X

IBM research honcho: From the Pentagon to the 'toy shop' (Q&A)

Zachary Lemnios, who was U.S. assistant secretary of defense for research and engineering, gave up his Pentagon role to become IBM's VP of research strategy. CNET asks him to compare the roles.

Daniel Terdiman Former Senior Writer / News
Daniel Terdiman is a senior writer at CNET News covering Twitter, Net culture, and everything in between.
Daniel Terdiman
6 min read

Since September 11, 2001, the American security apparatus has been focused largely on stopping terrorists from striking again. But some feel a more pressing danger may be that of cyber attacks, digital hacks that disable critical infrastructure and bring society to a crawl.

As the U.S. government has tried to shape its approach to such attacks, President Obama and the secretary of defense have relied on contributions from a number of people in the Pentagon and elsewhere for ideas on how to stop bad actors, be they from national governments or small terrorist groups. Among them was Assistant Secretary of Defense for Research and Engineering Zachary Lemnios, whose job was to help with Obama's national security plans, and whose main areas of focus were big data and cyber security.

Former U.S. Assistant Secretary of Defense for Research and Engineering Zachary Lemnios has joined IBM as vice president for research strategy. U.S. Department of Defense

Last Monday, Lemnios -- who previously had top jobs at DARPA, Ford, MIT's Lincoln Laboratory, and other large companies -- traded in his government paycheck for one from the private sector. Named IBM's new vice president for research strategy, Lemnios will take on responsibility for bolstering Big Blue's global research agenda, helping to plan and design the company's famous Grand Challenges -- which in the past have included Deep Blue, the first computer to beat a chess world champion, and Watson, which beat the best players in the world at Jeopardy.

CNET talked to Lemnios on his fifth day on the job, asking where he thinks America's defenses are against cyber attacks, how IBM compares to DARPA, and much more.

Q: Former U.S. spy chief John McConnell said that the U.S. faces a cyber 9/11 if we're not prepared. Given your past experience, what's your assessment of the U.S.'s readiness against cyber attack?
Zachary Lemnios: Cyber security is not an area that's going to be dominated by a government play. This is an area where industry, government, and the private sector have to take partnership roles, and it's where the Defense Department is pushing those ideas. One of the biggest challenges we have in cyber security is that it's hard to know when a network is secure. Most of what's done in the private sector, you have absolutely no idea when you're under attack. You know it afterward. So most of the tools we have might have provide some sort of alert, and there's great tools that law enforcement has to do forensic analysis. But there's very little that's done to project a threat or to project how a system might be vulnerable. There's a line of research and development that's going to push towards cognitive computing so that a system understands when it's under attack and how it might be able to deploy defenses in a measured way. For me, that's a really interesting field, but we're very early in the research stage.

Given the importance of what you were working on at the Defense Department, why did you choose to leave for the private sector?
Lemnios: I was there as a Senate-confirmed political appointee. I came into the department in July, 2009, and all political appointees evaporate at some point in time. For me it was just the right time to leave, for a variety of reasons. I looked at IBM and decided on here because first of all, there's a remarkably rich and talented technical staff. Secondly, there's a clear vision of where the company wants to go. And thirdly, there's the resources to actually deploy ideas, services, and products globally. So for me, it was a natural progression.

How would you describe your new role?
Lemnios: It is to drive the research strategy for IBM, to make the research investments that IBM makes more relevant, to accelerate the impact, and to accelerate the deployment of those ideas. And in any corporation, IBM especially, you want to be looking far enough out, but you don't want to look so far that those ideas never come to fruition.

You're going to have some major responsibility with the Grand Challenges. Looking back at Deep Blue and Watson, which were so successful, that must be a little challenging to take that program on. What does that feel like?
Lemnios: I'm in a toy shop. This is amazing. There's really good people, clarity, and there are people at IBM that really understand the impact of technology. For IBM, it's not a bumper sticker, it's really part of the DNA here. Think about going from atoms to applications in short time. When you look at Watson, the Jeopardy event was a remarkable culmination of really deep research and really great integration. But what's really exciting to me are the products that's going to spawn, and the way IBM will use the results of Watson in the medical field, in education, and in thinking about how we might apply the learning and the reasoning elements of Watson to other fields, and to change and open new businesses.

Are there are any particular areas that you'd like to see addressed with Grand Challenges?
Lemnios: It's doubling down on cognitive computing, big data, maybe mobile applications, maybe thinking about how do we actually intersect these fields. When you think about managing large data structures, and doing it in a cognitive way so the system understands -- you get a new piece of hardware, and it actually understands you -- I think that there will be lots of plays that we make over the next several years that intersect many of these individual themes.

Will you have authority to create new Grand Challenges?
Lemnios: It's a team effort. One thing that I've learned in the few days that I've been here is that not only is there rich talent, but there's a remarkable spirit of collaboration across IBM's global laboratories. It really is an ecosystem where ideas get matched up by bringing the right people together, and the depth is here to do that.

Do you have a portfolio for changing things in the way IBM does its work, and the direction of the research, or do you have to work within longstanding structures?
Lemnios: I don't think anything's static. If there's one thing you can bet on, it's that this is all about accelerating the impact of research, and by doing so, you have to think about the processes. It doesn't seem to me that IBM is one of those corporations that worships the infrastructure. It really is about making an impact, and if that means changing the way innovation occurs, if it means changing the ecosystem in terms of how you bring people to the table, both internally, and with customers, that's all part of the mix.

I would assume that with your background and experience, you're coming into IBM with an important personal and professional network? I'm curious, how will you work to help IBM work in concert with the government to continue supporting new technologies and disruptive efforts?
Lemnios: Yeah, I've got a pretty healthy network. I actually looked at that very carefully from a post-government employment perspective. I have no procurement restrictions with the government. I have restrictions that apply to the Obama pledge I took that prohibit me from soliciting and actively engaging with a government employee for two years, but that said, I can have lots of discussions on a broad set of topics, and will, with folks here without regard to going back and meeting with my former colleagues at DoD and across government.

You worked at DARPA, so what kinds of comparisons can you draw between DARPA and IBM?
Lemnios: I ran two of the offices at DARPA -- Microsystems technology office, and I stood up the information processing technology office -- and within IPTA, we funded a little application called Siri. I also funded a little bit of the early work on Watson, so I was pretty familiar with IBM. The difference is that DARPA is a sponsoring agencies, where you have about a hundred program managers who have all come in with great ideas. Their leverage is moving money to contractors who actually implement those ideas. IBM is a little bit different. IBM has people with good ideas, but also the resources, and the facilities, and the laboratory infrastructure internally to implement and execute those ideas.

You now work for a company, and your responsibility is to the shareholders. But do you still feel you have a responsibility to the American people, having spent so much time in government?
Lemnios: It was an absolute honor to be able to serve in the Pentagon and provide technology to our men and women in harm's way. Part of my DNA is still in that building, and where I'm called upon to help by the department, I'm more than willing to do that.