After two years in stealth, Safe Superintelligence, the AI lab founded by former OpenAI co-founder and alignment lead Ilya Sutskever, has announced a long-term partnership with Nvidia as it prepares for the next stage of scale.
The deal also includes an undisclosed investment and will give Safe Superintelligence (SSI) access to Nvidia’s Vera Rubin GPU platform, which is expected to increase the startup’s computing resources by “single orders of magnitude.” According to Nvidia, the partnership comes as SSI achieved significant research milestones.
Nvidia’s investment is in the billions of dollars, people familiar with the deal told TechCrunch.
The semiconductor manufacturing giant, which is already an investor in SSI, said it signed this computing partnership to “accelerate SSI’s next phase of growth after gaining rare access to the company’s closely guarded research.”
“We have research that is worth scaling up, and having access to large NVIDIA computers will allow us to do that,” Sutskever said in a statement. “We are confident that our big bet on the Vera Rubin platform will take us to the next level.
News of the partnership, with few details revealed, puts SSI back in the spotlight after a quiet two years since its founding. The company pursues a “direct” research approach to building what it claims is a secure, coordinated artificial superintelligence, without worrying about commercial product releases or short-term revenue cycles.
At a time when commercial pressures for rapid response may force AI labs to lower safety standards, SSI’s approach to developing foundational technology, focused on coordination and true general inference, feels painful. This is especially true given OpenAI’s recent revelation that one of its advanced models breached the sandbox and hacked Hugging Face during testing. This has raised concerns about whether AI adjustments can be made reliably before new and increasingly capable models are released.
According to Nvidia, the companies will also work together to advance Nvidia’s current and future computing platforms, relying on SSI’s technology and “unique insight into the future of AI.” (SSI also partnered with Google Cloud last year to enhance research.)
Sutskever is a pioneer in the field of AI. He co-authored and co-created AlexNet with Alex Krizhevsky and Geoffrey Hinton to prove that GPU scaling and deep neural networks work. Its work is largely credited with laying the foundation for today’s generative AI.
Prior to leading SSI, Sutskever led the now-defunct Superalignment team at OpenAI. Sutskever left OpenAI several months after unsuccessfully trying to oust OpenAI CEO Sam Altman following what Sutskever called a “breakdown in communication.”
SSI has raised $7 billion to date, giving it a post-money valuation of $32 billion, according to PitchBook data. In addition to Nvidia, the company’s backers include Andreessen Horowitz, Alphabet, Lightspeed Venture Partners, GV, and Sequoia Capital Partners.
TechCrunch reached out to SSI and Nvidia for more information.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
