Ethan Mollick: gap between labs' internal models and public releases may be widening
emollick · x · 2026-10-30
Wharton professor Ethan Mollick notes there is no good measure for the gap between labs' best internal models and the best models released to the public. He suspects it has grown considerably over the past year, especially after the temporary "Fable ban" — implying the public may be far behind frontier internal capabilities.
More from AGI Musings
- Toby Walsh discusses "Making Sense of AI" with MP Allegra Spender — TobyWalsh · 2026-10-30
- Founder dad on raising kids in the AI age: "I wish AI would take jobs faster" — alfred_lua · 2026-10-30
- "A New Dawn for Mathematical Progress": a deep, positive vision for AI-partnered mathematics — LiminalEntityX · 2026-10-30
- If 'total lightcone domination' were truly at stake, nation states would act very differently — rickasaurus · 2026-10-30
- littmath: recognizing AI risks in articles like this is a reason for (some) optimism — littmath · 2026-10-30
- Levie and Mollick: the gap between frontier AI users and everyone else keeps widening — mhmazur · 2026-10-30