Mollick: Meta's Muse nails the personal-assistant agent UX, backed by massive free compute
emollick · x · 2026-09-21
Wharton professor Ethan Mollick calls Meta's Muse an impressive implementation of the OpenClaw idea — an AI personal-assistant agent you chat with ongoingly — with an experience so accessible that people who didn't realize what AI could do for them can pick it up. He adds two context points: the labs are focused on enterprise because that's where willingness to pay for tokens is highest, so even though their systems can do what Muse does out of the box, they aren't as intuitive; and Meta is giving away a lot of compute to make the experience work this well.
Related event: Mollick praises Meta Muse as the best consumer AI personal assistant(2 posts)→
More from AGI Musings
- DeepSeek team told a VC two years ago that media generation was 'unserious' — michellechen · 2026-09-22
- UnitedHealth kept AI care-denial tool despite 90% appeal reversal rate, lawsuit says — aakashgupta · 2026-09-22
- Polymarket Puts Just 11% Odds on US-China Deal to Pace the AI Frontier — Polymarket · 2026-09-22
- ChatGPT and AlphaFold already helped design canine cancer mRNA vaccines, researcher notes — cloneofsimo · 2026-09-22
- LLMs have commoditized the "excited but harmless" corporate tone, and jobs will go with it — StewartalsopIII · 2026-09-22
- Musk: via user's IP and cookies, Amazon can't tell human buyers from AI agents — elonmusk · 2026-09-21