Killer Apps and the Game-Changing Potential of AI

Artificial intelligence has achieved some stunning victories in recent years. First, IBM’s Deep Blue trounced chess champion Garry Kasparov. Then, Watson defeated of two of Jeopardy’s top champions. And most recently, a sophisticated computer program called AlphaGo beat the world’s reigning Go champion in Seoul. Driven by big data, complex analytics, and increasingly sophisticated programming, AI is surprising even industry observers with its rapid advances. Can these technologies be applied to business and confer a competitive advantage?

Irving Wladawsky-Berger is uniquely positioned to answer that question. He spent 37 years with IBM, where he was responsible for identifying emerging technologies and monitoring the marketplace for developments that could affect the future of the IT industry. Since retiring, he acts as a strategic advisor for blue-chip clients, serves as a visiting lecturer at MIT’s Sloan School of Management, writes a weekly blog, and contributes regularly to the Wall Street Journal’s CIO Journal.

Wladawsky-Berger recently sat down with Martin Reeves, a BCG senior partner and the leader of the firm’s BCG Henderson Institute. Edited excerpts from that conversation follow.

What would you say are the limits of what is currently possible? What is it that we cannot do yet with artificial intelligence?

Throughout the history of computing, computing has always been very good at solving a problem when it’s relatively well defined.

The interesting question is, at what point can we start looking at our intelligent machines to invent their own games, to invent their own problems? Or, as you have written in your articles, to develop their own strategy instead of being superb tools for helping us solve a human strategy.

What would it take to convert the latest advantages of artificial intelligence into competitive advantage?

What is most important is to take the technology and make it part of an application that does something very useful, where the technology is the equivalent of the wires that make it work.

Let me give you an example. It turned out, in my opinion, the killer app for the Internet was this notion of customer self-service. Before the Internet, if you wanted to buy a car and didn’t know what car to buy, you had to get pamphlets, go to different car dealers, and get lots of information.

Once the Internet showed up and there were websites like Edmunds.com, we saw that you can do all the research yourself. By the time you went to get the car, you knew as much as the dealer—often more. I think the value to business is when the technology gets translated into something that people really appreciate without having to know the technology.

What might you speculate are some of the killer applications of artificial intelligence that will change the game for us?

A number of them are already there. For example, sophisticated searching that can put the search in a context instead of just using textual analysis. We’re so used to search, we don’t think of it as an AI application.

That’s a very simple application, but what always happens with computers is when you understand how the application works, then almost always it looks trivial. When you don’t know how it works, it looks like magic.

Navigation is a very sophisticated application that gets solved some place in the cloud with all kinds of interesting algorithms. I think we’re going to keep making more and more and more progress. Each step will likely feel incremental but when you look back, the sum total of these incremental steps will look phenomenal.

If I’m a CEO who wants to avoid jumping into the AI market too early but doesn’t want to miss the boat, how should I focus my efforts? What should companies be doing?

I think they need people out there experimenting, doing things, knowing what’s going on. I always tell people to hang out with the smartest people they know, because knowledge and ideas are absorbed by osmosis. You talk to people and you get ideas. Now, they won’t know what ideas they’re giving you that apply to your business. But you’ll know, and you’re the one who will connect what they’re saying to some problem in your business that you couldn’t solve before—and that now you might be able to. I would absolutely advise companies to experiment in the marketplace.