> I'm honestly not sure what the point of this paper is. They state it clearly in the abstract I think: "we must develop effective ways to describe their behav

> I'm honestly not sure what the point of this paper is. They state it clearly in the abstract I think: "we must develop effective ways to describe their behaviour in high-level terms without falling into the trap of anthropomorphism.", which seems pretty sensible to me. But from this high point it is mostly downhill, with the chief problem being that despite them wanting to avoid "anthropomorphism" they are still obviously using boatloads of it everywhere: > Role play is a useful framing for dialogue agents, allowing us to draw on the fund of folk psychological concepts we use to understand human behaviour—beliefs, desires, goals, ambitions, emotions and so on Astonishingly, after listing this collection, explicitly labelled "human behaviours", they claim there's no risk of anthropomorphism! I guess they mean that it's not anthropomorphism if you are just declaring the thing to be an actual human... Personally I think you'd be much better off seeking inspiration in the terminology of older fields that have dealt with objects pretending to be humans. For example, while I don't know much about theory of painting, I'm still pretty sure it doesn't try to discuss how people respond to a painting by hiding its author and upgrading the finished painting itself to an active agent with its own motivations...