After trillions of tokens of gradient descent, the model becomes the optimizer
DimitrisPapail · x · 2026-11-03
Dimitris Papail offers a pithy paradox: after trillions of tokens of gradient descent, the model converges to an optimizer that no longer needs gradients at all — a one-line distillation of the "models learning to self-improve" frontier idea.
More from Fun
- zeeg shows off SLAG, an AI-built Sonic-style party game with 6-player lobby — zeeg · 2026-11-03
- 'Definitely not Mario': AI-generated copyright-dodging game characters go viral — zeeg · 2026-11-03
- Everyone Is Building a VCS Replacement — and It's Turning Into a Following Frenzy — latkins · 2026-11-03
- Sentry CEO David Cramer open-sources Slag, a party platformer built with Claude Code — zeeg · 2026-11-03
- AI launch culture debate: paid engagement is a worse look than an earnest, cleanly executed launch — Jackyhuang · 2026-11-03
- Suffering is like pH: essential at 2/10, horrifying at 8/10 — louisvarge · 2026-11-03