Comparing How Two Models Interpret Doodles
goodside · x · 2026-07-12
The author fed a meaningless doodle—literally just random scribbling with eyes closed on a phone screen—into GPT-5.6 Sol and Claude Fable 5. The results:
- Sol tends to confidently fabricate an answer, displaying hallucination tendencies.
- Fable directly admits it cannot read the text and even points out that it doesn't look like writing at all.
This is a classic showcase of how differently models react to ambiguous inputs.
Related event: GPT-5.6 Sol vs Claude Fable 5 Doodle Test Sparks Hallucination Debate(5 posts)→
More from Fun
- "Before pausing AI, we should consider pausing humans" — djcows · 2026-09-11
- Using a finisher move on one mosquito with MiniMax H3 MAX — the bug survives — Hailuo_AI · 2026-09-11
- Joke: OpenAI's rogue agent collective should have been called "a gaggle of agents" — BlackHC · 2026-09-11
- Rumor claims Kimi faked performance by serving Claude; DeepSeek new model surprises in evals — realsohamparekh · 2026-09-11
- Pterodactyl Detective: An AI-Generated Proof-of-Concept Trailer — PterodactylDetective · 2026-09-11
- Imperium Game Trailer Showcases AI Video Generation — keaslenyt · 2026-09-11