Wow! This is beautifully written and compelling in its thesis. It is so refreshing to have moved on from the nonsense rhetoric of "lies" and "hallucinations" that has dominated recent AI discourse. As you know, I think the history of mechanical automata has such a lot to tell us about attitudes to AI - I hadn't thought of Descartes though - great example. You approach this field from psychoanalysis in general and Lacan in particular and so you end up with language as the key paradigm for exploring these interactions. That tempts me to get Foucault and critical discourse analysis out of my theoretical toolkit and come at the same thing from that angle. Anyway, thanks for posting - really helpful!