IBM's Watson Can Now Debate Its Opponents

We may earn a commission from links on this page.

Since defeating the world's greatest Jeopardy players, IBM's Watson has been busy at work in the healthcare industry. But now, the artificially intelligent computer has undergone a fairly substantive upgrade — one that enables it not just to extract information, but to "understand" and reason from it as well.

Last week at the Milken Institute Global Conference there was a session entitled "Why Tomorrow Won't Look Like Today: Things that Will Blow Your Mind." During the panel discussion, John Kelly III, the Senior Vice President and Director of IBM Research, asked: "Can a computer with access to large bodies of information like Wikipedia extract relevant information, digest and reason on that information and understand the context … and present it in natural language, with no human intervention?"


The answer would appear to be yes. As reported in KurzweilAI:

In a canned demo, Kelly chose a sample debate topic: "The sale of violent video games to minors should be banned." The Debater was tasked with presenting pros and cons for a debate on this question.

Speaking in nearly perfect English, Watson/The Debater replied: "Scanned approximately 4 million Wikipedia articles, returning ten most relevant articles. Scanned all 3,000 sentences in top ten articles. Detected sentences which contain candidate claims. Identified borders of candidate claims. Assessed pro and con polarity of candidate claims. Constructed demo speech with top claim predictions. Ready to deliver."


Watson then presented three relevant arguments in favor of banning violent video games for minors, but qualified its assessment by bring up several relevant counterarguments and considerations. In all, it was a fairly cogent review of the data.

Skip ahead to 45:25 in the video to see the result of Watson's analysis:

It's important to note, however, that Watson doesn't actually understand the content — it's just running algorithms on the data and conducting probabilistic analyses to reach conclusions. But it's a good example of machine/human collaboration — with this added caveat from K|AI:

"So Watson still needs human beings to work collaboratively with," said moderator Richard Sandler, Executive Vice President, Milken Family Foundation. "As of today," Kelly replied, apparently in jest.



You may also enjoy: