One of the government’s leading researchers and research administrators has been recognized with a new outside post. Dr. Simon Liu, the administrator of the Agricultural Research Service at the Agriculture Department, is now a fellow of the National Academy of Public Administration. He joined the Federal Drive with Tom Temin to discuss his work.

Interview Transcript:

Tom Temin: And let’s get a quick update of what’s been going on at the ARS in the past year since we last spoke.

Simon Liu: Well, it’s something I will continue to do the research to respond to any kind of emerging issues that is going to impact American people everyday from field to table. Give you one example. Currently, we are facing challenge in terms of avian influenza. So as a result, we involve into all kinds of testing to make sure that the milk we drink good to drink and then also the steak we consume are good and make sure that our animal, the poultry side, the cows have a good, good mechanism to protect them.

        Get tips and tactics to make informed IT and professional services buys across government in our Small Business Guide.

Tom Temin: Yeah, food is really a system, isn’t it? Because by the time it gets to one’s plate, it’s really multiple sources and multiple influences. And not just that cow or that ear of corn.

Simon Liu: Exactly. So we need to protect them all different fronts. The crops, the fruit, the dairy products, the meat products, the fishery, you name it.

Tom Temin: And what is your human capital situation? Do you have all the employees you need or what kind of talent are you looking for these days?

Simon Liu: We are much, much better off right now with our hiring strategy. We have done that especially in the past year or so. So right now, our current agency, we have roughly about 7,000 strong employees, including 2,000 PSG research scientists and post-docs. So we are in very good shape.

Tom Temin: And I wanted to ask you also the interaction you might have with the food safety organizations in some ways. Research is a safety factor, but your upstream of consumption. And recently there was some outbreaks of I think, some kind of poisoning from fast food and that involved an agricultural product or a couple of them. And that’s not really the ARS to deal with. But is there interaction because it is a chain of events in many ways?

Simon Liu: Yes, certainly as we do research to detect any kind of pathogens or any kind of virus, especially the food systems here. We also conduct the research to develop different type of technology to detect those pathogens and to do the packaging of our food products or the processing of food products. So that make sure that especially in the upstream of the food supply chain from field to the farm gate and then to the stores. So those are the area that is a part of our research.

Tom Temin: And the ARS was undergoing, I believe, some “physical plant” updates. Some modernization of its infrastructure. Give us an update on that.

Simon Liu: Infrastructure is very critical to us. Currently, our agency, we have 95 research locations. So we have first class researchers. So we need to make sure our research laboratory, they are up to date. Currently, our average research laboratory, the age is about 48 years old. So we continue to push to modernize our research laboratories. So we really appreciate and grateful for the congressional support to continue this path.

        Read more: Management

Tom Temin: All right. We’re speaking with Dr. Simon Lu. He’s the administrator of the Agricultural Research Service, part of the USDA. And now you’re a fellow of the National Academy of Public Administration. And the fellow sometimes get called in to do real work. It’s not just an honorarium. What have they said you might be called to do in your area of expertise?

Simon Liu: Certainly I’m very, very excited to become a member of the National Academy of Public Administration. So as far as my expertise, I will say I have broad expertise in terms of my education background. I attend 12 different universities from three countries in many different fields, including, of course, agriculture. I keep learning every day. And then computer science, electrical engineering, education, national security and government management and business administration. So I have broad expertise, broad background. I have worked for five federal agencies and also the industry, the private sectors. So with that background, as far as the federal government, I have worked for NASA as a contractor for about 10 years and then working for DOJ, Treasury Department, NIH and right now, I’m with the USDA. So broad experience. Now, throughout the past, I would say 24 years, I’ve been serving as a member of CSC. So I have engaging in many, many committees across the federal government. The committee is an interagency committee and many different tasks to do and in that review, provide a suggestion as solution to them. So I’m excited to become the member And then so that through this kind of things, this membership, I will be able to participate even more. So one of my goals is to be able to contribute to the federal community and also state and local.

Tom Temin: Sure. And thinking about all of the places you’ve been and the different assignments you’ve had and the experiences you’ve had nowadays, there seems to be one thing that is cutting across all of these endeavors, and that is the emergence of artificial intelligence. Do you find that that’s something that could be or will be or is already useful to the work at ARS? And certainly a lot of agencies have decided, ‘Yeah, this is something we need to explore.’

Simon Liu: Certainly. As a matter of fact, artificial intelligence was my piece of patient 30 some years ago. So for the past working 30 years, I worked for those five departments. Part of my work is to develop an artificial intelligence system. And in good old day, we used to AI to gauge the communication system. At DOJ, the same thing, we use the AI to detect the same kind of crime or maybe the drug coming into this country. And then also to predict the crime, the area that we need to install or may be able to enhance our security system. The same thing in the Treasury. Used AI to help to direct the budget allocations. And NIH, I work for them. We use AI to help to advance the developing of the medicine. And then within ARS, I must say that we are very, very proud of working with my boss to take action. We established the first AI Center of Excellence within ARS five years ago. At that time, we envisioned AI will play a major role. So we established Center of Excellence. And looking at the use case of AI, our goal is to harness the power of the AI to advance our research and to speed up our scientific discovery. So five year pass.

Tom Temin: Then what happened.

Simon Liu: Then currently, I must say that we have more than 100 research projects. They leverage AI to advance their research in many different fronts, from the breeding to fuse safety to the smart of our culture irrigation system to deal with climate change and many different fronts. So more than 100 research projects currently among the 95 research locations around the country. We have over 60% of the laboratory. They are using the AI to advance their research. So I’m so very proud of the accomplishments that we have done in the past five years. Essentially, we’ll focus on four different areas five years ago. One is to make sure that we have a strong, strong business record use case. Second, as you know that the AI without data, there’s no AI. So we make sure that we formulate collecting our data, making sure that we manage our data properly. So those data will be able to train our system and then also make sure that we have the skill set. We cannot compete with our satellite, we cannot compete with Google or Meta or other high-tech company in terms of salary benefit. So we need to train our internal people. So we established a very, very strong training program to make sure that our people get enough training, get to know AI and begin to get their hands dirty, and then to experiment with AI. And the last thing is that we need to make sure that we have a good architecture from two fronts. One is management side, organizational structure. Make sure we have a good organization structure to be able to direct the AI research and also the technical architecture, make sure that we have a platform to do so. As you know that right now in the area is the GPU was so hot in the industry, right?

Tom Temin: Sure.

        Sign up for our daily newsletter so you never miss a beat on all things federal

Simon Liu: And we have used GPU, I would say more than eight years. About eight years ago, we began to use that. So right now, we have GPU as a part of our high-performance computing infrastructure. So five years pass, I must say that we have accomplished quite a lot. I’m so pleased. The kind of effort we have.

Tom Temin: And let me ask that.

Simon Liu: As a matter of fact, next week, the 22nd this month, we are going to have an AI conference. This AI conference hosted by Texas A&M. And we love their facility to invite the people. We have hundreds of people. We anticipate there’ll be 400 plus people coming in from our intramural areas, research scientists, extramural program, USDA agency, and also our collaborator in the university. So we are very, very excited about it.

Tom Temin: We’re speaking with Dr. Simon Liu. He’s the administrator of the Agricultural Research Service, part of the USDA. And a final question on AI. Since the five years that you have been deeply into this across the department, the emergence of the large language model or the so-called generative AI. That seems to be a step function difference between classical AI and now. What’s your assessment of the large language models and the generative?

Simon Liu: I think we also have done with generative AI now using some kind of pilot project to do so. We feel that it’s very, very useful. For instance, we use the regenerative AI to do our documentation, automatic indexing, make sure that we have tons of documents. Documentation need to be curated. And in the past, it was either a human being or semiautomatic. When I was the regenerative AI, we can do faster and then also cheaper. So it was high quality. So this is on the paper side, on the particle side, on the Norwich side and the same thing. In the past, we spend a lot of effort to collect data and in many cases, our predictive modeling itself. We face the challenge of having enough data. Right now, we will be able to use the regenerative AI to generate that dataset. So help us to fine tune our predictive model. So certainly regenerative AI, we have been done some experiment with some success. But we’ll continue in this endeavor. I do think you have a lot of potential for us in terms of helping us to advance our research.

Tom Temin: And also the idea of maybe taking the data that you have going back years and years and years that could be revisited with a large language model. And maybe there’s insights that will emerge that you may not be aware of.

Simon Liu: Absolutely. If you look at the winner of Nobel Prize this year, they are doing exactly just that. I do see this. Nobel Prize in Physics. Their invention is looking at the AI, using the AI and training the AI was huge among that dataset. And then the algorithm somehow will be able to identify pockets from the data, and the market will help us to advance our research. Certainly, you’re absolutely right. Right now, we are looking back at our climate science data. We’re collecting climate science data, climate change data in some of the laboratories over 100 years old. So that means that we have other years of history of climate data. We right now are going back to those data we collected. Try just like you say, try to identify the market. Help us to find the things to pair our research for the future.

Tom Temin: Kind of gives new meaning to the amber waves of grain.

Simon Liu: Absolutely. Yeah, certainly.

Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.