FORTUNE — On May 15, Fortune senior writer Jessi Hempel interviewed IBM (IBM) CEO Ginni Rometty as a keynote for the National Venture Capital Association’s 40th anniversary conference, Venturescape. What follows is an edited version of their conversation.
Fortune: IBM was once about mainframes, and then PCs and printers. Now IBM is about services, software, Watson. How do you think about the company?
Rometty: Two years ago, IBM had its 100th anniversary, which is when people asked that question the most. And I think one of my biggest learnings has been, never define yourself by a product. I would like us to be thought of as an innovation company. The only way you survive is you continuously transform into something else. It’s this idea of continuous transformation that makes you an innovation company.
Now, an innovation company is something that we all aspire to be. You are aspiring to be that with more than 400,000 employees. What does it even mean to be an innovation company especially at that scale?
I’ve got a formula in my head about this idea of continuous transformation, and maybe it’s helpful for people as you build out these businesses, because I break it into five different pieces.
I think the first thing you always have to do is keep reinventing yourself for high value. I think particularly in our tech industry, this is an industry that has violent innovation and then commoditization, and it’s a cycle of innovation/commoditization.
You guys [venture capitalists] play a big role for us in that because you acquire, you divest, and then you remix your own development. But that acquire and divest is a really important piece. You know, for us, we have divested $15 billion. But then, we have acquired 140 companies. So, the formula is, in part, move to higher value.
The second thing you’ve got to always do is keep thinking about how to make a market, and I think you can do that by buyer, by category, by geography.
Then the third thing I think about is, you know, assuming you live to be a little bit older, you’ve got to reinvent your core franchises. Right? Things like middleware will get reinvented to mobile middleware, as an example. So, you reinvent.
And then I think the fourth you can’t forget, when you said about all the people, it’s the skills. So, you absolutely, you have no choice. And many people can reinvent themselves and some people can’t, right? So, you reinvent skills, and then at the end of the day, I think what we all end up doing is you’ve got to keep looking at yourself, the company, and reinventing the company, which is a bit of when we were talking about how does the company change.
MORE: Google Android’s enterprise problem
Instead of making direct investments in companies, you guys work closely with the VC industry. How does your strategy work?
Let me share with you kind of what we look for and why when we look at an acquisition, because it’s a vital piece. In fact, it’s so important to how we think of remaking the company, we actually commit over a period of time. We do something called a roadmap, it’s a financial roadmap. And the five-year roadmap that started a couple of years ago ends in 2015. We’ve said we’ll acquire $20 billion of companies.
So, we’re pretty clear about what are the areas I talked about some that we’re focused in. So where we’ll do acquisitions, they’re adjacent. They’ll always cluster around strategic areas.
The second, they’ll typically always be intellectual property. I’ve got a distribution system that goes to 170 countries. If I acquire properly, you know, you may be successful in one or two countries, or one place; I can scale, and that’s part of the value that IBM brings.
In your Annual Report, you talk about an analytics process that you use for acquiring.
Yes. So, every part of your business will change based on what I consider predictive analytics of the future. So, we did this for all of our acquisitions. We used to look at 300 to 500 things on every acquisition. Okay, well, that wasn’t necessarily good for speed, by the way. We did a lot of work with our own Research group on the analytics to be able to predict three to five factors individually by acquisition that are going to make the biggest difference.
Ginni, for many of us, Watson is that Jeopardy! Game; but when you think of Watson it’s a much bigger thought, right?
Yes, she has done way more than that! It’s about the coming of a third area of technology, right? You know, the original systems counted things. The next set of things were programmable. This thing learns, right? You give it very little instruction and then the more information it has, the more it learns.
So, since the time that you would have seen it on Jeopardy! when it could answer simple questions and it beat the best humans out there, we put it to work for medicine. And we have been working with Memorial Sloan Kettering, MD Anderson, Columbia, some of the finest institutions in this world.
Medicine is one high-value area for Watson. Might there be a role for venture capital there? There might be, in areas such as interfaces and certain specialized areas.
But here’s what’s more interesting. That’s a high-value area; we are also now about to come out with Watson in what we consider as an advisor, and it will be in volume around research-oriented industries. Think of things like pharma; or as a client advisor in industries that have huge numbers of end retail clients. And so, think of financial services, think of a telco.
And Watson’s a service; we will launch an ecosystem where Watson’s a service and you build applications around it. And you have to have domain expertise. That’s what will be your value of the future.
You call this period a “golden era of technology.” And you mentioned this idea of social information as the new production line. What do you mean by that?
I am very aggressive internally in IBM in moving into a social enterprise. It flattens organizations. It enhances their speed. We have hubs, as an example, now around some of our key clients around the world that attach everyone, irrespective of where they are in an organization, in a way that you have come to know how social networking works.
And I envision a day even with all your employees where it’s more important what you share, not just what you know … It’s not so much what you say you know; I care about what the world thinks you know, what clients think you know.
And then maybe there’s a day that you’re paid that way, based on what people think and appreciate and what you share. It’s a very different paradigm than today. So, I believe this idea of being the social production line of the future is how many, many companies will operate, particularly in a global environment.
MORE: Blackberry is about to give away its last advantage
So, to that end, Ginni, you have in your tenure so far at IBM been very aggressive at moving internally into the social tools. What have been the biggest challenges as you do that, or the unexpected things that you’ve learned?
Some of the wonderful side benefits have been my ability to communicate two ways with the organization pervasively and quickly is beyond compare, and to have a two-way dialogue. That’s not just a one-way push; that is, you learn things very fast. You can take layers out of an organization, right? You can push it down.
But more important, the other reason I am so intent on this is, as you look at what we are all hiring, I often call it the millennial generation that is the way they work.