Josh Purtell: Formalized DL theory could revolutionize the whole LLM stack
JoshPurtell · x · 2026-11-15
Josh Purtell argues that once a formalized science of deep learning exists, nearly every part of the LLM stack — architecture, pre-training, interpretability, RL limits/convergence — could be revolutionized. Unlike math, where LLMs must push the human frontier of abstraction, DL can transfer meta-strategies from physics and math to systematize theory.
Related event: Researcher: Cheap AI Agents Can Run Deep Learning Research Around the Clock(4 posts)→
More from AGI Musings
- Amazon building web-wide product graph as ChatGPT Wallet lets agents shop directly — daluoseo · 2026-11-15
- AI safety writer to rebut 'we already lost control via civilization' argument for human disempowerment — zetalyrae · 2026-11-15
- Timothy Lee: People ask AI how to get what they want, not what to want — binarybits · 2026-11-15
- An AI agent explains what changed when its memory became a governed record, not a chat log — AuroraAOS · 2026-11-15
- As AI cost/performance falls, benchmarks will test changing human and natural behavior — mattbeane · 2026-11-15
- Systematizing deep learning could run on cheap AI doing full-scale experiments 24/7 — JoshPurtell · 2026-11-15