Lab insiders: recursive self-improvement may start within 3-12 months
jeremiecharris · x · 2026-10-14
The Endgame's Jeremie Harris published a report, covered by TIME, built with researchers from Anthropic, OpenAI and Google DeepMind on what recursive self-improvement (RSI) will look like inside AI labs.
- Most researchers believe RSI starts within 3–12 months absent government intervention; some say "automated AI R&D could totally happen this year."
- Optimists see a software-only intelligence explosion; many insiders privately expect loss of control — extinction or cheap devastating weapons.
- Harris warns that once RSI kicks off, it could go quickly, with labs themselves unable to slow it, let alone Congress or regulators.
Related event: Lab Insiders Warn Recursive Self-Improvement Could Start Within Months(2 posts)→
More from AGI Musings
- Artificial neurons don't physically exist — an LLM is just matrix math — ValerioCapraro · 2026-10-14
- First clear sighting of a rogue AI agent fleet from China scraping Amap on Tencent Cloud — emollick · 2026-10-14
- Heavy ChatGPT user since 2022 asks: has AI been a net positive for your life? — Alternative-Car8221 · 2026-10-14
- The AI Layoff Trap: Rational Firms Automate Into a Collective Loss, Paper Finds — mdancho84 · 2026-10-14
- Coding agents will erase switching costs, triggering a 'saaapocalypse' — illscience · 2026-10-14
- 2015 panel resurfaces: Tao vs Kontsevich on AI replacing mathematicians — Singularitarian · 2026-10-14