We're in San Francisco at IBM Think 2019, and we met this impressive robot capable of defeating the best human debate.
IBM Think 2019 started in San Francisco, and it was great since Monday, February 11, we were lucky to see another great milestone in technology. Another round of robots vs. humans.
Yes, because Fayerwayer saw the first debate between AI "Project Debater" against Grand Finalist in the 2016 World Debate Championship and 2012 European Debate Champion, Harish Natarajan
The truth is that none of them had previous preparations and only 15 minutes before they were started, they got information and topic in the discussion
The proposed theme: Is state aid for pre-school education necessary?
"Project Debtor, there is one she was in favor of the proposal and the human Nataraja against.
While the public could vote online to reward the position of the liberal machine or conservative European champion.
How IBM's project debates work
Project Debates consists of some 10,000 million phrases taken from newspapers and academic publications.
Its creators Naon Slomin and Ranit Aharonov explain that, instead of trying to defeat humans, it is valuable to use an AI to contextualize large databases and thereby prepare a discourse capable of articulating reasoning.
For almost 30 minutes of debate, with 03:30 minutes for arguments, machine and man, there was no respite.
And given the final result and reversal of the original trend, it seems to be the person who won.
Of course, one surprised Natarajan at the end of the debate, of course, exclaimed in awe that he had never had such a hard opponent. She, the IBM robot made him extremely his ideas to disprove. A large amount of data in context was no more than a good dose of emotional intelligence, articulation of ideas and especially weight and gesticulation. Non-verbal communication that still destroys.