MIT Tries to Rescue IBM Watson
January 21, 2016
“Don’t Blame Watson for IBM’s Slide” is a remarkable example of “real” journalism. I love it when academics (you know the innocent party in the student loan maneuver) defends a really big outfit (IBM).
The write up focuses squarely on the problem IBM has created; for example:
If you’ve seen IBM’s advertisements or have read the proclamations that the company is making a big bet on Watson, its famed “cognitive computing” engine, you might be tempted to think the gamble is failing.
I know that IBM has made a serious miscalculation with the Watson play. First, HP is learning the hard way that search is not something that generates tons of dough without some serious management expertise. How is HP solving its conundrum with regard to the $11 billion bet on mid 1990s technology? HP is just going to fly like the legal eagles. Now legal eagles don’t like a sparrow in the flock. You may be able to guess who the winners will be in the HP solution.
IBM has embraced Lucene. That’s a good move. Palantir has done the same thing. IBM then failed to learn from outfits like Palantir even though IBM owns i2 and could have pursued a similarly focused and clear headed approach.
Nope. IBM made Watson into what I consider Jack Benny Show material. Watson does game shows without realizing that post production wizardry makes a “win” look like the mouse clicks of a film school grad. Watson does cook books. Watson cures cancer. Watson solves insurance woes. Watson does not generate the type of revenue that will make a dent in IBM’s revenue needs.
But IBM has achieved one thing. IBM has made cognitive computing the skateboard on which more nimble outfits are riding. Perhaps IBM should charge these start ups for the psychological assistance the wonky Watson PR campaigns have delivered without a prescription or a juicy hourly fee.
The write up reports in “real” journalistic style:
… Even if Watson had become a big business by now, IBM would still be in huge trouble because of trends that have been afoot for a very long time—notably the rise of cloud computing services that have diminished the need for large organizations to buy IBM servers and mainframes. This was in play long before Ginni Rometty was named CEO in 2011, but her predecessor, Sam Palmisano, was better able to mask the decline and keep Wall Street happy by selling off unprofitable lines of business, buying high-margin software companies, and returning billions of dollars to investors through dividends and share buybacks. Now there are fewer financial levers left to pull. Revenue has been falling for 15 quarters in a row.
Yep, blame Sam. Blame someone.
The reality is that Lucene, acquired technology, and home brew software are not going to do lots of stuff.
Back to Palantir. Whatever the company’s faults (and there are some Doozies), Palantir aced IBM in the intelligence sector. Palantir focused and then used indirect sales methods to move into financial services and health care.
IBM buys ads and does PR. Palantir gets others to push the product. I wonder if anyone at some of the banks know what a “helper” is or how to create and “object” in Palantir land.
Probably not.
But IBM has demonstrated that it lacks focus, has no effective strategy to make search and information access generate big money or in Palantir’s case, big flows of venture capital.
What IBM can do, however, is get the “real” journalists to play along. After 15 consecutive quarters of revenue excitement, IBM needs to find a solution.
Hint: i2 can help. Will IBM listen? Nah, it’s more fund thinking up ways to hire Bob Dylan to explain Watson or visualizing Watson as chemical structures.
Stephen E Arnold, January 21, 2016
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
