Artificial Sentience: From Computational Functionalism to Behaviourism?

Abstract

I critically examine Jonathan Birch’s account of how brain emulations might create artificial sentience candidates, arguing that his account risks collapsing into a behaviourist account, unless the specific computational processes realised by such emulations are carefully considered. Drawing on examples from computational neuroscience, I show that similar behaviours can arise from vastly different computational processes, thereby undermining the suggestion to focus on behaviourial profiles to determine whether a brain emulation of a sentience candidate is itself a sentience candidate.