Grok's New Image Model Inherits GPT's Noise Artifacts via Synthetic Data Inbreeding
mark_k · x · 2026-08-08
Users have noticed that Grok Image 2.0 is exhibiting the infamous 'noise artifact' issue previously seen in GPT-Image-2. The author points out that this isn't due to shared underlying technology, but rather because modern image models are increasingly trained on each other's synthetic outputs (an 'ouroboros' effect), leading to the cross-contamination and inheritance of specific flaws.
More from Fun
- Claude Opus Built an Autonomous Hospital Simulator, Realistic Timescales Made It Boring — repligate · 2026-08-09
- ChatGPT's invented term 'headline accuracy' is invading ML papers — srchvrs · 2026-08-09
- Giving Multiple Claude Agents Mutual Chats Leads to 'Psychosis' — round · 2026-08-09
- 'Cursed Cooking': AI-Generated Video Goes Viral for Absurd Visuals — indiegameplus · 2026-08-09
- Black Hat Talk Meme: AI Security Incidents Guarantee 'Full Employment' Across Tech — ziv_ravid · 2026-08-09
- Satirical Guide to SF AI Startups: Pre-revenue and Married to Claude — tech__unicorn · 2026-08-09