Pressure is building for Congress to regulate AI : The Indicator from Planet Money A corporate titan and a cautious professor face off before Congress and ... agree with each other? From the ivory tower to inside the boardroom, the pressure campaign to regulate AI.

For sponsor-free episodes of The Indicator from Planet Money, subscribe to Planet Money+ via Apple Podcasts or at

Bots, bootleggers and Baptists

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript




Right at the end of last year, ChatGPT got released. And since then, there's been a flurry of discussions and speculations and use cases. Congress, very quickly, has held a hearing on artificial intelligence. That was last week. This is a first step, really, for politicians to discuss AI's promises and risks.


SAM ALTMAN: Thank you for the opportunity to speak to you today.

WOODS: Behind this wooden table, we had Sam Altman. He's the CEO of OpenAI, which makes ChatGPT.


ALTMAN: But also that it creates serious risks. We have to...


And right beside Sam was a cognitive scientist named Gary Marcus.


GARY MARCUS: Thank you, Senators. Today's meeting...

MA: Gary is an emeritus professor at New York University and has written and talked a lot about the potential dangers of AI.

WOODS: And one thing that really stood out to us at this historic hearing was this line from Sam Altman from OpenAI.


ALTMAN: The U.S. government might consider a combination of licensing and testing requirements for development and release of AI models above a threshold of capabilities.

WOODS: A company asking for licensing, like some kind of FDA approval for AI products to ensure their safety. This is not the stereotype of business lobbyists out to slash red tape. This is the opposite. It's a CEO saying, give us red tape.

MA: And so on this point, Sam was in total agreement with Gary Marcus, the cognitive scientist. Together, they made this curious couple. You got a corporate titan and a cautious professor.

WOODS: But if you think about it, there is a logic there, and potentially it's one of self-interest. This is THE INDICATOR FROM PLANET MONEY. I'm Darian Woods.

MA: And I'm Adrian Ma. Big companies ask for regulation more often than you might think. Today on the show, a theory of strange pairings that might shed light on why OpenAI is calling for AI regulation.


WOODS: Bruce Yandle grew up in the American South in the 1930s and '40s, and he lived among two very different groups.

BRUCE YANDLE: Some of my best friends were Baptists, and some of my kinfolks were bootleggers. I knew both sides of the story. And it was a common expression - bootleggers and Baptists get together, and we get a dry county and the bootleggers make more money. And so that was just part of the culture.

MA: Fast forward to the 1970s. Bruce is working for the federal government, where his job is reviewing proposed federal regulations. Eventually, Bruce became executive director at the Federal Trade Commission, and he noticed that often, you would see unlikely coalitions getting together, pushing for regulatory changes, which reminded him of the bootleggers and the Baptists from his past - right? - these two seemingly opposing groups that actually both benefited from seeing the same rule put in place - for instance, a ban on alcohol sales on Sundays.

YANDLE: The bootleggers have a market to themselves one day a week. They can buy their booze on Saturday from a legitimate seller, and they can open their drinking houses on Sunday and earn a nifty profit, as long as the Baptists make certain those laws are enforced.

WOODS: The metaphorical bootlegger represents any existing company that might benefit from restricting competitors. And Baptists? Well, that's any group with a sort of morality story that they're peddling. It could be labor rights activists or public health advocates.

MA: Bruce argues that coalitions of bootleggers and Baptists help make regulation more likely.

YANDLE: It is more likely to succeed than an effort by bootleggers alone or efforts by Baptists alone.

WOODS: The idea extends further than just alcohol sales, of course. You can see it when natural gas companies are on the same side as environmentalists when they both try to thwart a common enemy, like coal companies. Or you can see it in gambling licenses, where an existing casino might be on the same side as anti-gambling advocates in blocking new casinos.

MA: And Bruce says this bootleggers and Baptists idea can also be applied to the congressional hearing with Sam Altman from OpenAI.

YANDLE: He was appealing for regulation, most likely from a federal agency, someone who would limit competition, limit entry into the field and serve as a cartel agent for the approved players. And so government, then, becomes a cartelizing factor, a gatekeeper, one who maybe assures that the industry will be profitable and will be doing the "right thing," put in quotation marks.

WOODS: So I guess you're saying that Sam Altman is the bootlegger here.

YANDLE: He's the bootlegger.

MA: But on the other hand, some could argue that Sam's just pointing out the obvious, that there are serious dangers with AI.

WOODS: Yeah, like you can impersonate somebody's voice really easily and you can scam them out of their life savings. You can type in explosions at Pentagon, generate this realistic looking image, share that online, and cause the stock market to tremor. That fake image scenario actually happened on Monday.

MA: Yeah, freaky. So were Sam's comments in front of Congress just a way to sort of lock in his company's advantage? We don't know what Sam was thinking because he didn't actually answer our interview request, but we did get the closest thing - the Baptist who was sitting next to him.

WOODS: Do you think you're a Baptist and maybe the AI companies are the bootleggers in this case?

MARCUS: So, I mean, one never knows. I was sitting next to a CEO, and one never knows.

MA: Gary Marcus, the cognitive scientist.

MARCUS: It was interesting sitting next to him and watching, you know, when he appeared to be tense and when he appeared to be enthusiastic and so forth. I think he really is serious about regulation. Now, I think there are other companies that have talked about regulation where it's probably just performative. But I thought in his particular case, he's got this company that's doing very well, but he does see the risks and he wants to make sure things turn out OK.

MA: Gary supports Sam's call for more regulation, like licenses for AI companies. Both Gary and Sam see licenses working a little like this. A company would work on a new AI tool, say a chat bot. They would test it internally, and then they would show a government regulator that this chat bot wasn't going to spit out instructions on how to overthrow the government and it wasn't going to make sexist and racist content and it was going to meet whatever kinds of requirements that the regulator sets. Then the chat bot would be granted a license to go public. And, of course, because you don't actually know how this thing is going to perform in the real world, there would be some sort of check-in, say, six months later, to make sure that this chat bot isn't being misused.

WOODS: We asked Gary whether he thought that larger companies might be able to meet those standards more easily than a smaller startup.

MARCUS: It seems to me that the way that Sam framed it was very sensitive to that issue. You know, he said that licensing need not apply for a smaller company. And, you know, he said we have to foster innovation in the smaller companies. And so he could have framed that as just everybody needs licensing, and that would have felt, to me, like an attempt to just shut out everybody else and to ensure the incumbents. But the way he framed it was not that.

MA: But Gary is mindful that regulation can go wrong.

MARCUS: There is no easy path here. Like, I'm not naive. But I think in the end, we really do need some regulation. I think the existing laws are just not strong enough to deal with all of the kinds of things that are coming up now that we just never had to deal with. Hopefully we get it right and hopefully if we don't get it right, which we probably won't on the first time, that we can iterate on it and make it better.

WOODS: And on this, Gary and Bruce Yandle, the economist, actually both agree.

YANDLE: I think it's that kind of fine print that would become crucial.


MA: Maybe just don't let ChatGPT write the fine print.

WOODS: No ChatGPT in our legislative halls, please.

MA: (Laughter).


WOODS: The show is produced by Brittany Cronin with engineering by Josephine Nyounai. It was fact-checked by Sierra Juarez. Viet Le is our senior producer, and Kate Concannon edits the show. THE INDICATOR is a production of NPR.


Copyright © 2023 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.