Josh Purtell argues LLMs are computationally too simple for consciousness, sparking alignment-circle debate
JoshPurtell · x · 2026-10-04
Developer Josh Purtell debated alignment-community figures (Rob Bensinger, Vladkf et al.) over a "scale → complexity → consciousness" gradient. He finds the gradient intuitively obvious and claims it's largely settled that LLMs are computationally simple, needing few "gears" — especially given how small they can be. To rebuild intuition, he rescales: a lock, key, or watch is obviously not conscious, a warehouse-sized LLM still isn't, but a star-sized one starts to feel plausible. An opinionated AI-circle debate on machine consciousness criteria rather than a settled conclusion.
Related event: Machine Consciousness Debate: Is an LLM a Book or a Star-Sized Brain?(26 posts)→
More from AGI Musings
- Oak Lab debuts with 'Making Learning Abundant' manifesto on efficient continual learning from raw experience — MartinKlissarov · 2026-11-19
- AI pundit mocks Europe: 'bred not to desire too much' — same arrogance as mocking SpaceX — teortaxesTex · 2026-11-19
- AI alignment debate: a dark forest scenario where superintelligences preemptively strike GPU-hoarding rivals — jessi_cata · 2026-11-19
- OpenAI and Notre Dame Law School to co-host AI-for-humanity dialogue on Dec 4 — wjscheirer · 2026-11-19
- NYT: Anthropic Shows Organizational Fractures Ahead of Historic IPO — kylekabasares · 2026-11-19
- The biggest advantage of a top lab isn't money — it's saving you years — madeofAjala · 2026-11-19