For IBM, Watson’s Success Is Not So Elementary

Darryl K. Taft covers IBM, big data and a number of other topics for TechWeekEurope and eWeek

Some people poured cold water on IBM’s New Watson AI Unit. Daryl Taft says the project needs time to prove itself

Last week IBM launched a new business unit to commercialise and begin to capitalise in earnest on its Watson cognitive computing technology.

Big Blue announced its IBM Watson Group at aNew York function. However, just prior to the launch there was a report (in the Wall Street Journal)calling the systems, software and services giant out for falling short on its effort to monetise the Watson technology.

I’m not an IBM apologist, but I believe that while the WSJ took what appears to be a well-intentioned look at what IBM has been doing with some of its early Watson projects in health care and financial services, it seems a bit unfair to criticise the company for not having made a ton of money on something it has consistently maintained was still in its early phases.

IBM WatsonCognitive computing? It’s still a learning experience

In any conversations I had with IBM executives about Watson, they always said commercialisation efforts had not been finalised and the early projects were learning experiences for both the clients and IBM. Nobody ever tried to claim that Watson was going to ring in the dough by the beginning of 2014. Were there high hopes for the technology? Sure, but there was also a lot of caution and conservatism about when it would be ready to present to the mainstream IBM customer.

In short, IBM played the Watson effort like a startup inside the company. Many startups take years to make any kind of money — indeed many carry on for a few years in stealth mode until they can get traction enough to make a splash.

Watson got its initial exposure by beating human competitors on the Jeopardy game showin 2011. IBM chose to get that exposure but afterwards, operating in stealth mode was not an option.

The company is not without its faults, but judging Watson now is premature. There was only one Watson-based product before this week—the Watson Engagement Advisor, announced last May. In my opinion, IBM didn’t go on the clock with Watson until November when it gave third-party developers access to Watson via the Watson Developer Cloud and announced an ecosystem for the technology.

Golden boy takes charge

Then this week came the new IBM Watson Group, with a new location in Manhattan’s Silicon Alley and an IBM golden boy in the name of Mike Rhodin to run it. Rhodin has a string of successes at IBM and if CEO Ginni Rometty put him in charge of this newly created unit, it’s because the company is serious about it and needs a proven leader in that position. To jump-start the operation, IBM has committed $1 billion (£610m) —$100 million of which will go to building up the ecosystem—and 2,000 employees, with one-third of IBM Research dedicated to Watson.

Watson JeopardySo don’t judge IBM’s success with Watson since Jeopardy back in 2011; we can begin to judge it since late last year. Don’t throw shade on the success of an effort that has only just been given the company’s blessing. Let’s have a look at where things are this time next year.

“The formation of a dedicated Watson business unit and subsequent investments in the suite further indicate IBM’s intent to transition from an outsourcing company to an outcomes-based vendor, a transition that TBR expects will happen over the next two years,” said Matthew Casey, a software analyst at Technology Business Research (TBR). “Commercialisation of Watson is the next logical step for IBM, based on the company’s investment in the platform to date and the increasingly disruptive nature of cloud. However, this will be a challenging task considering the cost and complexity to the end user of adopting Watson technologies.”

IBM has long been the big dog in enterprise IT. It may no longer be the case that you can never get fired for buying IBM, but if you choose IBM for your enterprise systems, you are typically in a solid position. Sure, in many new technologies such as big data, analytics and the cloud, IBM has serious competition. But it is not falling down on the job.

“IBM is placing big bets and big money on transforming computer interaction from tabulation and programming to deep engagement,” said Forrester senior analyst Michele Goetz, in a statement. “If they succeed, our interaction with technology will truly be personal through interactions and natural conversations that are suggestive, supportive, and as Terry Jones of Kayak explained, ‘makes you feel good’ about the experience.

“There are still hurdles such as expense, complexity, information access, coping with ambiguity and context, the supervision of learning, and the implications of suggestions that are unrecognised today. To work, the ecosystem has to be open and communal. Investment is needed beyond the platform for applications and devices to deliver on Watson’s value. IBM’s commitment and leadership are in place,” Goetz continued. “The question is if IBM and its partners can scale Watson to be something more than a complex custom solution to become a truly transformative approach to businesses and our way of life.”

So, yes, IBM faces challenges with Watson. But let’s give them a chance before we grade any progress.

How is your IBM knowledge? Try our quiz!

Originally published on eWeek.