Lance Bellers - Fotolia

News Stay informed about the latest enterprise technology news and product updates.

Government AI strategy includes tech transfer to private sector

The Department of Energy's chief commercialization officer works with the DoE's national laboratories on transferring AI and other technologies to the private sector.

CHICAGO -- The Department of Energy's first chief commercialization officer, Conner Prochaska, says the U.S. is positioned to win the AI technology race -- and believes it's critical it does, for the good of the world.

As director of the DoE's Office of Technology Transitions, Prochaska oversees, among other things, the commercialization of projects related to government AI strategy.

In this Q&A, conducted at the DoE's InnovationXLab Artificial Intelligence Summit here, Prochaska talks about government AI strategy and how basic science evolves into commercial products. He also reflects on the U.S. rivalry with China over AI, and how he sees the U.S. government AI strategy as embodying democratic values.

A former Naval intelligence officer and trained lawyer, Prochaska likens the AI race to the nuclear arms race of a previous era, and says the U.S. is committed to winning the battle. The U.S. is well positioned to do so, Prochaska says, with assets such as the DoE's new Artificial Intelligence and Technology Office, created in response to President Donald Trump's executive order on AI and exascale supercomputers. Here, he discusses the government AI strategy and supercomputing potential.

How do you move advanced technologies such as AI out of the public sector and into the private sector, and how is that part of the government AI strategy?

Headshot of Conner Prochaska, the Department of Energy's first chief commercialization officerConner Prochaska

Conner Prochaska: Let me take a step a little higher than AI. When we talk about the Department of Energy, and its history in the weapons program -- the origin of the department is the Atomic Energy Commission -- and this vast complex of 17 national labs, how do we get that to where it actually makes an impact?

There's a few steps from the basic research we do in the department to a product or a thing or change in the world, similar to NASA. At the Department of Energy we spend roughly $18 billion a year in research and development. Most of that research is basic science. There's not an end, like NASA does research to go to Mars, or how the Department of Defense does research to protect soldiers, sailors, airmen and marines and win wars. The Department of Energy does research for the betterment of humanity -- not to sound overly romantic about it. But there's a step beyond just the research and how we get that 

In the history of the department it was like, 'Well, somebody will come find us. We'll publish a paper and someone will come.' Now we're trying to make it a little more purposeful and less accidental. How do we make sure it's not some firewall between us and industry, and how do we grow that impact? Now bring that to artificial intelligence. When we talk about artificial intelligence, we scan our giant complex of 17 national labs and facilities; do we know everything that's going on? We're doing a lot, but it's pretty disparate across the department. Let's bring it together, so we do have a more purposeful path to artificial intelligence [and other technologies].

Can you talk about how your office and the national labs are involved in government AI strategy and the race for technology and AI supremacy with China?

U.S. taxpayers are footing the bill, they get the first benefit ... but we believe ... we are going to make the world a better place. And putting up the appropriate guard rails, we're not trying to weaponize science.
Conner ProchaskaChief commercialization officer, DoE

Prochaska: That's our mission. We're in a race. But we do want to be good partners. We want to share with the world.

At the end of the day, we believe that [AI and technology] should be lived, created and embedded with American values. We've got the tools, and we want to do this. U.S. taxpayers are footing the bill, they get the first benefit, but we believe that through that, we are going to make the world a better place. And putting up the appropriate guard rails, we're not trying to weaponize science. The scientific community is a worldwide community; there are no borders. But we have to acknowledge the reality, that some people have weaponized it and have turned it against us; we just have to pay attention. It's a similar argument to nuclear energy. The world is going to get nuclear energy one way or the other. They're either going to get it from us, or they're going to get it from someone else who doesn't care as much about security, nonproliferation and keeping people safe. Same thing with artificial intelligence.                                                      

How much of the government AI strategy plays into the work of your office?

Prochaska: When we talk about the Office of Technology Transitions headquarters, every lab also has a tech transfer team or commercialization team or an IP team, but our role at headquarters is to ensure the labs have the tools to be as effective as possible. Across the federal R&D system, how do we do better about getting stuff out of the lab and into the market? We do it through policy, we do it through access, our lab partnering, and finally promotion, which is what you're living and breathing right here, at our XLab [conference] on AI.

Can you talk about your office's work with national labs and how the labs are your main conduit for technology R&D?

Prochaska: We set the direction, we tell the labs what we're interested in doing, and the labs are the factory floor, [and] the customer is the U.S. taxpayer. We want to make sure we benefit the people who are paying the bills first, and then get that out to the greater world. Our national labs are the engine, and headquarters is the Department of Energy; 16 of the 17 national labs are doing artificial intelligence.

How important is exascale supercomputing to the DoE's technology strategy?

Prochaska: You can't put a figure or a thought process on it. It's invaluable. No place else is there a concentration of this kind of computing power. We already have four of the top 10 fastest supercomputers. And let's not forget those that aren't in the top 10 that we have and control.

Really the question is, how do we use these supercomputers? In 2021, the first exascale computer, here at the Argonne lab, quickly followed by one at Oak Ridge National Laboratory, and then quickly followed by one "El Capitan," [at Lawrence Livermore National Laboratory]. That computing power, we don't know what it's going to unlock. That kind of computing power has never existed.

How do the economics of technology transfer work? Is the government ever part of a commercial undertaking?

Prochaska: It depends. The federal government has no desire for business. We do put some restrictions on it in that we don't want it to go to a potentially adversarial country or organization, but we're not in the business of owning IP, we're not in the business of retaining IP. The 17 national labs are contractor run and those labs may retain some IP as a company in that they license it out, but the federal government doesn't have a dog in that fight.

The conference was held Oct. 2-3 at the Drake Hotel.

Dig Deeper on AI business strategies

Join the conversation

1 comment

Send me notifications when other members comment.

Please create a username to comment.

What role do you think is most appropriate for the government to play in AI research and development?
Cancel

-ADS BY GOOGLE

SearchBusinessAnalytics

SearchCIO

SearchDataManagement

SearchERP

Close