Problem Solving | Idea Generation | Training and Education | News and Information
|Watson, Creativity, and Jeopardy!|
Every time a thinking machine pulls off a stunt like IBM’s Deep Blue trouncing Gary Kasparov in 1997 or IBM’s Watson walloped undefeated Jeopardy! champ Ken Jennings and top money-winner Brad Rutter, I feel it my duty to assure my readers that a machine will never perform a creative act. Watson will never present us with an invention, innovation, or even a child’s poem.
Even with the triumph of Watson, it’s no more difficult to defend the human mind as the unique seat of creativity. On Jeopardy! we have seen a machine out-calculate two human brains. You may call calculation thinking but you may not call it creativity. Yes, Watson has come closer than Deep Blue to doing what we do when we doodle, laugh at a joke, or dream of being famous. But we will never catch Watson singing in the shower.
We see primates, a few other mammals, and even birds behave creatively, but machines—never.
How Watson Works
When Watson considers a Jeopardy! question (actually an answer to which it must provide the question), its 2,500 parallel processor cores, manipulate algorithms written by humans at about 33-billion operations a second. In mind-boggling time, Watson assigns a score to a host of possible responses. When it finds one or more that rise to a certain level of accuracy, it rings in and goes with the highest scoring response.
Richard Powers in What Is Artificial Intelligence? wonders “whether Watson is really answering questions at all or just noticing statistical correlations in vast amounts of data.”
I wonder, what’s the difference? When Jennings or Rutter responds, each puts to work some part of his 100-trillion neurological connections inside three pounds of grey matter. The way this soggy wad of neurons makes and selects responses remains a mystery, but the process may turn out to work not unlike the brains of IBM’s David Ferrucci and his team who created Watson. In their own image?
They have evolved, if you will, Watson’s tremendous new power of open-domain question answering from Deep Blue’s relatively simpler mathematical manipulation of limited, closed-set options. A huge step in the field of artificial intelligence. Still, in the end, Watson simply does more non-creative calculating faster.
How Creativity Works
Your brain serves your human needs. Even when we accomplish great feats of pure logic or engage in cold, rational discourse, we merely approach the dispassionate processing we see in Watson. It’s freedom from fear, pride, and the need to win is one of its greatest assets.
In contrast, driven by a need to perform, accomplish, and solve, Jennings, Rutter, and we lesser humans bring messy passion to the process. We think about how we think, care about what we do, and create what we desire. In this unique way, we are sloppier, error-prone, and creative.
I’ve playfully and ironically identified the source of our creative power as the Seven Creative Juices. Actually they were identified for me by early Christians as the Seven Deadly Sins. I simply point out that these human desires drive creativity as well as they pave the path to pain and demise.
Nonsense, playfulness, and heeding our desires help quiet our fears, lift our confidence, and allow our creative juices to flow. Practice, patience, and perseverance help but passion points the way and powers creativity. (Why have we assigned the letter P to so many of these essential elements?)
Just as the human desire for fun drives games like Deep Blue vs. Gary Kasparov, our need to satisfy our fundamental needs drives invention and innovation. Would anyone argue that Watson knows what it answers or cares whether or not it is correct?
Along with Watson’s developers, Jennings, Rutter, Trebek, his live and broadcast audiences a lot of people had a heck of a lot of fun. Some might call it frivolous. Those who invested the most in the game, sweated more and swelled with the greatest pride—the king of the Deadly Sins as well as the Creative Juices.
Feats of creativity—like Ferrucci’s, not Watson’s—raise more questions than answers. Now that we have Watson, what will we do with it? It obvious has the power to make Google search many times more accurate and efficient. When will it power our searches? What else can it do for us?
Lots of creative folks are working on these questions, I’m sure. And I’m just as sure that neither Watson nor his progeny will ever answer the big ones. They may contribute to questions like, what’s the best exercise program for me? should I see the doc about this discoloration on my ear lobe? But never, whom should I vote for? should stay with my partner? why is there something rather than nothing?
This Right Brain Workout appeared for the first time on IdeaConnection.