Ed. Note: Neil Bartlett was CTO and Head of Analytics for risk management systems developer Algorithmic when it sold itself to IBM in October 2011. With the acquisition, he became CTO & Head of Development for Risk Analytics at IBM Canada Ltd. This April he was one of eight IBM employees to be named IBM Fellows, becoming one of a very exclusive group of only 246 IBMers so honored in the 50 year history of the program. Shortly afterwards, SiliconAngle had the opportunity to interview him. This is an excerpt from the nine page transcription of that interview.

SiliconAngle: This is an honor….

Mr. Bartlett: Yes it is, and as a member of a new acquisition I am only just starting to realize how huge a privilege it is…. Prior to joining IBM I was used to being the spokesperson for Algorithmic on the technology and risk sides combined. IBM is a big company, and one of the attractions of joining it was the many assets, ideas, communities, people, software processes that we could access to make a better offering. The problem is that now you are a very, very small voice in a big organization. So one thing that this has changed for me is now I get back to being a bigger voice, and I get the resources and air time to talk about the ideas we want to promote.

SiliconAngle: Okay. So what are those messages?

Mr. Bartlett: Well, one of the key messages we are trying to get out there is an expanded view of risk management. Risk management as a discipline is siloed. We are trying to break out of that silo, and to do that we need the richer see of assets that IBM has. Those assets are scattered in different parts of IBM, so having a stronger voice means that it is easier to pull them together.

Then our longer term goal is that risk management is about how to manage uncertainty, and uncertainty is a key issue in the world. Every organization, and these days even individuals, need to understand how the future is going to play out. The future is not a stable state, there is a lot of change, and the issue is managing that. A lot of the techniques surrounding risk management can be applied to managing uncertainty, but … how do you have a conversation with those tools,… and how do you gain trust [in them]? That is a real emerging problem in the world.

There is so much emphasis on things like Big Data and analytics that are beginning to deliver information that is very, very important from vast amounts of data. So the Big Data initiative is exceedingly important to the world. The next phase has to be how do I engage with the models … such that as a human I can gain trust in that information. To take that and turn it into something that is leverageable by mere mortals in small companies that do not have true risk management teams I think is giving an important message to the world.

SiliconAngle: That is a very interesting idea. The new kinds of analysis around Big Data are qualitatively different in that they are forward looking and focused more on identifying change rather that the traditional rear-view mirror analysis that presumes that tomorrow will be like yesterday. What are the biggest problems in achieving this vision of actually using forward-looking analysis to anticipate and mitigate risk rather than reacting after the fact?

Mr. Bartlett: Risk quantification is an area that we have been deeply involved with for many years. But quantification is only one part of the issue. Many of the reasons for any form of risk management are social, and a lot of the issues are around the softer side of how the processes themselves work in terms of people, how the people can start to understand what the measurements are telling them and actually practice risk management that embodies what the measurements are saying. So it gets back to … how do people get an understanding of the uncertainty in the future and … trust that information & act upon it. … The biggest problem is that people weren’t really listening to what the measurements were telling them. So it is more about the people in the organizations than it is about the numbers.

It is a very complex issue, & that is why I love this space. It is not just a software problem, not just a math problem, not just a finance problem. It is all those things and more. And the rewards are huge, because is if you can stop the world from going bust or, even better, actually make the world work well, it is a great place to be.

SiliconAngle: That really is an inspirational vision. So how do you manage risk and, for instance, balance investing in preparations for events you know will happen that are not the end of the world, such as disk drive failures in a data center, against huge risks with a low probability such as say a super hurricane hitting the U.S. Northeast and putting half of New York City and coastal New Jersey underwater?

Mr. Bartlett: The way that gets solved today is on a statistical basis — mean time to failures and so forth. But ultimately how you would really want to be able to do that is to model the liabilities and the assets as well. You can model a data center or even a nuclear power plant, quite complex systems, and run them very well by looking not just at the failure of components but how the whole thing works. You can model them in financial terms and come up with the cost of dealing with extreme events. That is very similar to the types of modeling you would put in place for loss events in financial organizations. So there are techniques for taking that model and throwing really big stresses at it.

Let’s say I am considering building a data center in the E.U. What are the risks of putting it there versus putting it in Brazil or China or the States, factoring in things like the risk of political instabilities, for instance, that may or may not be there? Now of course the people who put data centers in do have models that do try to bring that stuff to bear. But how do you make that [modeling capability] freely available to everybody such that when you are choosing where to go on vacation next year, what are the risks associated with each destination and should you, for instance, buy travel insurance in case you loose your deposit?

SiliconAngle: Sort of risk-management-as-a-service for the masses. I see what you mean about making the world a better place. The IT industry and technology is evolving incredibly quickly with cloud, mobile, Big Data, incredible jumps in compute power, flash storage. What do you see for the future of analytics from the technical side? Is it, for instance, going the Watson route, so with the next generation people will be able to just ask a normal question out loud like, “Should we relocate our data center to Kansas or to Brazil?” And it will understand all the implications behind that question, do a complex set of analytics and come back with a real answer?

Mr. Bartlett: Watson obviously has a certain set of use cases that I think are very exciting. But I think the point is that computers will be helping out much more in making decisions in the next stage. What I am hoping is that we can stop focusing on the daily process of just dealing with the everyday risks and as humans start focusing on those outlying problems and keep our systems much more stable. Today when things are going well the system behaves itself. It is when big things go wrong that our systems don’t behave as they should. Fixing that requires more attention from humans. So I hope we are going to see computers take over more of the drudgery.

Now I really have to run, b it has been great talking to you. I have people knocking at my door.

SiliconAngle: Okay, thank you very much. I can see one way that becoming an IBM Fellow has changed your life. You are very much in demand. And I think I see what you meant by making the world a better place.