Obernolte Discusses the Future of Artificial Intelligence and the Role the Federal Government Should Play
WASHINGTON, DC – With each new week seeming to bring another new advancement in Artificial Intelligence, The Ripon Society hosted a dinner discussion Monday evening with one of the leading experts on AI in the U.S. House of Representatives.
The expert was Congressman Jay Obernolte (CA-23). First elected to Congress in 2020, Obernolte is an entrepreneur and computer scientist by training who has spent the past 30 years running his own video game development company. He is also the only member on Capitol Hill with a master’s degree in AI.
“I think it’s never been clear to people before just how disruptive AI is going to be,” he stated in remarks to kick off the discussion. “And it’s definitely going to be disruptive. Technological innovations always are disruptive, going back most recently to the internet, but also all the way back to the printing press. But you know, disruption is not necessarily pejorative. Disruption can be a good thing.
“Think about the fact that in the entire history of the United States, every single major expansion of our economy has been preceded by a substantial gain in worker productivity. It is a hard and fast rule. You need that gain in productivity that leads to a gain in corporate profitability that leads to gains in compensation. It leads to investment gains, it expands the entire economy, and makes the pie bigger for everyone. So, every single time that’s happened in the history of the United States, it’s been heralded by a gain in worker productivity.”
“I believe that AI is going to be the next big catalyst to economic prosperity, leading to not only a growth of prosperity, but a decline in poverty and the things that AI can catalyze, like universal education, the eradication of poverty, and being able to get our arms around the problem of homelessness. I think that AI can help with all of these things. So, I’m an AI optimist.”
Because of his experience and background, Obernolte has been given a leadership role on this issue on Capitol Hill. He serves as Vice Chair of the Artificial Intelligence Caucus, and has also been tasked with leading a House Task Force that will examine both benefits and risks of AI, and what, if any, guardrails may be needed down the road.
“If we look at the pantheon of the things that we have to guard against,” the California lawmaker observed, “they’re substantial, but they do not include things like an army of evil robots with red laser eyes rising up to take over the world out of the Terminator. That’s not what keeps me up at night. What does keep me up at night are things like the way that AI can be used to pierce through digital data privacy and re-aggregate personal data that was supposedly disaggregated, and then use that to build models of human behavior that they can be used to influence behavior and influence public opinion.”
“That is a very real risk that we need to be aware of and that we need to guard against. I worry about the potential of AI to be used for the spread of mis- and dis-information. I worry about the fact that AI, because it is such a productivity enhancer, can be used to enhance the productivity of people that do not mean well.”
Beyond the possible risks of Artificial Intelligence, Obernolte said there are also a host of “legal and creative questions” that need to be answered. They are questions, he believes, that it is the responsibility of the federal government to address.
“When AI is used to create something, who owns the copyright for that work?” he asked. “When copyrighted material is used to train an AI algorithm, who owns the AI algorithm that results from that training? How much transparency is required in that process of training, and what compensation is due to the people who created the material that’s used to train the AI?
“These are all unanswered questions that I think the federal government has a responsibility to answer. These questions are going to arise regardless of whether or not we in government act. If we do not fulfill our responsibility to clarify the legal landscape, we’re just going to ask the courts to intervene in something that the courts have no precedent for intervening in. We need to do our jobs and make sure that we come up with thoughtful, deliberate, well-researched answers to those questions.”
What we do not need, Obernolte added, is for Congress to follow the example being set overseas.
“The European Union has gotten way out over its skis in trying to regulate AI,” he stated bluntly. “The result has been to increase pressure on Congress [to act] … If we go down the wrong path, if we overregulate, if we set up a legal framework that makes it impossible to employ AI to increase productivity, we are never going to get to the pot of gold at the end of the rainbow. All of the different benefits that we talked about, government has the capacity of killing if we do it wrong.”
“I want to take a thoughtful approach to regulation. To me, when we talk about regulating, we should never regulate for the sake of regulation. We always should have an end result in mind. So when you look at AI and you ask me what do I think our goals are, I think the role of government is definitely to mitigate against consumer harms. That’s something that’s a well-established role of government. I think people across the political spectrum would agree that that is an appropriate role of government. To me, that means we have to understand what harms we want to prevent from occurring. We just mentioned some of them. Once we agree on those things, we can craft a regulatory framework that guards against those potential harms.”
According to Obernolte, this framework will not be created overnight – something else, he noted, that regulators in Europe do not seem to grasp.
“Another thing that I think is really important to recognize is that this is not the work of weeks or months or even years,” he stated. “One of the things that the EU is doing that’s potentially dangerous is pretending they can craft this AI Act — they can put it together, debate it, pass it, and then think that the problem is solved. AI is moving so rapidly, that it would be a fool’s errand to pretend that we can pass an AI Act and be done with our job of regulating it. I think that incrementalism needs to be our approach … And so, I see the task of regulating AI as not just passing one bill, but passing 10 bills this year, and 10 bills next year, and 10 bills a year after that.”
Obernolte concluded by making one final point – because the development of AI technologies is so expensive, lawmakers on Capitol Hill not only need to take the time, but need to work together to get it right.
“We can’t expect developers to make that kind of investment and take that kind of risk if the perception is that tomorrow Congress could take an action that would make that investment worthless,” he warned. “So we need to make sure that whatever we craft is durable and it doesn’t change with a small change in the balance of power between the House and the Senate or the House and the White House, which means it needs to be broadly bipartisan and it needs to be broadly bicameral. That’s the challenge ahead of us.”
To view Congressman Obernolte’s remarks before The Ripon Society on Monday evening, please click the link below:
The Ripon Society is a public policy organization that was founded in 1962 and takes its name from the town where the Republican Party was born in 1854 – Ripon, Wisconsin. One of the main goals of The Ripon Society is to promote the ideas and principles that have made America great and contributed to the GOP’s success. These ideas include keeping our nation secure, keeping taxes low and having a federal government that is smaller, smarter and more accountable to the people.