How IBM Plans to Serve the 'Making Better Decisions' Business Market

NEWS ANALYSIS: 'Cognitive to the Core' is an appropos unifying line for a massive company with 375,000 full-time employees scattered all over the world.

IBM.SF4

LAS VEGAS--IBM is always looking at new markets to enter, but the most important one it sees right now is one you won't find the analysts or investors discussing.

Big Blue calls it the "making better business decisions" market. That may sound like an amorphous marketing term, but IBM considers it real enough to have designed an entire multibillion-dollar enterprise product and service strategy around it.

This strategy, of course, all starts with the IBM Watson machine-learning franchise and trickles both up (into the IBM Blue cloud) and down into hundreds of follow-on software products and services. Every product or solution on the company's menu now touches analytics in some way/shape/form, because this is what it perceives business will want in the next decade of the 21st century. Enterprises increasingly want their IT to guide them into making the scientifically correct business decisions that will eventually enrich themselves and their investors.

If one had to put an all-inclusive headline on IBM's statement to the world March 21 at its Interconnect 2017 conference, it would have to be "Cognitive to the Core." It's an appropos unifying line for a massive company with 375,000 full-time employees scattered all over the world.

Enterprises Need to Do More Than Just Understand Their Data

"For all of you who don't always look at market sizes, the market for IT is $1.4 trillion. But the market to make better decisions is $2 trillion," IBM CEO and President of the Board Ginny Rometty (at left in photo, with Salesforce CEO and founder Mark Benioff) told a full-house crowd of 10,000 during the main keynote March 21. "That's why you need that range of capabilities (of business intelligence, machine learning and cognitive computing).

"The second thing is, if you're 'cognitive to the core,' it can't be just looking at your data and understanding it. You almost have to use all of the senses that you and I have. The ability to see, to feel, to hear, and to read. This is what Watson does."

Rometty said Watson's ability to look at images "is better than any other AI engine out there.  We can describe what's in an image, or if it's a melanoma--95 percent accuracy on telling you what it is. Our ability to listen with Watson is almost human. What percentage of words do we miss in the course of a human conversation? The answer is five percent. We just finished a study that reported Watson has a 5.5 percent error rate--almost human."

People have to be able to read, Rometty said, and so do computing systems. "Watson reads 12 languages right now, which represents 90 percent of the internet population. And what I mean by feel and touch, we are the leader in motion, in patterns and sensors. So whether that's an escalator, a ball bearing, an elevator--that kind of feel," Rometty said.

Why Domain and Industry Knowledge is So Important

But the most important attribute about Watson that differentiates itself is its knowledge of industries and domains, Rometty said.

"If you look at all the data in the world, 20 percent can be accessed by a public search engine. Eighty percent of it sits with you," Rometty said. "It is proprietary. But all the value that we talk about is sitting in that 80 percent, for the most part.

"So Watson has been trained by experts to understand the language of an industry and understand that data. When I say trained by experts, I think this will come back to us in the decades ahead; it will matter. For example, 20 of the greatest cancer centers have helped train Watson in that (sector). That matters, because you will come to ask one day, 'Who trained this system? Where did that data come from?' You wouldn't go to your tax preparer for your oncology information," Rometty said.

"This is how we're using Watson to make the IBM cloud the best one for cognitive workloads."

Photo credit: Chris Preimesberger, eWEEK

Chris Preimesberger

Chris J. Preimesberger

Chris J. Preimesberger is Editor of Features & Analysis at eWEEK, responsible in large part for the publication's coverage areas. In his 12 years and more than 3,900 stories at eWEEK, he...