IBM's Debating Robot Takes on Its First Human Opponents

IBM's latest artificial intelligence robot showcased its skills against two world-champion debaters.

IBM's Debating Robot Takes on Its First Human Opponents
IBM

IBM's Watson took on Jeopardy contestants. Other artificial intelligence systems can play Go and chess with ease against the world's best players. And now, IBM's latest machine is taking on a human opponent in the realm of debate. 

IBM's Project Debater -- an AI specifically tailored to argue and debate in a formal setting -- debated in public for the first time earlier this week. The Project Debater system went toe-to-toe with duo Dan Zafrir and Noa Ovadia. 

The Project Debater AI has been in the works since 2012, and its performance against world-champion debaters was worth the wait. The robot took on two separate debates and subjects ranged from medicine to education.

"There is a lot at stake today," the robot said in opening statements. "Especially for me."

The AI system even cracked a few jokes. At one point, it regretted not being able to say an injustice made its blood boil because "I have no blood."

“It’s clear that something like that is relevant to anything that has to do with decision making,” said Ranit Aharonov, manager of the debating technologies team at IBM in Haifa, Israel.

IBM has been rather secretive as to how it would ultimately apply Project Debater in the future. It's not being deployed commercially anytime soon. Given the company's history with advanced AI, them deciding to keep the debating bot under wraps for a little while longer makes sense. While Watson's appearance on Jeopardy in 2011 fascinated devoted quiz show fans, IBM hasn't explained how much they use the super-intelligent and incredibly fast AI.

Project Debater digests large swaths of information on a particular topic, scanning newspapers, academic papers, and formal scholarship to decide what snippets it could piece together in its argument. Then, a secondary algorithm culls the repetitions in the data. A voice recognition system comes into play during the debate itself by listening to the opponent. (This could quickly become the trickiest part for the robot to consistently get right, especially if it mishears its human opponent.)

Project Debater doesn't exist solely to make AI systems better, however. IBM officials said the system could improve human argumentation as well.

"The world is awash with information, misinformation, and superficial thinking. Project Debater pushes the frontiers of AI to facilitate intelligent debate so we can build well-informed arguments and make better decisions," the company explained

Blockchain

IBM Unveils the World's Smallest Computer That Is Tinier Than a Salt Grain

IBM continued: "The rise of one-sided and doctored narratives is challenging society and our platforms. Too often, we talk past one another. We need a smarter way. New developments in language and reasoning in AI can help shine a light in the darkness of distorted facts to provide diverse, well-informed viewpoints -- both the pro and the con."

The company also makes it very clear that Project Debater's AI goes without the emotional pull often found in rousing political debates or petty arguments. It provides meaningful insights on a variety of topics, and it manages to do so with grammatically correct English, IBM noted.

Via: IBM