Machine ethics: A solution in search of a problem?
Machine ethics: A solution in search of a problem?
In March 2020 Seth Lazar presented a paper on machine ethics to an interdisciplinary conference at CMU. His respondent was Professor Jonathan Cohen (Princeton). Here's the abstract:This will be a mostly critical paper focused on the shortcomings of the existing machine ethics literature, from both practical and theoretical perspectives. Though it may seem like flogging a dead horse, I'll raise some new objections to 'top-down' approaches to formalising moral reasoning and constraints. I'll also criticise a few different machine-learning based approaches to machine ethics. Reinforcement learning, inverse reinforcement learning, and evolutionary/game theoretic efforts will take a kicking too. I'll aim to close by trying to reformulate the problem that I think machine ethics should be trying to solve, and suggesting some ways forward.