Aneesh Chopra was the Secretary of Technology for Virginia before becoming the first Chief Technology Officer of the United States. Today he is the president of CareJourney, a company that uses open health data to provide patient-specific insights and workflow automation for care organizations. Tweet him at @aneeshchopra.

From The Commons: Aneesh is also the author of Innovative State: How New Technologies Can Transform Government ––required reading for anyone working in government technology and innovation who wants to understand the origins of a lot of the work we do. We sat down with him to ask about the early versions of digital innovation in government, and where he thinks the field should head now.

What did technology and innovation look like in government in 2006, when you first joined?

When Tim Kaine became Governor of Virginia, he asked me to be Secretary of Technology. I was given a $3 million innovation fund to test ideas that the government hadn’t tried before. I had a choice: in my few years there, I could hit one or two projects out of the park, or I could do a lot of small things that created a culture of innovation, inspiring others to take the ball and run with it once I was gone. I see moments when going deep in a couple areas and nailing them is exceptional. But my personal preference is to run many plays, even if a lot won’t go anywhere.

Tell us about a small play that worked really well.

The Chair of the Virginia Board of Education was running a campaign for GED attainment. I met with him and asked a naive question: “How do you get a GED?” He said you could go to a community college, you could go to a high school, or an adult center, you could do it on your own. All I heard was “on your own,” so I asked about that.

It turns out the Virginia Board of Education had bought 26 instructional videos. You could rent them out on VHS, learn the GED, and take your test. The DIY model was super cheap, but really outdated. I said, “Nearly everyone in Virginia has cable TV, do you think we could put the GED course on demand?” And his eyes lit up and he said, “Wow, could we do this?” I said, “Why not?”

In less than 90 days, we got the two biggest cable companies in Virginia involved. The governor filmed an intro and we rolled it out for free. There was no procurement, and we didn’t need a digital services team. Thousands of people started watching the course. It was just something so simple, because the Chair of the Board of Education was passionate, but didn’t know there was a technical path to get there.

So then you became the first CTO for the U.S. Did you take the small plays approach again?

I took a different tactic. We had this SWAT team, which was me, Jeff Zients, and Vivek Kundra. Rahm Emanuel, our Chief of Staff, wanted to use us to solve a bunch of problems. In 2009, he saw a New York Times story that the VA couldn’t process benefits for the GI bill. It was September, and veterans who thought they could go to college and get their benefits were homeless, not getting food, and dying in the streets. Rahm read that and called an emergency meeting with the SWAT team, even though we were supposed to be working on open data in government, not serving as an emergency fix-it squad for tech issues. Rahm said, “Get on Air Force Two.” We flew to a processing center in St. Louis, with no context or history of the problem, and just had to fix it. It was a tragedy at every level. The people processing applications had 25 different applications open at the same time. They were looking up one piece of data here, one there, and one there, then manually copy-and-pasting into other applications. Each request took three hours to process, even though they only needed six pieces of information. It could have easily been automated.

We created and launched a solution, but the IT culture at the VA was at the corner of toxic and hateful. So while the technology worked, it wasn’t celebrated, and it wasn’t scaled. That’s why the SWAT team approach isn’t the best strategy for this work. You can create a single program like that, but without a culture to reinforce it, there’s no sustaining it.

You’ve worked in and around government technology for a while now, and have had some time away from the White House to process. What approach do you think is most effective for solving the kinds of intractable problems we tend to face?

If you have an outcome objective— you want to boost third grade reading scores, you want to lower unnecessary hospitalization rates, you want to advance clean energy deployment, whatever it is— the government has a few role options. It can be as an operator, providing direct service. It can act as a regulator, setting standards. Or it can act as a convener, encouraging the industry to do something on its own. That last approach really invites more stakeholders to the table. That’s that’s what animates my time and energy these days— it’s just on the other side of that exact same problem solving coin.