Artificial intelligence (AI) has been a hot topic of discussion for years, but at the conference TV antenna by 2019 In San Francisco (California), it was technology itself that debated. IBM Project Debater, The first artificial intelligence system (AI) that could argue with people on complex issues, faced its rival with a larger diameter Monday, February 11 at the Yerba Buena Center for the Arts.
The opponent of the Debater Project was the Grand Final of the World Debating Championship 2016 and the 2012 European Discussion Champion, Plow Natarajan. None of them prepared the subject of the debate in advance If genes are to be subsidized. Each side had only 15 minutes to prepare their arguments for and against a particular thesis. To defeat Natragen, the system was forced to gather relevant facts and opinions, give them form with built-in arguments, and then communicate their ideas clearly and convincingly.
"It's hard to prepare for a discussion with an opponent that is almost unknown, but I imagine that the Debbeiter project will challenge"Said Natarajan ahead of the competition. "Chess, Lego, L'Opardy, Starcraft – The challenges of the man in front of the machines ended in victory machines."
The argument was divided between an initial statement of four minutes, a delay of equal duration and a two-minute summary. John Donaben, Four Emmy Award winners receive a series of discussions Modi'in in the United States, Discount the session.
After the debate was announced Harish Netergen by the audience. Yet, The Dovrat Project He pointed out that he had learned more than the answers provided by artificial intelligence.
The Project Debater Knowledge Base is complex About 10 million phrases taken from newspapers and academic publications. In development since 2012, the Debater project is IBM's next milestone in the AI field, following previous advances such as Deep Blue (1996/1997) and Watson in danger! (2011). The goal of the project's supervisor is to build a system that helps people make informed decisions when the answers are not black or white.
The first discussion of the Debbeiter project took place last June in front of a small audience against Israel's champions Noa Ovadia and Dan Tsafrir. Against its employees was the machine to argue in favor of government subsidies for space research.
"Subsidizing space exploration is like investing in good tires"Said the machine, in the form of a 1.5 meter high monolithic meter, which would enrich the human brain, inspire the young and be a" very solid investment, "he said, making it even more important than public roads, schools or health services. Ovadia was tried in the hearing by the mass of journalists who witnessed the incident.
The iBM computer was performing better in the second hearing against Zafrir, this time claiming the benefits of telemedicine. The machine noted sources, a joy for the audience and even some hearing jokes in the process. In general, The audience thought that the person would better communicate the claim, but the computer came out decisively in the "enrichment of knowledge". At least nine members of the audience have changed their minds on the computer.
"Humans tend to be better at using a more expressive language, a more original language"Said Drio Gil, IBM Research Director," They bring their personal experience as a way to illustrate this. The machine does not live in the real world and has no life that it can exploit. "
Unlike machines like Watson, IBM says it is breaking new ground by creating a system that deals with deeper human practice, such as rhetoric and theoretical analysis. "If you think about it, the rules of the debate are much more ambiguous than the rules of board game" Said Ranit Aharonov, who runs the project.
"The art of discussion is probably more complicated for a machine than any of these games," agreed European champion Natarajan in an article published on Sunday.
"In essence, the success of the discussion has three components. First, the debater needs to process large amounts of information and build relevant arguments. Second, the complex argument explains complex arguments in a clear and structured manner. Third, it is necessary to make these arguments important to the public. It requires careful use of language, emotions, rhetoric and concrete examples of real life. While the machine must excel first, the latter two can be a challenge. However, I'm sure there will soon be a point where AI will be better than humans in all three cases"Added Dovrat's rival.