The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Ilya Sutskever’s undisclosed project turned out to be Safe Superintelligence Inc. (SSI), an AI company he co-founded in June 2024 with Daniel Levy and Daniel Gross. SSI says it is focused on one goal: building safe superintelligence, rather than selling consumer AI products. As of August 18, 2026, the company has announced an investment and computing partnership with NVIDIA, but it has not publicly described a model, technical breakthrough or product.
Why did Ilya Sutskever leave OpenAI?
Sutskever, an OpenAI co-founder and its chief scientist, announced on May 14, 2024, that he was leaving after nearly a decade at the company. He said he was pursuing a “very personally meaningful” project but did not identify it. OpenAI named Jakub Pachocki as his successor. Associated Press reporting on his departure described the announcement and its timing.
As an Amazon Associate I earn from qualifying purchases.
The departure followed months of turmoil after OpenAI’s November 2023 board attempt to remove CEO Sam Altman. Sutskever took part in that board decision and later said he regretted his involvement. The episode sharpened public debate over OpenAI’s governance and the balance between commercial product development and AI safety.
That context matters, but it does not establish a single cause for Sutskever’s resignation. Reporting connected his exit to wider tensions over AI safety and OpenAI’s direction; Sutskever did not publicly say that a dispute with Altman, or a belief that OpenAI had abandoned safety, was the reason he left. Jan Leike, who co-led OpenAI’s Superalignment team with Sutskever, resigned shortly afterward and criticized the company’s safety priorities. The departures are evidence of safety-related disagreements, not proof of Sutskever’s precise private motive. AP’s account of the broader dispute provides that context.
#1 Best Overall
What was the mysterious project?
On June 19, 2024, roughly a month after his departure, Sutskever revealed Safe Superintelligence Inc. The company’s co-founders are Sutskever, former OpenAI researcher Daniel Levy and former Apple AI executive and Cue co-founder Daniel Gross. Its identity was initially unclear because Sutskever had supplied no company name, technology description, funding details or launch date when he announced his exit—not because public evidence pointed to a particular secret breakthrough.
SSI’s stated mission is to build safe superintelligence. The company describes its organizational model as “one goal and one product,” meaning it does not present a conventional sequence of consumer products or interim releases. SSI says it intends to develop safety and capabilities together, with safety keeping pace with or ahead of increasing capabilities. Its public site lists offices in Palo Alto and Tel Aviv and emphasizes recruiting. SSI’s website describes the company’s mission and structure; TechCrunch’s launch coverage reported its formation as a for-profit research company.
Rank #2
What does “safe superintelligence” mean?
Superintelligence generally means an AI system that substantially exceeds human intellectual ability across a broad range of tasks. “Safe superintelligence” is SSI’s mission and branding, not a universally agreed scientific category with a standard test or definition.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →In SSI’s framing, safety is a core technical problem to pursue alongside capability development, not simply a compliance step added before a product launch. The company has not publicly explained how it defines safety operationally, what alignment techniques it uses, or what evidence would count as success. Its stated aim is therefore not the same as a demonstrated result.
How does SSI’s approach differ from OpenAI’s?
The contrast is organizational, not proof that one company is safe and the other is not. OpenAI combines research with deployed products and commercial activity; SSI says it is concentrating on a single long-term research objective without an ordinary product roadmap. OpenAI evolved from a nonprofit into a complex commercial structure, while contemporary launch coverage described SSI as a for-profit company organized around its research goal.
| Question | OpenAI at the time | SSI’s stated model |
|---|---|---|
| Primary activity | Research alongside deployed AI products | A single research objective: safe superintelligence |
| Product cycle | Consumer and commercial products | No stated intermediate product roadmap |
| Safety framing | Safety research within a broader AI company | Safety and capability development as the company’s sole focus |
| Corporate structure | A nonprofit-origin organization that evolved into a complex commercial structure | A for-profit company, according to contemporary launch coverage |
| Public technical disclosure | Products and research released selectively | Little technical detail disclosed publicly |
SSI says its business model is intended to insulate safety, security and progress from short-term commercial pressure. That is an organizational intention, not a guarantee: a for-profit research company can still face pressure from investors, competition, hiring needs and the cost of computing infrastructure. A single-purpose lab may avoid some product deadlines, but its limited public disclosure also makes outside assessment harder.
Who funded SSI?
Axios reported on September 5, 2024, that SSI had raised more than $1 billion in its first financing round. The reported investors included Andreessen Horowitz, DST Global, Sequoia Capital, SV Angel, and NFDG, the partnership associated with Daniel Gross and Nat Friedman. This is reported financing, not a disclosed valuation or evidence of unlimited resources. Axios’s financing report also noted the challenge of funding a long research program without near-term product revenue.
Free tools Windows power users keep installed
One-click scans. No signup required.
NVIDIA’s July 2026 announcement separately said it had invested in SSI, without disclosing the amount. The announcement described an investment and strategic partnership, not an acquisition.
Best Value
What changed with NVIDIA in 2026?
On July 27, 2026, NVIDIA and SSI announced a long-term strategic partnership. NVIDIA said SSI would gain access to its Vera Rubin computing platform and that the arrangement would increase SSI’s computing capacity by an order of magnitude—roughly tenfold. Neither the starting capacity nor the final hardware count, training budget or independently verified compute figure was provided.
NVIDIA also said SSI had spent two years quietly developing a new research direction and that NVIDIA had rare access to that closely guarded work before agreeing to the partnership. The announcement suggests SSI is moving beyond its quiet startup phase toward larger-scale research, backed by additional infrastructure. It does not reveal what the research direction is, establish that it works, or show that SSI has achieved artificial general intelligence or superintelligence. No technical paper, benchmark, safety case or independent audit accompanied the announcement. NVIDIA’s announcement is the source for the investment, Vera Rubin access and compute claim; those claims are company statements, not independent verification.
What has SSI publicly disclosed—and what remains unknown?
| Publicly disclosed | Not publicly established |
|---|---|
| SSI’s stated goal is safe superintelligence | Model architecture or training data |
| The company says it develops safety and capability together | Its alignment methods or operational definition of safety |
| It lists Palo Alto and Tel Aviv as locations | Compute budget, system scale or training milestones |
| Axios reported more than $1 billion in 2024 financing; NVIDIA later announced an investment | NVIDIA’s investment amount or SSI’s valuation |
| NVIDIA announced a partnership involving access to Vera Rubin systems | Independent evidence for the announced compute increase or research results |
| NVIDIA identified Sutskever as SSI’s co-founder and CEO and Daniel Levy as a company leader | A named public model, API, consumer application, benchmark suite or release date |
As of August 18, 2026, SSI’s public website describes its mission and recruiting but does not name a public model, consumer application, API, benchmark suite or downloadable research system. That describes its public disclosures; it does not establish whether private systems exist. Without published methods, evaluations or milestones, outsiders cannot independently assess what SSI is building, how it measures safety, or whether its approach is working.
Why the company’s secrecy matters
A long-horizon research lab can, in principle, avoid the pressure to ship unfinished products and give researchers room to pursue unconventional ideas. SSI’s reported financing and NVIDIA partnership also give it access to substantial capital and computing resources. Those inputs can support ambitious work, but they are not evidence of a technical result.
The trade-off is accountability. Keeping methods private makes it harder for researchers and the public to scrutinize safety claims. A company with no near-term product revenue may need further financing, and a concentrated effort around one high-risk objective can limit external checks. SSI has not yet publicly shown that its approach works or released evidence by which outside observers could evaluate it.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
