Head of School, Associate Professor Seth Lazar, will be presenting a public talk next Thursday 13 December on artificial intelligence and ethics, What’s wrong with machine ethics? This will coincide with the Anatomy of Artificial Intelligence lecture, presented by Professor Kate Crawford, later that evening.

Machine intelligence is a decision technology. Practically all of the decisions that we would use AI and Machine Learning to support have moral implications. Some of them are explicitly moral. And we are developing autonomous systems that use machine intelligence to make morally-loaded decisions independently. Small wonder, then, that the call for more ethical AI is now practically universal. Machine ethics is the field of philosophy and computer science with the task of making machine intelligence more ethical. In this paper, I introduce machine ethics and explain its shortcomings. I don't have an easy fix. One notable problem, however, is how far removed machine ethics is from leading contemporary work in moral and political philosophy. Fixing this problem (as, for example, the Humanising Machine Intelligence project proposes to do) may help us make significant progress towards ethical machine intelligence.

Associate Professor Seth Lazar is head of the School of Philosophy at the ANU, and leader of the Humanising Machine Intelligence project. He has worked on the ethics of war, risk, and now machine intelligence. His papers appear in the top international philosophy journals; he edited the Oxford Handbook of Ethics of War; and he has written one monograph with Oxford University Press (Sparing Civilians, 2015), and has another under contract (Duty under Doubt: Deontological Decision-Making with Imperfect Information).