Rumors: Google, Anthropic, OpenAI already using RSI; author asks about China and compute
kimmonismus · x · 2026-10-10
The author relays rumors that Google, Anthropic, and OpenAI are all already using recursive self-improvement (RSI): Google's rapid progress and OpenAI's recent math results are allegedly linked to RSI, while Anthropic has reportedly begun training a larger model ("Fable 6") to catch up. He raises two questions: China is notably quiet on RSI (though a few firms like GLM are pushing real steps), and once RSI arrives, compute and energy matter even more — with NVIDIA the biggest beneficiary.
Related event: Rumors: Major AI Labs Nearing or Achieving Recursive Self-Improvement(2 posts)→
More from AGI Musings
- Anil Seth: AI could get 1000x more intelligent without ever becoming conscious — anilkseth · 2026-10-10
- Personal AI agents will fragment like 2023 chatbots: taste will beat raw horsepower — thisiskp_ · 2026-10-10
- Nick Bostrom's 'mind crime' risk: unknowingly torturing billions of conscious machines — cccalum · 2026-10-10
- Baumol's cost disease explains Japan's cheap goods and pricey services — and what's valuable in the AI era — oran_ge · 2026-10-10
- Synths were once sold as a cheap orchestra replacement — generative vocal pop is repeating history — jasonbaldridge · 2026-10-10
- AI commentators: AGI disruption only goes mainstream when half of white-collar jobs vanish — iruletheworldmo · 2026-10-10