Schmidhuber: banning superintelligence is infeasible as compute gets 10x cheaper every 5 years
SchmidhuberAI · x · 2026-09-24
Responding to Polymarket's "ban superintelligence" debate, Jürgen Schmidhuber argues a ban is technically infeasible: compute gets 10x cheaper every 5 years, and soon millions will run frontier labs on desktops, executing recursive self-improvement (RSI) algorithms.
He recaps his own RSI timeline: first concrete RSI algorithms published in 1987 (when compute was 100,000,000x more expensive), self-modifying policies in 1994, gradient descent-based RSI in neural nets since 1992, asymptotically optimal curriculum-learning RSI since 2002, the mathematically optimal self-referential Gödel Machine since 2003, RSI with artificial curiosity/intrinsic motivation since 1990/1997, and recent work since 2020. Details in technical note IDSIA-9-26.
More from AGI Musings
- Prediction: hospitals will keep their own tested open-model formularies in two years — MaziyarPanahi · 2026-11-12
- Packy revives Qwest CEO's 1998 'no bandwidth glut' quote for the AI oversupply debate — prestonpesek · 2026-11-12
- Stanford's James Zou joins The Economist podcast on AI agents in science — anshulkundaje · 2026-11-12
- a16z maps the B2A economy: AI agents are now most of web traffic and soon its buyers — venturetwins · 2026-11-12
- Jason: Anthropic is rebuilding every SaaS product into Claude—free for life — smtabatabaie · 2026-11-12
- Fleuret: if AI designs drugs as fast as it makes dumb videos, regulation will buckle — francoisfleuret · 2026-11-12