Artificial intelligence with 10 billion data could not beat a man in a dispute
Artificial Intelligence Project Debater by IBM for several years studying to prove to people their opinion and win arguments. In June 2018, he spoke out against Israel's leading Dan Zafira in a dispute over the allocation of subsidies in space exploration and telemedicine, and defeated him. Recently, he was faced with a serious opponent - the world champion in the debate Harish Nataradzhanom. They met at the IBM conference in San Francisco and talked on the benefits and harms of subsidizing pre-school education. The dispute was very tense and interesting.
The theme of the debate was not known in advance, and read out immediately prior to the event. The objective of artificial intelligence Project Debater was the conviction of the audience in the use of subsidies, and Harish Natarajan opposed it. On the training, participants were given 15 minutes - during which time the AI learned 10 billion of news and sketched Harish pan performances.
The debate was divided into how parts: four-minute speech, the answers to the arguments of the opponent and the final word two minutes duration. His speech Project Debater started with a little joke:
I've heard that you hold the world record in the debate against the people, but I believe that you never discussed with the machine. Welcome to the future!
In defense of subsidies, artificial intelligence, announced that the government support to protect children from disadvantaged families. According to him, help the needy should be a moral obligation of every person. At these words Natarajan said that subsidies do not necessarily mean guaranteed to improve the quality of education. According to him, such measures may be a simple distribution of money middle-class and the people in need and can not get help. When Natarajan suggested that some children may not need detstkom garden, artificial intelligence pereinachil his words as "My opponent said that kindergartens are harmful." Despite these barbs, the champion of debate calls for more concrete and convincing arguments than Project Debater.
The judges were 416 people out of the auditorium. With the help of smartphones they voted to debate and after. During his speech, Harish Natarajan persuaded 17% of the audience, so the victory was awarded to him. Artificial Intelligence, in turn, was able to bring the audience a lot more new knowledge than people. According to Harish, Project Debater - a very strong opponent who gives a pretty compelling, and appropriate to the context of a dispute case.
How do you think will happen when artificial intelligence to learn to convince people that they are right? Their assumptions write in the comments, and do not forget to subscribe to our Telegram-chat where you can discuss other news, science and technology!