'The craftsmanship of training very deep models is lost in the LLM age'
_arohan_ · x · 2026-09-04
xidulu observes that while everyone talks about looping for more depth at inference time, the craftsmanship of training a very deep model appears to be lost in the LLM age — modern practice favors iterative looping over architectural depth.
More from AGI Musings
- AI Now warns public backlash against AI is deeper than a PR problem — AINowInstitute · 2026-09-04
- Go pros' decision quality and new-move invention rate jumped after AlphaGo — burny_tech · 2026-09-04
- Researchers: Chain-of-Thought Legibility Is Too Fragile to Underpin Long-Term AI Safety — tszzl · 2026-09-04
- VCs are 'paralyzed' on AGI: many lack any hypothesis about whether it's real or coming — km · 2026-09-04
- We built artificial intelligence first, then decided to do artificial psychology on it — round · 2026-09-04
- "Everything will be K-shaped": AI job market splits between learners and laggards — round · 2026-09-04