• September 28, 2020

Live from IBM Insight2015: Whither Watson?

IBM Watson

IBM’s Watson has been on my radar for quite a while now.  For long I have been quite confused as to what Watson actually is. It appeared to be a strange multi-shaped thing with a varying definition, depending of how you looked at it. The example application and services were sometimes quite exotic; I still have a card on my desk from last year with a recipe for creating a beer based mousse, courtesy of Chef Watson, which can generate entirely new taste experiences. Other aspects were rather elusive, such as when I tried to find out how to build my own application and got the impression that it would be too hard for me to do my own knowledge management.

Fortunately this year at Insight I did get a lot of answers. At the expo I found out that Watson is actually a portfolio of applications that share a common concept, but are often quite disparate. So it becomes quite essential for any aspiring Watson user to understand what these parts are and not let yourself be confused.

The actual value proposition (which up to now had been quite vague to me) has been starting to emerge in the form of some real applications that were on display . I’ll pick out three of them here.

One is wine4.me, a personal wine recommendation application. It uses Watson to learn about your tastes, and then recommends wines that you might enjoy. By rating wines you’ve already had, the app will learn more about your preferences. There is no need for specialized sommelier language–just tell it what you liked, what you didn’t like, and it will be able to recommend other wines for you. This is great for casual wine drinkers like myself who never could get over the idea that wines might taste like raspberries or pencil shavings.

Another very convincing application of Watson technology is Ivy by Go Moment. Ivy is an SMS based messaging app that engages with hotel guests the moment they check in. A little bit like Cortana or Siri, it acts like a concierge, managing requests like show bookings, taxis, or information on events in town. Most importantly, it can identify if there is anything the client is not happy about and pass these comments on straight to the hotel management for a human response, well before the guest leaves a bad review on TripAdvisor. There are a lot of smarts in that application that Go Moment, unlike Microsoft or Apple, would not have been able to build themselves. Instead they leverage Watson technology, which goes a long way to prove its value.

But the big boy here is the combination of Watson with the services of The Weather Company (which IBM just acquired, excluding the TV channel). Weather is big data and it is real time. I can’t say how big its big data is, but I suspect that they are in the top tier. It might seem strange that IBM, an IT company, would buy a weather forecasting company. But then maybe it’s not.

First, weather data is an information resource that touches all of us, and therefore is valuable to all of us. IBM showed examples of how the American Red cross uses weather data to predict resource allocation needs to deal with extreme weather events (get a text that says: “there is an active tornado warning for your area”). Airlines use this information to predict turbulence and to optimize fuel consumption. Roofing companies use it to predict mid to long-term supply demand.

So for IBM, the purchase of The Weather Company is basically a case of buying the equivalent of fuel to drive its predictive services part of the company. It does make me wonder if this is just the beginning for IBM’s predictive services arm to buy other such businesses. Who might be the next one? Or rather not who, but what, as I think that the Internet of Things will become the biggest driver of supplying information that needs to be managed–the amount of data is potentially so huge we will almost have to rely on analytics capabilities such as Watson.

And that starts to make me think that the “traditional” data science approaches of using R, Python, Scikit etc. may have a limited time in the sun. In only a few years, they may be relegated to a more niche position, with total analytics environments becoming the de facto standard where analytics development and application is driven by conversation between users and recommendation engines similar in concept like IVY above but more conversant in statistics and data science.


Stephan Meyn 18 Posts

Stephan is a Professional Services Consultant for the Legasuite range. HE enjoys being a polyglot and works with other technologies: cloud computing, text analytics, machine learning, and big data are all part of his activities at Rocket Software.


  • Avatar

    Ian Reply

    October 29, 2015 at 3:10 pm

    very interesting post Stephan. Didn’t make insight this week but must look up some pieces

Leave a Comment

Your email address will not be published. Required fields are marked *