by Diann Daniel

IBM Supercomputer to Compete on Jeopardy

Apr 27, 20092 mins
Data CenterEnterprise ApplicationsOutsourcing

What's the best way to showcase your AI-powered supercomputer? For IBM, that solution will be to pit its question answering system, Watson, against humans on a popular game show. n

IBM today announced that it aims to take artificial intelligence to a new level with its supercomputer, codenamed “Watson,” sending the technology to compete in the world of game shows.

Related Articles

15 Must-See Sights in Google Earth

Fitness Tech: 15 of the Hottest Health and Exercise Gadgets, Gizmos and Apps

MIT Wearable Gadget Gives You Sixth Sense, a la Minority Report

The computer has the ability to analyze complex questions and form answers so well that it can compete with humans on the game show Jeopardy, IBM says, and a machine vs. human competition on the show is in the works.

Watson, in its final stages of development, is a question answering system that has been in development for nearly two years. Underneath Watson’s hood, natural language processing and other technologies analyze meanings behind words. This gives Watson the ability to identify relevant and irrelevant content, interpret ambiguous expression and puns, decompose questions into sub-questions, and synthesize information to form an answer, IBM says. Watson considers massive volumes and varieties of natural language, then quickly analyzes and scores supporting or refuting evidence—in a matter of seconds.

Jeopardy is the perfect forum in which to display Watson’s advances in artificial intelligence, according to IBM. Contestants on the show must quickly and accurately answer the problems posed by host Alex Trebeck by drawing on a wide range of knowledge to quickly analyze irony, riddles, subtle meaning and complexities.

IBM says Watson can do the same–in other words, do complex human-like thinking that computers have not traditionally done well–with precision and speed. Especially important will be Watson’s ability to rate its confidence in the answers it produces.

“The challenge is to build a system that, unlike systems before it, can rival the human mind’s ability to determine precise answers to natural language questions and to compute accurate confidences in the answers,” said Dr. David Ferrucci, leader of the IBM Watson project team, in a press release. “This confidence processing ability is key. It greatly distinguishes the IBM approach from conventional search, and is critical to implementing useful business applications of question answering.”

Details are still forthcoming regarding exact timing for Watson’s debut on Jeopardy.