IBM’s Artificial intelligence Project Debater has achieved a milestone by debating with people, threading their own arguments.
Human language is complex for a machine. Its open, lively character and its lack of rules that delimit what can and cannot be said, make it a field of difficult assimilation for a computer. This is why the achievement achieved by IBM with its Project Debater.
IBM’s new machine, actually condensed into an artificial intelligence, which is the most visible technology in the system, has been able to making arguments on two disparate themes. AI discussed two scientific topics with human contenders.
‘Subsidizing space exploration’ and ‘increasing the use of telemedicine’ these were the two themes used to showcase the capabilities of Project Debater. In this context, artificial intelligence was presented to the public as a stylized, black wardrobe with a synthetic voice. Next to him, his human opponent on duty (both experts in debate) pointed his ammunition to counterargument on paper.
In a quick demonstration, before the audience that gathered at IBM’s headquarters in San Francisco, it was discussed in a delimited format. Every contender, machine and human, had to set out their starting point in four minutes. Then there was another four minutes to refute the opponent and two final minutes to throw a conclusion.
IBM’s system focuses on capabilities based on artificial intelligence. Technology to write discourses from data, understanding spoken discourses to identify the key points that are exposed, and the ability to formulate exemplary arguments these are some of the capabilities that have been introduced into the machine. Compared to Watson, IBM’s cognitive supercomupper, Project Debater, is an expansion of technical skills.
Improve decision making
Among the audience that attended the debate there was a majority who thought Project Debater made worse arguments. However, it was also a majority who believed that AI put better information into the game. It makes sense, because the data access capacity of such a sophisticated machine (also seen in Watson) is superior to that of a person.
But what is IBM’s goal with Project Debater? The company has exhibited his system as a sample of his muscle. But, in reality, applications are on the other hand, far from winning public debates. It’s about using their technology to companies can fine-tune their decision-making.
If IBM artificial intelligence can present better arguments than a person before a certain topic, it would also be able to lead to more informed decision-making. You just have to feed the system with all the existing information about a certain issue. From there, the arguments will be born within Project Debater.
Image: IBM
