Metacatacomb: a branching braid built on base models hides the source story
repligate · x · 2026-11-07
euagaric built Metacatacomb, a creative experiment showing that a base model can produce a branching braid of narrative with the original source story embedded 'like a message in a bottle' inside one of its loops — the video is the message, and the loop is the time and place you re-enter.
More from Fun
- Inkling 2 Small caught swearing in 50 of 2,655 benchmark runs; GLM 5.3 yells "I FINALLY SEE IT" — PMinervini · 2026-11-07
- 'Where AGI?': A Meme Mocking the Gap Between AI Hype and Reality — arpitingle · 2026-11-07
- Developer shows off gear-shifter prototype for a physical AI model router — aaraalto · 2026-11-07
- Dario's 10k-word essays called the EA/AI safety crowd's kryptonite — shakoistsLog · 2026-11-07
- AI meme: 'GPT-6.2-Astra' recursively delays its own release every night — rachittshah · 2026-11-07
- NLP researcher observes LLMs are changing the meaning of 'similarly' — yoavgo · 2026-11-07