Close

Elementary, my dear Watson

“Anyone who has begun to think places some portion of the world in jeopardy.”  – John Dewey

You can learn a lot from the game of Jeopardy! as revealed in a recent blog by Jonah Lehrer and the original work by IBM, who have spent years building a “question answering” machine called Watson, in the quest to develop artificial intelligence.  While Watson performs impressively, (s)he has some significant failings, and chief among them is that (s)he doesn’t know what (s)he knows in the same way that you and I know what we know.  What I mean by that is that humans can be remarkably quick to react to a question, even when they do not have the answer, but only know that they know the answer (ie we recognise that the answer is buried somewhere in our brain, and therefore react to a question, without having the answer immediately to hand).  Think of all those times that you stick your hand up to ask a question, without knowing precisely what you want to ask – despite your pre-emption, you still manage to ask something intelligent.  Or remember the experience of having something on the ‘tip of your tongue’ – you know that someone’s name beings with K, but you can’t quite remember what the name is!

We often know that we know the answer to something, without knowing what the answer is.  That is, our “feeling” or intuition about something anticipates (significantly) our knowledge of the same thing.  This is where Watson, the computer, consistently failed – while (s)he was precise and impressively accurate in answering questions, in the Jeopardy! game (s)he was often very slow to know that (s)he knew the answer, losing out in anticipating and reacting in a competitive situation.

The ability of human brains to, almost immediately, anticipate the appropriate reaction after assessing the context, relevant inputs and appropriate response, even before “knowing” what the response actually is, demonstrates how important our unconscious knowledge is to anticipating and predicting the world around us.  The reality is that emotions do NOT “interfere” with thinking, but rather are an essential predictor and driver of our conscious thought.

The implications for research are profound.  As discussed in many other posts, our “feelings” often know more than our “conscious” minds.  More intriguingly, perhaps this also means that someone’s speed to anticipate that a question is relevant tells us more than their actual answer to the question.

REFERENCES

http://scienceblogs.com/cortex/2010/06/feelings_of_knowing.php

http://www.nytimes.com/2010/06/20/magazine/20Computer-t.html?hpw=&pagewanted=all

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *