Editor’s note: This is a lightly edited transcript version of Barbara’s recent Optimistic Outlook podcast conversation with U.S. Congressman Jay Obernolte: Subscribe to the podcast on your favorite platform.
Representing California, Congressman Jay Obernolte is chairman of the House Science, Space, and Technology Committee Research and Technology Subcommittee, and a member of the Committee on Energy and Commerce, before which I testified in February.
Last year, he spearheaded the bipartisan House Report on Artificial Intelligence that explores responsible AI adoption across sectors.
Congressman Obernolte is also the founder of FarSight Studios, an American video-game developer.
Barbara Humpton:
Please give our audience the 30,000-foot view of how you see the opportunities and risks for AI in American businesses.
Congressman Obernolte:
AI is going to be completely transformative not just for American business, but for American society. And, unfortunately, I think we've done a very poor job of articulating the optimistic case for AI. We as Americans have had our opinion about the downsides of AI colored by the last 50 years of science fiction and pop culture. So, if you ask someone about AI, you'll get something out of the Terminator films where the army of evil robots is rising up to take over the world, but you don't get the optimistic case.
And the optimistic case is this: First of all, AI is already the most powerful tool for the dissemination of human knowledge that mankind has ever invented. It can teach you anything that you want to learn about, in whatever learning style is optimal for you. Second, it's going to completely democratize access, not only to information and education, but expert opinion. And that is going to completely change our society.
But what is also happening is that AI will shortly become the most powerful tool for the enhancement of human productivity that mankind has ever invented. And that is going to be incredibly empowering. I believe that that is going to create a rising wave of prosperity that literally lifts all the boats in America.
Barbara Humpton:
You led the bipartisan effort in Congress to produce the House report on artificial intelligence, exploring AI use across multiple sectors from healthcare to finance to agriculture. And the report also offered a vision for responsible AI adoption, the innovation and governance that has to be balanced as this technology enters our world. When you think about the report now, what are the highlights that really jump out to you?
Congressman Obernolte:
Our task force was charged with the job of coming up with, by the end of 2024, a report proposing a federal regulatory framework for AI. So, I had 24 very talented and dedicated members, 12 on each side of the aisle. And we had 25 hearings last year and delivered a 270-page report in December that fulfilled our charge.
Just hit the high points of this report: Americans are under the misperception that AI is unregulated in our country. That is completely untrue. We have adopted a very different approach than other entities like the European Union in embracing decentralized sectoral regulation.
What that means is that we believe in regulating outcomes, not tools. AI is a very powerful tool, but at the end of the day, it’s just a tool. And when you're thinking about the risks of AI deployment, which is what you have to understand to be able to talk about what our obligation is as regulators to mitigate those risks, it matters very much what you're going to do with that AI. An algorithm that is unacceptably risky in one usage context—for example, AI in a medical device that's going to be implanted in someone's body, to name a particularly high-risk deployment—could be completely benign in another usage, like in a video game. The sectoral regulators know this.
We do already have AI being used in contexts like medical devices. In fact, a lot of people are shocked when I tell them that the U.S. Food and Drug Administration (FDA), which obviously is our regulator that has authority over the regulation of medical devices, has already processed over a thousand applications for the use of AI in medical devices.
So, one of the things in our report that we are embracing is to continue this decentralized, sectoral approach to regulation. Here's the reason why.
The European Union has taken a very different approach. They've centralized everything; they have their sectoral regulators, but they're creating a new regulator for AI. As a result, you're going to need another license from that regulator for anything but the lowest-risk usage of AI.
We think that that's the wrong approach because we think it's a lot easier to teach the FDA, for example, what it might not already know about AI than it would be to teach a brand-new entity everything the FDA has learned over decades of ensuring patient safety.
That's true of other sectoral regulators, too. The National Highway Traffic Safety Administration (NHTSA) is already dealing with the use of AI in autonomously driven vehicles. The Federal Aviation Administration (FAA) is dealing with the use of AI and aircraft avionics. The list goes on and on and on.
Barbara Humpton:
I think that makes a world of sense. What's also interesting is when we think about it from an end-user's perspective, often we're not even aware of the AI's existence. Look at examples of the mapping algorithms that we use, or, frankly, just the simple search capabilities we've all become accustomed to. This is the first technology that actually comes to people instead of people having to go to the technology to incorporate it.
Congressman Obernolte:
It’s interesting that no one has come up with a federal definition in statute for what AI actually means, or what AI is. And that's because AI is in the eye of the beholder. Is it a neural network? Is it an algorithm that makes a computer seem like a person? Everyone has a different definition.
To me, AI is anything that you interact with a machine but seems like you're interacting with a human. When you boil it down to that, you can see why AI is going to be so impactful, because it's going to create a way that allows us a very human interface for dealing with complicated technology.
Barbara Humpton:
I was thrilled to have the opportunity earlier this year to testify before your committee to talk about AI and manufacturing. You were really delving into the impact of AI on manufacturing. And we had been really trying to draw a distinction between what we called, at the time, “consumer AI”— the AI that we're using in our private lives, in our commercial lives, in our entertainment lives—and the AI that's being introduced into the world of manufacturing, where we have control over the data that's used to train the foundational models, and where we have control over how that is used and over human intervention in the manufacturing environment. The day I testified, you followed up with a question that made clear you were thinking this way toward the sectoral approaches.
Congressman Obernolte:
And it's an excellent point that a sectoral approach to regulation allows you to differentiate between those two uses, because the thing about industrial AI is that the environment it's operating in is very constrained. If you imagine AI controlling a couple of robot arms on a manufacturing floor, with the caution tape laid out around it to make sure humans don't walk in there, and the robot arms are building a widget, that's a great example of industrial AI.
What we as regulators would do when evaluating the risks of that is to ask, “What are the worst things that can happen?” What are, as the lawyers say, the parade of horribles? And what are we going to do to mitigate those? Well, because the environment is constrained, the things that can go wrong are much more narrow. As long as you are doing quality control on the widget that the AI is building, you've really controlled for a lot of the risks of AI deployment in that case.
If you contrast that to a more consumer application of AI, like in a chatbot, you have no idea about risks. Could the chatbot help someone build a biological weapon? Could it cause a teen to adopt a suicidal ideation? Could it exhibit some really troubling racial biases? Because the deployment of AI in the consumer context is so much more open-ended, we have to worry about those things, whereas in industrial AI it's much simpler to get your arms around.
A balanced vision for regulating AI
By: Barbara Humpton, President and CEO, Siemens USA
To me, AI is anything that you interact with a machine but seems like you're interacting with a human. When you boil it down to that, you can see why AI is going to be so impactful, because it's going to create a way that allows us a very human interface for dealing with complicated technology.
Barbara Humpton:
I've often heard you talk about the importance of public-private partnerships for advancing American AI leadership. Looking at this landscape, one of our national goals is to maintain that leadership role. As I recall, your task force’s AI report has a great section discussing the importance of government, university, and industry ecosystems for R&D. I'd love to hear your view of how businesses and government can be working together to develop and expand these ecosystems.
Congressman Obernolte:
I think it's going to take all of us working together and pushing in the same direction to maintain U.S. dominance in AI development and deployment. One of the things that we are very worried about is that there will be a lack of competition in frontier AI development. It's gotten so expensive in terms of computing power and data to develop a frontier-AI algorithm that we're concerned that only the biggest tech companies will be able to have the resources to do that.
That’s very unhealthy, in particular, when it comes to academia and entrepreneurialism, because those are two of the forces that have really driven advances in American scientific discovery over the last hundred years, and we want to make sure that that tradition continues. When you talk about academia, in particular, having research done in academic settings comes along with some very desirable things like transparency, publication, and peer review. Those are things that you lose if that research is done only behind closed doors by the biggest corporations.
One of the things that we say in the task-force report is to recommend the codification of a pilot program that we're currently running called NARE, the National AI Research Resource, that makes pools of data available to academics and to entrepreneurs to make sure that such research continues to be done outside of the big corporations. This is very important.
And, by the way, to make the point—given that I'm a small-government conservative myself—all of these resources are donated. All the federal government does is to be the gatekeeper in arranging access to these resources. That's because the purveyors of these resources share our goal of making sure that entrepreneurs and academics continue to have access to them.
Barbara Humpton:
It's this innovation ecosystem that has truly brought the U.S. to this point of global leadership. Now, let’s switch our focus for just a minute and talk about the impact AI will have on people.
You've heard me say in my own testimony that I believe AI is not going to replace people. We know that every wave of new tools and technology that have been introduced have actually expanded what's humanly possible.
I'll share with you one quick story about the use of AI in my company’s operations.
We had researchers working on how to put a copilot into a manufacturing environment. In a factory in Germany, where there was a first and second shift operating, the researchers installed the new copilot, and people started to use it. In satisfaction surveys, the feedback from both first and second shift was great. But the researchers noticed something odd in the data. Second shift was using it more than first shift. Why was that?
So the researchers actually did interviews with the second shift personnel, who said, "In the previous times, if something had gone wrong during second shift, we had to call the experts—here it is, the middle of the night, and we don't want to bother them, so we use the copilot.”
This is the kind of thing where, yes, AI is absolutely changing the way we work, but it's also truly making people more productive and raising the level of expertise. Are you hearing stories like this?
Congressman Obernolte:
Very much. We have a whole chapter in our task-force report on education and workforce issues. And the point that we make very strongly is that we think that, as you say, AI is going to create a lot more jobs than it displaces, but it will be disruptive as new technologies always are.
Going all the way back to the advent of the printing press, technologies are disruptive, but they have created far more jobs than they have disrupted. We, as a society, are going to have an obligation to, first of all, equip our young people with the tools that they need to succeed in an AI-powered workspace. We need a commitment to reskilling because there will be some jobs that are displaced, and people are going to need help to be reskilled in other fields. We're going to need to beef up their social safety net to make sure that no one falls through the cracks.
But at the end of the day, we are going to create more jobs than we displace. I think that a lot of the angst around this topic is misplaced. I'll give you a specific example from my home state of California.
We had a very disruptive strike by Hollywood scriptwriters last year. One of the central issues of the strike was that the script writers were objecting to the use of AI by movie studios for the automated writing of movie scripts. Kind of unspoken in this was the assumption that greater use of AI would create fewer jobs for Hollywood scriptwriters.
I thought, how short-sighted is that? When I look at the future, ten years from now, I envision turning on Netflix and watching a show that's created by generative AI in real time just for me, based on what I feel like watching today and what I've enjoyed watching in the past—rather than watching a show that was created once for everybody. We're going to have the technology to do that. And I'm sure that there are going to be innovative entrepreneurs that take advantage of that.
That's many, many more jobs for scriptwriters. And, yes, scriptwriters will know how to use AI to empower their productivity. That holds true of lots of different fields.
I'll give you another example. There is an annual survey of graduating high-school seniors. One of the questions is, “What career do you most aspire to?” When I was growing up, I was at the end of the Apollo space program, and the number-one career that young people aspired to was astronaut. We all wanted to be an astronaut. Very few of us actually became astronauts, but it drove an entire generation of Americans into the STEM fields, which was incredibly useful for our country.
Can you guess last year what the number one answer of graduating high school seniors in America was?
Barbara Humpton:
I don't have a clue.
Congressman Obernolte:
Social-media influencer.
So, ignoring for a moment the somewhat disturbing things that answer says about our young people, think about this: Before the advent of the internet, how you would articulate to someone what a social-media influencer is, what they do, and why they provide a value to a society that people will pay them for?
Imagine how difficult it would be to have that conversation. You don't have the vocabulary unless you understand the internet and the way that has brought people together and connected people and the way that people live their lives through the internet now, right?
So, fast forward—there's going to be an entire spectrum of jobs that are created because of AI. We don't even have the vocabulary now to describe what those jobs are going to be, but we can look back through history, not just the internet, but the Industrial Revolution, the printing press—all of these technological revolutions have created displacement, but also enabled human success and created many, many more jobs than they've displaced. I am 100-percent confident that's going to be true of AI as well.
Technologies are disruptive, but they have created far more jobs than they have disrupted. We, as a society, are going to have an obligation to, first of all, equip our young people with the tools that they need to succeed in an AI-powered workspace. We need a commitment to reskilling because there will be some jobs that are displaced, and people are going to need help to be reskilled in other fields. We're going to need to beef up their social safety net to make sure that no one falls through the cracks.
Barbara Humpton:
I've been often asked by parents what advice I would give their children. What should they study in order to be prepared for the future? And I've been saying to people: tell your kids to play their video games. What do you think about this theory?
Congressman Obernolte:
I would say: Become accustomed to being able to articulate to a machine how to solve a problem, and then having the machine go do it.
We used to call that computer programming. I think that now when we say programming, we are referring more to what my generation used to call scripting. That’s a much higher level. When I started programming, it was in a very low-level language called “assembly language,” which is the language that the chips speak. Later I got higher-level languages like C, and then object-oriented languages.
With AI, it's going to be another layer above that where you'll be able to procedurally articulate to AI what you want it to do. The people that are the most effective in using AI will be the ones that are the best at articulating what they want the AI to do and how to do it. And those are going to be incredibly useful skills to have.
But let me also pair that with something else. As someone who's hired thousands and thousands of people in my career, both in the private sector and for the last 20 years in government office, I can tell you this: Please, kids, learn how to write. That is one of the most in-demand job skills right now. But young people are not graduating with the ability to be able to take something that they want to say and communicate it. And that's still going to be very important with AI.
This is an interesting point because there's this huge angst among schoolteachers about kids who use AI to write essays. It's almost like the debate that we had when I was growing up about whether or not we should be allowed to use calculators to do long division. And we were told, no, you have to learn this arcane process with decimal points and how you carry things. But we had the last laugh because it turns out now everyone carries a calculator around with them as part of their cell phone. Probably no one knows how to do long division anymore once they become an adult.
AI will be similar. We’ll come to appreciate that there are things that AI does for us that free us up to think about concepts at a higher level. So instead of thinking about how you’re going to articulate a particular point, you'll be thinking about how to persuade.
You and I are having a conversation. How am I going to convince you that my point of view is right? I want to tell you a story; I want to tell you this voyage of discovery that I took to try and get you to come along with me. And then I might use AI to help me with the nuts and bolts of making my case. The people that are going to be most successful using AI will know how to do that.
Barbara Humpton:
The 1980s version of me that used computer punch cards is absolutely embracing this new era we're entering. Beyond that, I wonder whether, on some level, a liberal-arts education come back into vogue.
In speaking to educators, we’re learning about personalized education. Each of us can have AI tutors who learn our learning style and then are able to bring us along a path at our own speed. There are just so many possibilities available to us that it brings me to what I love to ask each of my guests on the podcast, because I love to learn from my guests about their optimistic outlook.
As a member of Congress, you have an incredible opportunity to influence American innovation. What are you most optimistic about in the era of AI?
Congressman Obernolte:
I’ll first address the question you raised about a liberal-arts education. And my answer is: yes and no.
AI is going to completely revolutionize our approach to education. For the last 150 years, however, we have embraced this model that says that when you're young, you go acquire this thing that we call an education. You do that by starting in high school, and then you go to college, and then maybe you keep going to graduate school. Then you graduate and you use that education through the rest of a career.
That model, through the lens of AI-empowered learning, is completely outdated for two reasons.
One, we are going to have the opportunity to be lifelong learners. We can learn anything we want, anytime we want. So why on earth would we devote all of this time when we're young, just learning stuff, when learning itself is something that we're going to do our whole lives?
Two, there is going to be the necessity to embracing lifelong education. The pace of technology is just going to continue to accelerate, and there is going to be no possible way that one can acquire at the very beginning, in youth, everything one needs to know and not have a commitment to learning as they go along. The statistics say right now that the average senior graduating high school this year will have three careers in their lifetime—not three different jobs or three different employers, but three different careers, which is an incredible thing.
But you asked for the optimistic case. So, think about this:
In the 250 years that the United States has been a country, if you go back and you look at our economic history, you'll find that every major expansion of our gross domestic product has been heralded by a corresponding increase in the productivity of our workforce. The two are directly tied together. The productivity of the American worker fuels the growth of our economy.
For the last six years, since the beginning of COVID, we have been in a gradual decline in worker productivity. I think that AI is going to be the next catalyst for a huge explosion in productivity that not only lifts our gross domestic product in our economy but lifts the economic fortunes of everyone in America. That is an incredibly powerful thing.
Published: August 26, 2025