I’ve previously written about my first job after college, working in the field of artificial intelligence. AI work had an energy and optimism that was very exciting. But, over time, the field slowed down, faced with complex problems that seemed insurmountable.
But over the past ten years, AI been picking up steam again. Researchers are using increasingly large databases of information, and developing new ways of analyzing and learning from them. And they’re working on question-answering machines.
Earlier this week, The New York Times published a really interesting article, What is I.B.M.’s Watson? Watson, part of IBM Research’s DeepQa project, is a “Jeopardy!” playing AI system. The article is a great overview of the field, and an in-depth view of the specific challenges of “Jeopardy!” There’s also a nice video, How Does Watson Work?
It’s fascinating to watch how using voice as the means of interacting with Watson changes the experience. Unlike the text-based interaction of the [somewhat] comparable Wolfram Alpha, voice recognition and speech output make Watson feel much more engaging. Especially in a game context. And combined with the other interface elements (buzzers, etc) of “Jeopardy!” It’s a great example of mixed technology and appropriate interactivity.