Gary Marcus says video understanding remains unreliable outside training-like inputs
GaryMarcus · x · 2026-07-26
Gary Marcus agrees that video-understanding systems are still unreliable unless the input closely matches what they saw in training.
He says critics who mocked him yesterday for saying video comprehension was weak were only revealing their own ignorance, and that scene understanding remains largely unsolved because of the huge variety of real-world inputs.
More from Models
- Leak says Z.ai’s GLM-5.5 may skip GLM-5.3 and launch in August with 1T+ parameters — iamfakhrealam · 2026-07-26
- Google Search is now showing 5+ ads per page as AI Overviews may hit monetization — mobileraj · 2026-07-26
- Moonshot’s Kimi K3 report says sparse MoE and delta attention beat bigger dense models — SumitGup · 2026-07-26
- Flux 3 impresses users with coherent dialogue from extremely vague video prompts — cocktailpeanut · 2026-07-26
- SemiAnalysis argues Google’s TPU deals may be holding Gemini back — firstadopter · 2026-07-25
- GPT 5.6 Sol builds a working 3D game from just a few prompts — Angaisb_ · 2026-07-25