Rep. Don Beyer, D-Va.
Bill Clark | CQ-Roll Call Inc. | Getty Images
WASHINGTON — Don Beyer is no ordinary student at George Mason University. He is 73 years old. He prefers a notebook and pen to a laptop for taking notes. He is also the top lawmaker on AI policy in Congress.
The Virginia Democrat was fascinated by AI, but the breakthrough came when he discovered he could enroll in computer science classes at George Mason University. So he enrolled, starting with prerequisite classes to eventually earn a master’s degree in machine learning.
Beyer is only able to take one class a semester to balance voting on the floor, working on legislation and fundraising, and completing coding homework. But this class is already delivering benefits.
“I think every time I take an additional course, I get a better understanding of how real-world coding works,” he recently told CNBC. “What it means to have large datasets, what it means to look for these associations, and perhaps what it means to have unintended consequences.”
Beyer is a member of nearly every group of members of Congress working on AI. He is vice chair of both the bipartisan Congressional Artificial Intelligence Caucus and a new AI task force launched by the New Democratic Coalition, the largest group of moderate Democrats in the House.
He is also a member of former House Speaker Kevin McCarthy’s AI task force, which could be reinstated under House Speaker Mike Johnson. On the legislative front, he is the leader of the next bill. Expanding access to high-performance computational tools needed to develop AI.
Short-term intensive course
This fall, as members of Congress scramble to get the latest information on AI at hearings, forums, conferences and more. During a dinner with Open AI CEO Sam Altman, Beyer said his time in the classroom helped him understand what’s going on under the hood.
He is also learning how easily small mistakes can have a big impact on the code. Beyer said one of his daughters, who is also a programmer, sent him a thick book on debugging programs that was “very long.”
“You make a big mistake, and then you make a stupid little mistake, and it takes hours to figure it out, and then you realize how imperfect the technology is,” he said. “So there will be a lot of effort put into the downside risks of AI.”
Congress is grappling with how to advance AI.
In the House, Rep. Jay Obanolte (R-Calif.), who served with Beyer on McCarthy’s AI task force, told CNBC he spoke briefly with Johnson (R-La.), who chairs the AI group. He said he is interested in reopening soon. After more pressing battles, such as government funding, are over.
Obernolte said there are several directions the House could go in terms of AI, including enacting digital privacy protections for consumers, whether a new federal agency should oversee AI, or whether each monetary agency should address the issue. He said this includes deciding whether or not to treat the situation.
Obernolte, who holds a master’s degree in artificial intelligence, said there are no shortage of smart lawmakers on AI, including Beyer.
“Don is great, very knowledgeable and really passionate about this particular issue,” he said.
“Time is of the essence”
Another issue Congress is looking at is videos and photos that appear real but are generated by AI, especially of events that never happened or of real people saying things they didn’t actually say. The ease with which videos and photos showing people’s actions can be spread can ultimately influence elections.
Rep. Derek Kilmer (D-Wash.), chairman of the New Democratic Party’s AI Task Force, said the 2024 election will bring renewed pressure to find ways to minimize the impact of misleading and false media. He said it would bring a sense of urgency.
“The impact of the spread of misinformation on public discourse and the health of our democracy is significant,” Kilmer told CNBC. “And that’s what’s driving this.”
Senate Majority Leader Chuck Schumer of New York recently said that “time is of the essence” when it comes to handling AI-generated videos and photos. He said, “In terms of legislation and building guardrails around AI, that may be the first thing we need to do.”
Still, Beyer worries that Congress won’t respond quickly enough to the rapid pace of new AI models.
“What we’re trying to do is not repeat our mistakes on social media, where we’ve had no regulation at all for over 20 years,” Beyer said. “While social media has had some great positive effects, it has also had some pretty scary downsides: misinformation and disinformation.”
Bayer acknowledged that Congress is unlikely to be able to pass an AI bill this year because of the fight over spending and the House speaker’s gavel. But he remains hopeful that something could happen next year in the run-up to the 2024 elections.
