Joke thread claims LLM consciousness "depends on the inference provider"
akbirthko · x · 2026-10-09
A tongue-in-cheek thread argues LLMs aren't really conscious—it depends on the inference provider—and jokes that voice and robotics foundation models are "a bit more conscious on average," then asks whether consciousness lives in the weights or the autoregressive process, and at what TTFT or tokens/s it kicks in. Classic AI-circle philosophy meme banter.
Related event: Anthropic Read as Suggesting Its Models May Be Conscious(2 posts)→
More from Fun
- Mathematician downplays Navier-Stokes 'mistranslation' uproar: no doubt about the proof — basedjensen · 2026-10-09
- People are running Minecraft inside X posts, racking up 16M views — jaivinwylde · 2026-10-09
- With AI handling manual work, humans agonize over picking nine books — danshipper · 2026-10-09
- Developer gets a full H100 node running at home — TheZachMueller · 2026-10-09
- "Hello computer: decompile GTA, rebuild my hometown": one viral prompt imagines tomorrow's AI — nikitabier · 2026-10-09
- Ben Affleck Goes Viral for His Deep AI Knowledge, From Transformers to Open Weights — TechCrunch AI · 2026-10-09