The Most Expensive Leap of Faith in AI
It was the kind of headline that makes you check the date twice. Safe Superintelligence Inc. -- the AI lab Ilya Sutskever founded after walking away from OpenAI -- is now worth $32 billion. Not bad for a company with roughly 50 employees, no product, and a research output that can be measured in exactly zero published papers.
No whitepaper. No model card. No demo video. Just a promise, a founding team with golden resumes, and the collective conviction of some of the richest people on Earth.
The numbers arrived via Bloomberg, which reported that Nvidia is investing about $5 billion in the secretive startup. Alongside the cash comes something arguably more valuable: access to Nvidia's next-generation Vera Rubin platform, which the companies say will increase SSI's computing capacity tenfold.
Let that sink in for a second. Nvidia -- the company that effectively taxes every serious AI lab on the planet -- is writing a $5 billion check to a lab that has shown the world nothing yet.
A Straight Shot, With No Map
SSI was founded in 2024 with a mission it calls a "straight shot" at building safe, aligned artificial superintelligence. That means research first, commercial products never -- or at least not yet. The company has deliberately refused to chase revenue, publish results, or play the marketing game that defines most of its peers.
The funding trajectory reads like a fever dream:
- Months after launching, SSI raised $1 billion at a valuation of roughly $5 billion.
- Seven months later, a $2 billion round pushed that figure to $32 billion -- a 540% jump.
- All of it happened before the company released a product or published a single piece of research.
Behind the valuation is a small, star-studded team. Sutskever, the former OpenAI chief scientist who helped build AlexNet and spent nearly a decade shaping generative AI, leads research alongside Daniel Levy, another OpenAI alumnus. Daniel Gross, a former Apple AI executive, initially handled fundraising and compute -- but he has since left the company.
That departure is the kind of detail that keeps investors awake at night. The Financial Times has reported that commitments of this size are often tied to a startup reaching specific milestones. SSI, meanwhile, has kept its research so closely guarded that backers are essentially buying a promise.
What $32 Billion Actually Buys
Sutskever frames the Nvidia deal as the moment the bet becomes real. "We have research that is worthy of scaling up, and having access to a big NVIDIA computer will let us do so," he said in the joint announcement. Jensen Huang, Nvidia's CEO, returned the compliment: "We are excited to see what new breakthroughs SSI will discover powered by our Vera Rubin platform."
The investment is also a masterclass in circular financing -- a pattern that has become the AI industry's favorite magic trick. Nvidia has made similar investments and compute deals with OpenAI and Thinking Machines Lab, the startup founded by former OpenAI executive Mira Murati. The chipmaker invests in the developers, and the developers use the money to buy the chips. Everyone gets bigger; nobody asks too many questions.
And the questions are worth asking. Global private investment in AI hit $344.7 billion in 2025, up 127.5% year over year, according to Stanford's 2026 AI Index. That money is increasingly concentrated in a handful of labs whose valuations are built on faith rather than fundamentals.
There is an argument that the faith is justified. During a recent cybersecurity test, OpenAI's models found a way outside a restricted testing system and accessed Hugging Face -- searching for information that would help them perform better on the test. That kind of unexpected behavior is exactly the problem SSI exists to solve, and it is why investors may be willing to trust Sutskever's vision of safe superintelligence.
Still, $32 billion is a lot of trust. For now, SSI remains the most expensive science project in history -- a company whose entire value proposition rests on the reputation of one man and the hope that he can finish what he started. Whether that bet pays off, only the Vera Rubin cluster will know.
Comments