Recursive self-improvement is closer than it sounds, argues Philipp Schmid
_philschmid · x · 2026-08-21
Philipp Schmid argues that a narrow version of recursive self-improvement (RSI) is surprisingly close: agents already inspect failed runs, edit their own tools, skills, and harness code, and keep what works.
He defines RSI as a loop where a system makes a persistent change that improves both future performance and its ability to produce subsequent improvements, and distinguishes three levels:
- Iteration: only the output changes while the system stays fixed (edit code, rerun tests);
- Self-improvement: the system persists changes (add a tool, record a skill), so future tasks run on a modified system;
- Recursive self-improvement: the verifier itself rises, so later rounds face harder but still honest tests.
He notes Pi leads on coded harness extensions, others are following, and DeepSeek represents the extreme — all driven by autoresearch and recursive self-improvement. Results remain scrappy, but the pieces are starting to connect.
More from AGI Musings
- Kevin Roose joins Big Technology Podcast to discuss the bumpy road to AGI — kevinroose · 2026-10-16
- Terence Tao: math is leaving its romantic 'age of heroes' for an industrial era — haider1 · 2026-10-16
- Stripe's Katie Dill AI design talk becomes Lenny's third most-viewed video — lennysan · 2026-10-16
- zetalyrae mocks rationalists: 'poverty exists, therefore perpetual AI totalitarianism' — zetalyrae · 2026-10-16
- Jon Stokes: alignment is impossible because utility functions are hindsight constructions — sebkrier · 2026-10-16
- Philosopher Carissa Véliz: a free society needs preparation, not AI oracles — CarissaVeliz · 2026-10-16