After roughly five years of work, IBM has unveiled AI technology capable of debating humans.
Dubbed Project Debater, the technology was demonstated in San Francisco this week, where it was tasked with making an argument in favor of government-subsidized space research — a subject area that the AI had not studied prior to the debate.
"Subsidizing space exploration is like investing in really good tires," argued the computer system in a female voice and housed in a 5-foot tall machine with TV screens mounted on its sides.
After delivering an opening argument using information culled from the technology’s vast database of newspapers, journals and other sources, the computer listened to and then rebutted the human debater’s counter-argument.
Developers at IBM believes that the system is capable of tackling the deeper practices of human analysis and rhetoric, capable of discussing questions with no right or wrong answers.
"If you think of the rules of debate, they're far more open-ended than the rules of a board game," said Ranit Aharonov, who manages the debater project.
Yet, the developers of the system realize that despite being able to better access numbers and other evidence in support of its arguments, there are some features of the system that still need work.
"Humans tend to be better at using more expressive language, more original language," said Dario Gil, IBM's vice president of AI research. "They bring in their own personal experience as a way to illustrate the point. The machine doesn't live in the real world or have a life that it's able to tap into."
For more on the technology, watch the accompanying video.