What if those dystopian science-fiction films depicting out-of-control killer robots foreshadow our future? An unusual meeting of artificial-intelligence researchers earlier this year focused on discussing the dangers of ever-smarter and more-pervasive computers, in the hope of avoiding such nightmare scenarios.
The two-day meeting took place quietly back in February, but The New York Times broke news of it over the weekend. Eighteen top researchers from college and business labs attended the invitation-only event, sponsored by the Association for the Advancement of Artificial Intelligence.
Tom Mitchell, professor of artificial intelligence and machine learning at Carnegie Mellon University, attended the meeting. He said it was a far-ranging discussion, including fears along the lines of the plot of the 1980s film Robocop, in which a gun-wielding machine turns against its creators. Mr. Mitchell said that his own biggest concerns were more immediate — such as preventing misuse of the data that cellphone companies can glean about users' whereabouts from today's smartphones. “I don't think we should stop using cellphones with GPS sensors,” he said. “But we as a society should start taking a much more serious look at the issues raised by them.”
Martha Pollack, incoming president of the association that sponsored the meeting, did not attend the event, but she has seen a draft of a report from the meeting scheduled to be released later this year. “Artificial intelligence is getting successful, so the ethical issues become more pressing,” said Ms. Pollack, who is dean of the School of Information at the University of Michigan at Ann Arbor.
She said the meeting was the first of its kind for the artificial-intelligence association, though she said researchers regularly have informal discussions about the implications of their work.
One Response to Top Computer Scientists Met This Year to Discuss Dangers of Their Creations
oiegroup - July 29, 2009 at 11:13 am
NASA sponsored a similar conference in the early 1990s; I believe the results are unclassified and are probably still available from the Govt. Printing Office. The primary focus of the conference was to evaluate whether Vernor Vinge’s predictions of a technological singularity were anything to be concerned with. To NASA’s surprise, almost every one of the scientists evaluating Vinge’s speculations and warnings backed him up. My guess is that there have been more behind the scenese follow-ups from that NASA meeting since then, and it’s nice to see that the issue remains on someone’s radar, you know, just in case Vinge was right…