10 years, 10 days ago

Watson: Impressive Finding not Thinking

Link: http://blogs.gartner.com/mike-rollings/2011/02/14/watson-impressive-finding-not-thinking/

National Public Radio (NPR) seems to wake my imagination.  This morning they had a story about IBM’s Watson. Watson is IBM’s computer that is squaring off against two Jeopardy champions – the shows air for the next 3 days.

I wonder how many people will begin to believe that Watson actually thinks? I’m sure it will be a dazzling display of finding. Watson has to find all kinds of things to process the question (e.g. noun, verb, the question) and it has to process even more to find, weigh and select possible answers from historical information. But it is not thinking. (Everybody stop reading and pat yourself on the head. We humans still excel at that over machines).

My favorite example about Watson’s lack of thinking from NPR was “What do grasshoppers eat?”. This question had the response “Kosher” (a sign that many people like the cake?). Sure, you can add more programming, but as the story pointed out humans bring an enormous amount of context to the examination of a question. I would also say that humans add an enormous amount of context to the examination of a question. This subtle and important difference is what makes computers like Watson important helpers but not replacements for humans. We can associate other experiences and insights to make discoveries.

I don’t think that Watson’s creators hope to replace humans, but many people may romanticize about the potential to do so. Instead, I like to think about the augmentation of human cognition and our ability to develop an insight into uncharted territory.

Augmentation is the job for a computer. It is easy to over-emphasize facts that come from a computer to be ‘the truth’. Yet, no matter how powerful machines become, we must reserve the development of truth to a discourse between two or more humans who develop a shared context.