Saturday, February 19, 2011

Watson

It was a sad week in our household this week. It was inevitable, that the day would come, a day which many others have experienced, but a new one for me.

The elevator operator.

The pony express rider.

The medicinal leech farmer.

Where are they now? Gone. Lost in the sands of time as they were replaced by more modern solutions (although I'm confident the occasional medicinal leech farmer hobbyist is still out there).

And now that IBM has introduced its Watson computer, and Watson has handily destroyed the best Jeopardy players of all time, it looks like the era of the quiz show trivia specialist is coming to its end.

We've had a good run, the Van Doren cheating scandal notwithstanding. Our shows were broadcast to millions, and the adoring public basked in the glow of our useless knowledge. The capital of Afghanistan. The more obscure works of Liszt. The name of Hitler's dog. These facts were ours, weapons put to use in epic nerd battles of syndicated television.

Now it's all just stuff in a computer.

And I suppose that's OK. If it took a team of the most brilliant scientists in the country several years and $100 million dollars to build a machine the size of several rooms that works to do something I can do rolling out of bed, that's not the worst thing in the world.

It's also not that much of a surprise. With more powerful computers and more closely tied information networks, data gathering and storage is becoming easier and easier. Whereas in the past, the knowledge and ability to recall specific pieces of information was a highly valued skill (a good doctor who could diagnose an illness, a stockbroker who knows tons of stocks), now the facts themselves are table stakes. Everyone has them, and advances like Watson makes it even easier to get the answers you need.

But, as we saw when Watson called Toronto a U.S. city last night, the technology isn't perfect yet. And until those tools are foolproof, the valuable skill won't be knowing information (though it would help), it will be the ability to use the right tools in the right ways to find the information. Because, simply put, there's a crapload of data out there, and unless you know how to find what you need it's really tough to find anything (how many ehow search results can there really be?)

But I also think that this premium on the ability to use information search tools is a temporary one. As the tools get even better, they'll become even more precise and able to understand our language even better (imagine a more confident Watson, one that knows Toronto is in Canada).

So what happens then?

Then we'll have all kinds of data and information. Complete overload. At that point, finding information should be easier, but you'll need the skills to analyze it and critically examine it.

As I've seen so often in consulting, data in the wrong hands can be very dangerous.

So you'll need smart people who can actually look at and interpret data. This isn't just anyone. It's a set of people who pay attention to detail, who can recognize patterns, and who spot crappy analysis a mile a way (i.e., the people who read footnotes)

In short, the right people are those with skills like my fiancee, who specializes in data analytics. The wrong people are those that don't want to understand detail and have no interest in thinking about where their data comes from.

But that's OK. I'm fine with retiring early.