IBM Artificial Intelligence Takes on Human Debate Champs | Tech News
IBM’s Project Debater, “the first AI system that can debate humans on complex topics,” was put to the test last night at an event in San Francisco.
The AI went head-to-head with professional debaters Noa Ovadia and Dan Zafrir on two topics: whether space exploration should be subsidized, and whether we should increase the use of telemedicine.
“In both debates, the audience voted Project Debater to be worse at delivery but better in terms of the amount of information it conveyed,” according to The Guardian. “And despite several robotic slip-ups, the audience voted the AI to be more persuasive (in terms of changing the audience’s position) than its human opponent, Zafrir, in the second debate.”
Following previous AI milestones like Watson’s 2011 Jeopardy! win, IBM now aims to demonstrate it’s possible for humans to have meaningful discussions with a machine.
“The goal is to build a system that helps people make evidence-based decisions when the answers aren’t black-and-white,” IBM says. “In the future, we believe machines will be able to help humans with many important decisions we make daily.”
The competition followed a traditional debate format: Project Debater and its human challenger each presented a four-minute opening speech, then a four-minute rebuttal, and a two-minute closing statement. The system mines “hundreds of millions of articles from numerous well-known newspapers and magazines” searching for the most relevant points to support or contest the topic at hand.
“The debate topic is not known in advance, and Project Debater is not pre-trained on any specific topic,” IBM says.
Project Debater, which has been in the works since 2012, can “automatically generate a whole speech, reminiscent of an opinion article, and deliver it persuasively.” The system then uses Watson Speech to Text to “listen” to the human debater’s side and prepare its rebuttal. IBM says Project Debater can listen to and unpack up to four minutes of continuous spoken speech.
“Because it is a machine, Project Debater is impartial—it isn’t out to prove a position or to be ‘right,'” IBM says. “It is designed to help expand minds and help people see more than one side of an issue.”