It was the strangest experience for me to see Watson in its finale. I’d been watching the machine for so long, had grown accustomed to its brilliance, its occasional lapses, its quirks. And suddenly I found myself sharing this with the whole world. And they weren’t always responding the way I expected. For example, when Watson screwed up the Final Jeopardy clue on the second night, providing Toronto for an answer under the U.S. Cities category, I found myself feeling defensive. People were making fun of Watson, and criticizing it! And this after an absolutely dominant performance.
But for me, the machine’s lapses are fun, and even a bit comforting. It’s a reminder about the tremendous amount of work it has to do to come up with a correct response. It’s hard. And if Watson got them all right, people wouldn’t appreciate this. What’s more, Watson’s mistakes are a useful reminder to us that there are plenty of areas where the computer is clueless–and needs human help and supervision. There’s a place for us in a world full of Watson-like computers. They’re brilliant. And for me, story of how this one was built was fascinating. But in the end, Watson, with all its strengths, is still a tool.