CompArch 2013

The CompArch logo
17–21 June 2013 : Vancouver, British Columbia, Canada

Thursday, 20 June Keynote

Ethics outside the box:
Empirical tools for an ethics of artificial agents

Peter Danielson, University of British Columbia


Software introduces new kinds of agents: artificial software agents (ASA). To create them responsibility, engineers need an ethics of software agency. This pragmatic need for guidance and regulation conflicts with the weakness of moral science. We do not know much about how ethics informs interactions with and between artificial agents, most importantly, how people will regard ASA as agents: their agents (strictly speaking) and also their competitive and cooperative partners. Naturally, we want to deal with these new problems with our old ethical tools, but this conservative strategy may not work, and if not, may lead to catastrophic failure to anticipate the emerging moral landscape. (Just ask the creators of genetically modified foods.)

This talk will look at the box or frame of traditional ethics and some ways to use experimental data to get outside it. We focus on three surprising features of human-ASA interaction that disturb received ethical paradigms:

  1. Blame someone: the victim or the engineer (overactive deontology)
  2. Stop acting: we demand perfection (utopian consequentialism)
  3. Embrace obscurity (the Ultimatum Game)

We show how a experimental, cognitive and game theoretic approach to ethics can situate and explain these problems. However, it cannot solve them; for that we need new methods.

  1. Anonymous input of moral and value data (clickers)
  2. Democratic scale deliberation (N-Reasons)

Both of which demand reformed research ethics.

As some of the data discussed comes from the Robot Ethics survey, you will be better informed and engaged if before the talk you visit:
The class for the conference is: CompArch.


Peter Danielson Peter Danielson is the Mary and W. Maurice Young Professor of Applied Ethics at the W. Maurice Young Centre for Applied Ethics in the School of Population & Public Health, Faculty of Medicine at the University of British Columbia. He is a member of the Institute for Computing, Information and Cognitive Systems and teaches in the Cognitive Systems Program. Danielson studied philosophy at the University of Michigan, Princeton University and the University of Toronto. His has taught and researched philosophy, computer science, applied ethics and cognitive systems at York and UBC and has been a visitor at University of Toronto, SONY computer systems, Tokyo, and the International Institute for Applied Systems Analysis (IIASA), Vienna.

Danielson’s ongoing research program, Artificial Ethics, focuses on the interactions of more and less ethical real and artificial agents. He using methods of evolutionary game theory, agent-based modeling, experimental ethics and computer-mediated public participation to contributive to a cognitive systems approach to ethics. He is the author of Artificial morality (Routledge, 1992) and the editor of Modeling Rationality, Morality, and Evolution (Oxford University Press, 1998) Recent papers on ethics and genomics, ethics of data mining, robot ethics and the methodology of public participation in ethical decision making.

Danielson has led the design, prototyping and deployment of a series of computational tools that combine research with teaching and learning enhancements, ranging from game theory tournaments and evolutionary game theory simulators, to the N-Reasons software platform to support public participation in ethical decision making and classroom tools to support anonymity and automated grading. The N-Reasons platform has been used by to support empirical studies in ethics by a wide variety of researchers in science and technology ethics. His research team is currently working on iClickers for games and experiments, enhancing the N-Reasons platform, and AutoCanon, a platform for autonomously generating and grading course texts.