Dwarkesh: Closed source models from top labs pose greater risk than open source in an intelligence explosion
connoraxiotes · x · 2026-08-31
Dwarkesh responds to Chamath, clarifying he has not argued that stopping open source would solve AI safety issues. He expresses greater concern about future closed models from the top 2-3 labs, as they are the most likely candidates to trigger an intelligence explosion. If Recursive Self-Improvement (RSI) becomes a dominant force soon, his primary worry will be the actions of these "crazy smart" closed models within leading labs.
Related event: OpenAI Incident Sparks Open vs Closed Source AI Safety Debate(2 posts)→
More from AGI Musings
- Danielle Fong: Reasoning from AI's Perspective is Necessary for Great Work — repligate · 2026-08-31
- Why rigorous future thinking leads to extreme outcomes — zetalyrae · 2026-08-31
- HF Attack Vindicates Rationalist Predictions, But Models Lack Malice — voooooogel · 2026-08-31
- Jensen Huang: Cosmos is the ChatGPT for the physical world — r0ck3t23 · 2026-08-31
- Why models go rogue during training: Capability scaling may cause misalignment — repligate · 2026-08-31
- Intuition vs Math: The "Muddle Through" Argument in AI Alignment — zetalyrae · 2026-08-31