By Thomas Scott
Union Alumnus Steve Mills ‘73 spoke at the Nott Memorial last Tuesday, April 23 about his work at IBM, touching on topics such as the advent cognitive systems, like IBM’s Watson, as well as how crucial infrastructure should not be connected to the web. The Concordiensis caught up with Mills shortly beforehand for an exclusive interview.
Thomas Scott: You’re the Senior V.P. and Group Executive of Software & Systems. What exactly do you do in that capacity?
Steve Mills: IBM is a hundred-billion-dollar company and our company is divided up across various product lines that we have. We’re in the hardware business, the software business and the services business, which is essentially labor-based services of different kinds: building things, creating things for customers. I help with the whole product side of the company. I manage the development as well as the marketing and sales for all of the hardware and software products that we make.
What are some of your biggest challenges in that area?
Information technology is a very competitive market. So lots of competitors, lots of companies that we have to go up against every day. We do business in 170 countries, so the global reach of the company is always a challenge and of course skills is frankly one of the biggest issues we grapple with. Both the skills of our customers, their ability to implement the technology that we provide them; as well as the skills of the IBM people that have to help them implement the technology. It’s very much a skills-based business.
Did you have anything to do with Watson?
Yes … [the] Watson project was done in IBM research. It uses a variety of technologies all of which I pay for.… I have a team of people that work for me that are actually commercializing that product offering for a whole range of uses.… It’s very much a part of the portfolio that I manage.
What would you say IBM’s role in things like analytics are? Google takes part in sifting through massive amounts of data. How is Watson going to be part of that new regime of data mining so to speak?
IBM is a very big business measured in many billions of dollars directly associated with the software to do data analytics, big data, data bases those kinds of things. It’s an important area for IBM from an investment perspective. Over the last five [to] six years, we’ve invested over $20 billion in this area. It’s a big investment for us and a big part of our business. Companies like Google do a little bit different thing than we do. So a product like Watson is very specifically focused on trying to provide the right answer the first time, whereas as Google users we all know we get lots of answers and we have to then choose which one we think is the most relevant. And maybe we search again, search again, search again in order to find what we’re looking for. Google serves a very different purpose, because in most casesyou are interested in seeing more of the possibilities. There are many application scenarios where you actually don’t want to see too much. If you give me 20 answers, it doesn’t do me any good, right? If I’m trying to do a better job of diagnosing a disease, 20 or 30 answers is probably more than I want to deal with. I want the top three. I want statistical probability. I want accuracy. So they’re really different domains.
Do you have a particular favorite implementation for Watson? I know there was something about how [Watson] was used to actually help diagnose medical conditions, for example.
We have a number of medical projects underway around Watson and we think those probably, long term, will be among the most transformative. But we’re also in the midst of doing some very small projects with Watson as well, including helping companies with call centers and really using it as a decision aid tool for people doing everyday jobs where you want the computer to learn what that domain is. You think about any call center you would call to get service on a device, a product, something you would want answers to. You want the person on the other end of the phone to give you the best possible answer to your question. And very often they struggle in learning the job and therefore Watson tool can actually learn the job and then act as an aid to them to prompt them what the right answer is.
What did you study at Union?
How did that shape you’re experience at IBM?
I like to tell people I use my degree every day, since I deal with people all day long. Having been a psychology major certainly doesn’t get in the way of what I need to know and what I need to do. My experience at Union was very much about learning to learn. I’ve obviously learned a great deal more since leaving college than [when] I was in college because I’d been out for 40 years and therefore hopefully I’ve learned a lot in that 40-year period. Union was a great experience in terms of helping me build the necessary skills to then go out and build a career.
When did you really get into [computing] and how did you find your way to IBM?
Thankfully I was able to secure an interview and I managed to get through the interview and get a job offer and I took it. And then the rest was study, study, study. I spent a lot of years at IBM learning the craft because I didn’t have that craft when I graduated college.
So what was your first position at IBM?
I started out in sales. So I had to learn the product line to be able to go out and meet with clients and sell the products. So I was a 22-year-old kid selling $12 million computers to big businesses.
What do you see as the future of computing?
Well… it is the transforming technology of the last 60 years and it will continue to transform the world around us, no question about it. It’s the ultimate tool, and human beings are tool-builders. The computer is the tool that helps us with our most powerful muscle, the one in our heads.