Blogger argues AI images for physical goods you sell is AI's dumbest use case
AaronBergman18 · x · 2026-11-11
Aaron Bergman argues that generating AI images of the physical goods you're selling is the dumbest possible use of AI. His thread: retail images exist to distinguish your specific item from the general concept, while AI images merely approximate how light bounces off a generic object — something photography already does better. Most customers already know what a "chair" or "tikka masala" is; AI images are inherently misleading or useless for conveying further visual detail. He grants AI images' real value is representing things you don't have in front of you — exactly the opposite of product photos.
More from Multimodal
- MiniMax H3-Image goes open-weight: native 4K×4K image gen with 16 reference images — socialwithaayan · 2026-11-11
- Five one-shot Fable 5.2 demos: foldable phone, power map, game and more — FinanceYF5 · 2026-11-11
- ElevenLabs Scribe v3 Live tops STT leaderboard with 1.89% WER across 21 models — lukeharries · 2026-11-11
- Four imaginary tokens for Midjourney v8.2: Mapdrift, Ruinwake, Sensecrossing, Shadowdelay — LudovicCreator · 2026-11-11
- Higgsfield + Fable demo stuns: video model fused with LLM nails sarcasm and body language — RileyRalmuto · 2026-11-11
- AI-generated game audio: 544 voice lines, 772 SFX, 37 tracks via ElevenLabs and fal — FinanceYF5 · 2026-11-11