SAN FRANCISCO (CNN) - La gente discute de manera excelente.But an IBM project shows that computers are also becoming quite good in it.
On Monday, Harish Natarajan, a great finalist in the 2016 World Debate Championship faced the Project Debater of IBM, a computer promoted by the company as the first artificial intelligence system created to discuss significantly with humans.Natarajan won, but the computer demonstrated an ability to present increasingly complex arguments by artificial intelligence.
Project Debater, who has been in process since 2012, is designed to present his own coherent and convincing speeches, while taking the arguments of a human opponent and creates his own refutation.Even formulates your own closing argument.
To generate their arguments and refutations, Project Debater uses articles of newspapers and magazines of its own database, and also takes into account the nuances of the human opponent's arguments.It is not connected to the Internet and cannot take arguments of sites like Wikipedia.
The debate on Monday, organized by the non -profit company organized by the Intelligence Squared Us debate, was held in front of an audience in the Yerba Buena Center for the Arts of San Francisco.The issue of debate, if the preschool school had to receive subsidies or not, was not revealed to the artificial intelligence system or Natarajan until 15 minutes before they took the stage.Project Debater argued in favor of subsidies.Un aeropuerto estrenará como valet parking a un robotPublicidad
The debate followed the traditional structure: each part presented a four -minute opening speech, followed by a four -minute refutation to the other arguments.Finally they presented a closure of two minutes each.The audience voted on the one hand at the beginning of the debate and again at the end.
Lee: Mexico, with the potential to be a leader in the thanks to their young people
"Greetings, Harish," Project Debater began, speaking with a generally monotonous, female voice.AI argued, among other things, that subsidized preschool can help break the poverty cycle.He spoke in complete sentences and was based on a series of studies (including those of the EU disease control centers).
Natarajan followed him, arguing against the resolution, saying that subsidies would consume resources that middle -class families could use for other things.He also argued that subsidizing preschool does not imply that all children can attend.
"There will still be individuals who will be left due to market realities," he said.
The refutation segment of the debate was where some of the great differences between the human being and the computer (beyond the appearance and vocal abilities) were exposed.Natarajan addressed specific parts of Project Debater's arguments and rejected them, such as saying that it is not realistic to expect a government to have a budget without restrictions for assistance programs.
Project Debater's refutation, although expressed with eloquence, seemed more like a continuation of his initial argument than a true refutation of Natarajan's points.He kept his best arguments against his closing statement.Although out of order, all the elements of a debate seemed to be present.
While waiting for the final vote, Natarajan said it had been interesting that Project Debater could contextualize the information and obtain research details."It could be incredibly powerful" combine their abilities with those of a human.Prior to the debate, 79% of the audience approved that preschool was subsidiary and 13% disagree.In the end, 62% of those present were in favor and 30% disagree.Because this style of debate is scored on the side that wins more percentage points, Natarajan took the victory.
Lee: truly intelligent artificial intelligence: A risk to the human?
Project Debater shows how artificial intelligence systems have become increasingly flexible in recent years.The artificial intelligence that we are used to seeing, as digital attendees integrated in smart speakers, can only be used very limitedly, such as answering specific questions.But the IBM system shows how technology can also be used to explore problems that do not necessarily have a single answer.This could help people find new ways of working with computers and use AI to help us find more solutions to problems.
"It is really pushing the limits of the types of artificial intelligence systems that are more interactive with us and that can understand us better," said Darío Gil, director of IBM Research to First Move from CNN on Monday.