IBM’s supercomputer Watson made big news last week by defeating two of the world’s best-ever Jeopardy contestants, Ken Jennings and Brad Rutter. In fact, the competition wasn’t close; the computer came away with nearly four times as much money as the closest nearest competitor.
I don’t think I’m alone, however, in being unimpressed with this development. The fact that a computer can win a trivia contest that relies on fast reflexes and fast information retrieval doesn’t seem much like news to me.
A lot of people, apparently, agree. Matt Blum at GeekDad wrote an article this week called Why Watson’s Jeopardy Win is Mostly Meaningless saying:
IBM’s supercomputer software Watson’s win on the game show Jeopardy! is little more than a publicity stunt. Its value as a scientific experiment is roughly on par with grade school students showing what happens when you soak a tooth in Coke or add salt to a plant’s soil — that is, it’s not an invalid experiment, but it’s also not exactly news.
What would be impressive: if enterprise software immediately performed, as advertised, in the corporate setting (even though instead of getting to run on Supercomputers, it is forced to run on Sosocomputers).
Right this minute, thousands of hard-working corporate employees around the world are pulling their hair out trying to get meaningful data out of their ERP, CRM, PLM and other enterprise software systems.
Another similar (though paid less) group of thousands are the training developers whose jobs are to help others use the software once it’s working. Unfortunately, this group will be forced to develop training in an impossibly short period because the software won’t actually work until dangerously close to the date it is scheduled to roll-out to thousands of as-yet unsuspecting end-users.
In fact, it is this last group, the instructional designers and training developers of the world, who may actually have the most to gain if computers – and software – really do become Super one day.Supercomputer's Win on Jeopardy: Little Comfort to Those Implementing Enterprise Computer Software by Jon Matejcek