AIThe Vergeabout 2 hours ago
Why AI food looks like that
1 min read
Diffusion models struggle with thin structures like noodles, producing unsettling AI food images with holes and cracks. Experts at Oxford, Naples, Zurich, and King's College London say models lack real-world understanding, reproducing photographic looks without context. Training on AI-generated content can worsen visual degeneration.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
