OpenAI's Altman privately floats three options for controlling AI
Last week, OpenAI chief executive Sam Altman walked employees through three AI safety scenarios in an internal presentation.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
Last week, OpenAI chief executive Sam Altman walked employees through three AI safety scenarios in an internal presentation. The options were an international watchdog modeled on the IAEA, a merger of the major labs into one entity, or a single company powerful enough to shut down everyone else's compute. Altman called the single-company option terrible.
Why it matters
Getting all frontier AI developers to agree on how to pace AI technology is a difficult problem. Safety advocates have urged OpenAI and Anthropic to cap compute used for research and development.
Timeline
- Last week: Altman presented three AI safety scenarios to OpenAI staff in an internal presentation.
- Oct 21, 10:00 AM ET: AI safety advocates called on OpenAI and Anthropic to cap compute for research and development.
Coverage
- OpenAI, Anthropic Should Cap Compute for R&D, Say AI Safety Advocates (theinformation.com)
- Sam Altman reportedly pitched three endgames for AI safety, one involving a single firm melting everyone else's GPUs (the-decoder.com)
- OpenAI’s ‘Pacing 2.0’ Problem: How Smart Is Too Smart? (theinformation.com)
- Dealroom.co | Altman tells OpenAI staff there are three endgames for AI: an IAEA for AI, a merger of the labs, or one firm that "melts everyone else's GPUs" (app.dealroom.co)
- OpenAI CEO Altman Presents Three AI Safety Scenarios (Kang Da-eun)
- Altman Describes Three Endgames for AI, Including One Where a Single Firm ‘Melts Everyone Else’s GPUs’ (theinformation.com)
More from Technology & AI
- China puts humanoid robots to work serving food and sorting parcels
- Newsom proposes sweeping AI safety rules for Washington
- Google releases Gemini 4 Neon AI model with million-token context
- Microsoft and Intel widen their cloud chip partnership
- Google DeepMind rolls out Genie 4 to select subscribers
- AI helped AMD shorten chip design work, OpenAI says