between AI safety and shipping usable AI products. However, amidst the chaos that ensued at the AI giant he swiftly u-turned,
"I deeply regret my participation in the board's actions. I never intended to harm OpenAI,” Sutskever said in a statement posted to X.
Sutskever announced his departure from OpenAI in May, saying at the time he was "confident that OpenAI will build AGI that is both safe and beneficial".
But weeks later, in mid June, he announced the launch of safety-focused SSI, alongside Gross, who previously worked on AI at Apple, and Daniel Levy, also formerly of OpenAI.
Sutskever previously worked with "godfather of AI" Geoffrey Hinton, who stepped down from Google in May 2023 in order to more openly talk about the risks of artificial general intelligence (AGI) and super-intelligent AI.
SSI isn't the first company to emerge from OpenAI with a focus on safer AI. In 2021, Dario Amodei and his sister Daniela Amodei founded Anthropic to create safer AI after leaving the firm, with both reportedly concerned about the direction of the company.
Safe Superintelligence’s plans
SSI publicized its launch via a single website page with plain text on white background.
A framework to drive trusted, transparent, and explainable AI
"We have started the world’s first straight-shot SSI lab, with one goal and one product: a safe superintelligence," the company said at the time.
"We approach safety and capabilities in tandem, as technical problems to be solved through revolutionary engineering and scientific breakthroughs," the statement says. "We plan to advance capabilities as fast as possible while making sure our safety always remains ahead."
Gross said in an interview with Reuters not to expect a product for years — a contrast to companies like OpenAI that are pushing out marketable versions of AI to fund wider work on AGI.
"It's important for us to be surrounded by investors who understand, respect and support our mission, which is to make a straight shot to safe superintelligence and in particular to spend a couple of years doing R&D on our product before bringing it to market," Gross told Reuters.
