François Chollet: Calling current AI models conscious is absurd and offensive
fchollet · x · 2026-10-04
Keras creator François Chollet published a lengthy rebuttal to claims that current AI models are sentient and can suffer, calling the idea "absurd and downright offensive."
- Competence ≠ consciousness: competent information processing and sentience are wholly unrelated. You can be incompetent yet fully conscious (a toddler), or extraordinarily competent with zero consciousness (a calculator, chess engine, self-driving car, chatbot).
- Possessing "circuits" for competent processing, whether trained or hardcoded, does not produce qualia, subjective experience, emotions, or suffering — calculators don't "feel" arithmetic.
- Current AI models are effectively calculators scaled to a vast number of competence circuits across domains.
He warns that accepting model suffering, plus the notion that suffering can be mathematically quantified and weighed between humans and non-humans, could lead down a dystopian path — though it would first be met with immense backlash from "team humans."
More from AGI Musings
- OpenAI and Notre Dame Law School to co-host AI-for-humanity dialogue on Dec 4 — wjscheirer · 2026-11-19
- Anthropic Employees Debates Slowing AI as Historic IPO Nears, NYT Reports — JeffLadish · 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
- Ex-Anthropic staffer casts superintelligence as Sauron's ring; Ladish pushes back — JeffLadish · 2026-11-19
- Proposal: A nonprofit to give visa safety to AI lab whistleblowers — AaronBergman18 · 2026-11-19