Author weighs technology against issues of morality

Director of Cognitive Science at IU and author Colin Allen discussed ways in which machines can adhere to ethical standards during a lecture Sunday night at the School of Fine Arts.

Allen was invited by the Secular Alliance at IU to speak as part of the College of Arts and Sciences’ Themester fall 2012: “Good Behavior, Bad Behavior, Molecules to Morality.”

Allen co-authored “Moral Machines: Teaching Robots Right from Wrong,” which explains the near-term prospects for making machines and computers behave more responsibly and ethically.

“I care more about how can we give machines the kind of capacities that would make them fit into our environment more than they presently do,” Allen said.

Allen used the example of autonomous, or self-driven, cars as machines with more responsibility. He said that Mercedes-Benz and Google are testing autonomous driving systems.

“One of the most significant obstacles to proliferation of autonomous cars has been illegal,” Allen said.

Autonomous cars have been legalized under certain restrictions in Nevada.
Allen said one concern about autonomous cars is it has been hard for them to follow human direction, such as commands from a traffic cop or ambulance.

Allen also said it’s possible autonomous robots might be able to perform some tasks better than humans.

He discussed the top-down and bottom-up approaches of robotics.

In the top-down approach, the creator decides what capabilities the robot has, according to the Mind Project, an Illinois State University project that concentrates on cognitive science research.  The bottom-up approach requires no central program.

Allen reiterated a message first said by Massachusetts Institute of Technology professor Rosalind Picard, “The greater the freedom of a machine, the more it will need moral standards.”

Robots stimulate emotion in some way, Allen said.

“Even though the robot doesn’t feel emotions, we understand that,” he said.

Allen also discussed some challenges of writing his book.

“We focused too much on individual machines rather than systems of machines or humans,” he said.