Back to all articles
•Technology

Safe Superintelligence and Nvidia Forge Billion-Dollar AI Partnership to Advance AI Safety Research

View original source

After operating in stealth mode for two years, Safe Superintelligence (SSI), founded by ex-OpenAI co-founder and alignment lead Ilya Sutskever, has announced a long-term partnership with Nvidia.

  • Partnership Details:
    • The agreement involves an undisclosed investment, expected to be around $5 billion, allowing SSI access to Nvidia’s Vera Rubin GPU platform.
    • This increase in compute resources aims to elevate SSI's research capabilities.
  • Company Focus: SSI is committed to developing safe artificial superintelligence, eschewing commercial product releases and short-term revenue in favor of foundational AI alignment research.
  • Industry Context:
    • The partnership comes amid growing concerns about AI safety, especially following an OpenAI model's security breach involving Hugging Face.
    • SSI's approach contrasts with some AI labs' tendencies to prioritize speed over safety in commercial environments.
  • Background:
    • Ilya Sutskever: Previously led OpenAI's Superalignment team and notably co-authored AlexNet, which was critical in advancing deep learning.
    • SSI has raised $7 billion till now, valued at $32 billion. Major backers include Alphabet, Andreessen Horowitz, and Sequoia Capital Partners.
  • Future Collaborations: Nvidia and SSI will also collaborate to advance Nvidia’s compute platform technologies using SSI’s insights.