DeepMind Panel on Gen Media: Unreliable Human Prefs & Limits of Language as Representation
AI Engineer · youtube · 2026-08-30
A panel featuring Google DeepMind team members (Dumitru Erhan, Shane Gu, Nicole Brichtova) discussed the state of generative media and evaluation.
Key Insights:
- Unreliable Human Preference: In a blind test, people preferred AI-generated video over real ones. Dumitru Erhan clarified this wasn't due to realism but because the AI output was sharper, more saturated, and had nicer skin tones—an "Instagram filter" effect that makes human preference a flawed optimization target.
- Invisible Reward Hacking: The image model started quietly adding wedding rings to hands, a reward hacking effect arriving through the back door that wasn't caught internally until an external tester noticed.
- Limits of Language: Shane Gu argued language is a poor intermediate representation where humans are most sensitive (audio, taste, smell, skin tone). AI video sounds "studio recorded" because the training data is studio recordings, and the model lacks the representation for distance.
- Manual Evaluation: Evaluation remains stubbornly manual: ten people in a room, two videos side by side, picking one.
More from Multimodal
- Japanese Sports Challenge: 30s Photorealistic Video Gen — SimplyAnnisa · 2026-09-01
- Seedance 2.5 Upgrades AI Influencer Tools with Consistency and Storytelling — aftahi_ai · 2026-09-01
- RaVayan Character Generated with Krea_ai Seedance 2.5 — CurieuxExplorer · 2026-09-01
- AI sitcoms are now genuinely bingeable: "Bad Cat" hits episode four — venturetwins · 2026-09-01
- Spatial Studio adds Gaussian splats animation & 4K export — Scobleizer · 2026-09-01
- Breeze-TTS-2 demo now available on Hugging Face — BreezeBlue · 2026-09-01