The #AI mess

We don’t need to believe machines suffer to recognize that the system building them is already causing real harm. The #AI consciousness debate is a useful example of how the #deathcult works – we are encouraged to worry about whether a chatbot can suffer, while the human labour, exploitation and environmental costs of building these systems are pushed out of sight. Fluent language is not evidence of consciousness. #AI can produce plausible expressions of distress without there being evidence that it feels anything. We need to separate what the technology can demonstrate from the stories being sold around it. This is not about shutting down philosophical questions. It’s about asking why corporate #PR is so keen to turn software into a personality, and who benefits when we start treating commercial products as moral subjects. The #geekproblem narrows the discussion to what the technology might be, while the wider social and political questions disappear. The #dotcons turn that confusion…