So, Watson wins at Jeopardy. Now what to do with Mr. Watson?
The last time IBM did something truly stunning, “Deep Blue” defeated Russian Grandmaster Gary Kasparov in 1996. To celebrate its victory, IBM dumped the machine in the trash and went forward in its quest to sell large blocks of time to Fortune 500 companies.
In all fairness, IBM just does not know how to compete with products. The products it now sells are difficult to replicate platforms like mainframes, servers, esoteric software. To its credit, IBM at least knows this and has dumped mass market products to which it lost bruising battles to Microsoft and hardware manufacturers.
So, it comes as no surprise to me to see IBM pitching “Health care applications” and other arcane uses for its incredible achievement. The people in that ivory tower just have no idea how people use computers these days.
How about a search engine for Watson?
Google’s natural language processing sucks. Nobody else even comes close to Watson’s ability. The keyword + hyperlink “votes” model just does not cut it when it comes time to answer natural language questions. There is just too much garbage out there with almost everyone trying to manipulate search results for their commercial gain.
A search engine that allows people to get answers quickly will get immediate market share with a giant advertising bounty to boot. If there was ever a Google killer, this would be it. If executed well, Google’s primary business could be easily leapfrogged in no time. Clearly a lot of work would have to be done to make it all work, but the upside is so huge, it is almost a sin not to do it. Why not setup a funded independent unit that is partially venture funded with a world class management team?
But, visionaries, these people at IBM are not. They have become a giant Body Shop (aka “services”). It amazes me, that they have an R&D budget focused on Artificial Intelligence. They know nothing about commercializing these advances.
I never understood why the Chess grandmaster was not put on the web during its early days. It would have more than paid its way at $10 per game. You would have had people like me lining up all day to play against a grandmaster. Where on earth do you get that opportunity? It would have given IBM a connection with end users during the early days, and a way to pitch its platforms – if not make money on Chess. Instead, “Deep Blue” made it into the dumpster.
I don’t have much hope for Watson. There are just too many bodies to sell at IBM.
