Watson

As part of day 2 of the Future IT camp, learners got a chance to learn more about IBM's Watson, an A.I. computer known for it's ability to answer questions posed in natural language. Watson is also known for competing on Jeopardy and beating the two best Jeopardy contestants ever: Brad Rutter and Ken Jennings. The reason why Watson is of great interest to the learners is that it uses a Linux based operating system (SUSE Linux Enterprise Server 11) and the open source Apache Hadoop framework. Watson also serves as proof that open source can be used for any project regardless of complexity. Though Watson swept his competition on Jeopardy, designing an A.I. computer system that is able to respond to questions of any format and respond accordingly was an extremely difficult task. Most A.I. systems simply simulate intelligence by basing their behavior on hard coded rules that have been defined by its designer. However, it seems like an insurmountable task to define a rule for every situation in life. Therefore, David Ferrucci and his team of researchers changed their approach and introduced a new A.I. concept called machine learning. Through machine learning, A.I. systems no longer depend on hard coded rules, but are capable of evolving behaviors based on empirical data. This ability was key to Watson's success on Jeopardy.