Meta's Devansh Tandon: LLM recommenders are AI's biggest consumer app, 100x cheaper than chat
AI Engineer · youtube · 2026-09-25
Meta Recommendations Research lead Devansh Tandon argues LLM recommenders will be AI's biggest consumer app: 4 of the world's 10 most-used apps are feeds, and feeds are up to 100x cheaper per engagement hour than chat apps since they decode pointers to existing content rather than generating tokens.
Key points
- Four S-curves: traditional recsys, LLM-inspired, LLM-native, agentic
- Recipe: semantic IDs (a 3-min Reel from 10,000 tokens to 10), pre-train bilingual on English + catalog, post-train a chain-of-thought re-ranker
- Meta reports 30% more watch time on Instagram Reels from this approach
- Steerable feeds: Instagram's Your Algorithm lets users inspect and edit what the model thinks they like; prompted playlists, custom feeds, Ask DoorDash
More from Apps
- Designer builds 100k-site German heritage map with Claude, zero code written by hand — neutronenkind · 2026-11-11
- God of War creator David Jaffe hails AI game-making tool Spawn as the future he fantasized about — majidmanzarpour · 2026-11-11
- Valius Uses AI to Help an 11-Year-Old Boy Fight Osteosarcoma — anshulkundaje · 2026-11-11
- AnswerThis MCP brings clickable, line-level paper citations into ChatGPT and Claude — Faheem_uh · 2026-11-11
- 8 hidden Android features most users don't know, with setup steps — LearnWithBishal · 2026-11-11
- Ramp hid an AI-only offer on its site; AI mentioned it to buyers 370 times in 32 days — VeryWellVersed · 2026-11-11