UT Professor Explores Question of Robot Intelligence

KNOXVILLE – When the first American flag is planted in Martian soil, it may not be put there by a person, but by a robot.

University of Tennessee Professor Bruce MacLennan said he has little reason to doubt that while the technology to do so doesn’t exist yet, such an advance is a possibility.

However, according to MacLennan, a professor of computer science, before we advance to that point, we need a theoretical understanding of intelligence. He’ll discuss that issue this Friday at the UT Science Forum, held at noon in Thompson-Boling Arena.

MacLennan said before we can ship robots into space or even place them in the real world, scientists must first understand what it means to have a mind.

“The bigger question is, is it really intelligent or is it just acting like it’s intelligent?” he said. “What I want to address is a question of scientific methodology. I want to know how can you frame the question scientifically and how could you address it scientifically.”

MacLennan said there have been decades of highly regarded research and testing that have tried to develop a formula that accurately measures and interprets artificial intelligence. But he believes some of those tests are flawed as the scientists were less interested in the robots’ ability to function physically.

“For me, one of the fundamental signs of intelligence is the ability to operate competently in the real, natural world,” said MacLennan. “Robots that are working in very constrained or controlled environments or just sitting on your desk really don-t provide convincing evidence of intelligence. It’s not so much what you say, but what you do in the actual physical world.”

MacLennan said understanding emotional responses and the processes of the mind are other obstacles that must be considered before constructing a genuinely intelligent robot is possible. He noted that having a better grasp on what it is to have a mind will help humans better understand their own intelligence.

“This question sheds light on our understanding of our own humanness,” he said. “It’s a test case–if you can’t answer the question for robots, then it shows there is a real gap in the understanding of the mind.”

MacLennan said he hopes to raise these issues in an effort to encourage attendees to form an opinion for themselves.

“I hope they see some of the complexity and the subtlety, and that it’s not just an isolated question in artificial intelligence, but has connections to science and the humanities,” he said. “I hope it will get them thinking, even if they have made up their minds already.”

The UT Science Forum is a weekly, non-technical lecture and discussion designed to help others better understand research across many disciplines. It is held every Friday at noon in Thompson-Boling Arena, dining rooms C and D. Attendees may bring their own lunch or purchase it at the arena. Each presentation should last around 40 minutes followed by a question and answer session.

Additional upcoming Science Forums follow:
– “Fusion Energy Research: Status and Outlook,” Friday, March 31, Stanley Milora, director, Fusion Energy Division, Oak Ridge National Laboratory.
– “Who Says You Can’t Microwave a Fork?–Microwaving Metals at Y-12,” Friday, April 7, Ed Ripley, nuclear metallurgist, Y-12 National Security Complex.
– “When Ants Rule the World–Oh Wait, They Already Do,” Friday, April 21, Nathan Sanders, assistant professor of ecology and evolutionary biology.

Contacts:

Jay Mayfield, media relations (865-974-9409, jay.mayfield@tennessee.edu)

Bruce MacLennan, professor of computer science (865-974-5067, bmaclenn@utk.edu)

Mark Littmann, forum program chairman, (865-974-8156, littmann@utk.edu)