Home / Business / IBM AI can not beat the champion of human debate

IBM AI can not beat the champion of human debate

Each side received 15 minutes to prepare for the debate, after which they presented a four minute opening statement, a four minute rebuttal and a two minute summary. The audience of 100 people, meanwhile, was composed of the main participants of the Bay Area schools and more than one hundred journalists.

Miss Debater (formerly known as Project Debater) drew arguments from her database of 10 billion prayers taken from newspapers and academic journals. A female voice emanating from the human-sized black box emitted its responses, while three blue balls floated around its screen.

The confrontation was the last event in IBM's "big challenge" series that pitted humans against their intelligent machines. In 1996, his computer system beat the great chess master Garry Kasparov, although the Russian later accused the IBM team of cheating, something that the company denies to this day: then he retracted some of his accusations. Then, in 2011, his supercomputer Watson beat two record winners Danger! contestants

In the run-up to Monday's fight, Natarajan suggested that debating can be a harder battlefield for AI than Go and video games. "However, the debate is … more complicated for a machine than any of those … In its essence, the successful debate involves three components," he wrote in a LinkedIn post.

"First, a commentator needs to process large amounts of information and construct relevant arguments." Second, the debate implies [explaining] Complex arguments in a clear and structured way. Third, you need to make those arguments important to the audience. This requires the careful use of language, emotions, rhetoric and examples. While a machine should excel in the first, the last two can be difficult. "

But he also admitted that "soon there will come a point where artificial intelligence will be better than humans in all three". Does that mean it would be a game for a rematch?

Source link