While there's a lot of excitement about the potential of a natural-language analytics platform such as IBM's Watson supercomputer, all the hype surrounding Watson throws a spotlight on the need for much more sophisticated approaches to data management.
In order for something like Watson to truly add value in a business context, it needs to be programmed like any other machine. Unfortunately, not much of the data in a corporate environment is all that well managed, notes Planet Data Solutions CEO Howard Reissner.
Planet Data, which specializes in data discovery, has a lot of experience with data management issues. Not only will something like Watson first require all the relevant data for an application to be collected, it then has to be stored, cleansed, formatted and then extracted into a metadata server. This is one of those famous IT tasks that is a lot easier said than done, notes Reissner.
The reality is, says Reissner, even with Watson, if you wind up putting garbage in that's still what will come out. So before anybody gets too excited about eliminating the need for certain classes of professionals such as lawyers, Reissner says businesses should take stock of what something like Watson can realistically do and what will really be involved when it comes time to first set up something like Watson and then actually keep it relevant.
One day Watson might reduce the number of hours required to research a case, which would lead to fewer billable hours and probably fewer attorneys. But that's not the same thing as eliminating the need for an attorney to actually argue the case. Similarly, Watson could reduce the size of the medical staff needed to research and diagnose ailments, but that's hardly the same thing as eliminating the need for doctors.
We live in a world where platforms such as Watson offer great potential. But as Reissner notes, there's a lot of extremely complex work that needs to be done in terms of making Watson do something that has more value than simply winning a few games of "Jeopardy!."