RSI debate sweeps Silicon Valley: Anthropic says AI writes 80% of its code
新智元 · wechat · 2026-10-13
A roundup of the RSI (recursive self-improvement) discussion sweeping AI circles. Substantive points:
- Anthropic disclosed that Claude wrote 80%+ of merged internal code by May; by August, AI served as lead developer on 26% of R&D with 90%+ human-AI collaboration. Anthropic says full RSI hasn't arrived.
- OpenAI chief scientist Jakub Pachocki reportedly wrote internally that current progress could lead straight into RSI while chain-of-thought monitoring is "progressively diminishing."
- A Google engineer claims agents now handle most performance optimization at frontier labs, with release cycles cut from 125 to 44 days.
Much of the rest — rumors about secret models, Grok, and doomsday framing — is speculation and should be treated skeptically.
More from AGI Musings
- Stanford and Databricks use AI agents to reproduce medical findings and test analytical robustness — ChengleiSi · 2026-10-23
- ISMS Fellow Avi Goldfarb: AI disruption raises returns to exploring research directions — avicgoldfarb · 2026-10-23
- signull on AI-found PHP exploits: economic liability is the forcing function legacy software needs — signulll · 2026-10-23
- TheStalwart: Talking computers are sparking a new philosophical boom, echoing ancient Greece — toptickcrypto · 2026-10-23
- AI cyberattacks could paradoxically accelerate AI development in an offense-defense arms race — kimmonismus · 2026-10-23
- Economist on AI detector Pangram: not a filter, but a useful verification signal — paulnovosad · 2026-10-23