‘Oppenheimer’ and the politics of tech

With help from Derek Robertson

If you think the relationship between tech leaders and Washington is fraught now, the movie “Oppenheimer” — the weekend’s #2 top-grossing release, after the Barbie movie — offers a reminder that it was once much, much tenser.

The Robert Oppenheimer biopic, if you’ve somehow missed both the PR onslaught and the three-hour film itself, tells the story of the atom bomb’s invention. But the film, based on the monumental 2005 biography “American Prometheus,” actually revolves around a political mess in the Manhattan Project’s aftermath.

Oppenheimer was the physicist in charge of the project, but in the years that followed, he became outspoken about the dangers of a nuclear arms race between the U.S. and the Soviet Union, warning that it threatened humanity’s survival.

Then in 1954, with Washington in the grip of Sen. Joseph McCarthy’s anti-communist crusade, the Energy Department stripped Oppenheimer of his security clearance on the basis of his personal associations with communists. (In December, the Biden administration made the symbolic gesture of reversing the decision, saying it resulted from a “flawed process.”)

That all happened seven decades ago, but it’s lost on precisely nobody that Washington is, once again, grappling with the proliferation of another technology that carries potentially existential stakes.

The film’s director, Christopher Nolan, has noted “very strong parallels” between Oppenheimer and the AI pioneers who are now warning about the technology’s dangers.

On Thursday, Democratic Sen. Ed Markey of Massachusetts met with “American Prometheus” co-author Kai Bird in Washington to discuss Markey’s proposal to amend the National Defense Authorization Act, which is currently before the Senate, to ban the use of AI in making nuclear launch decisions.

DFD spoke to Bird Friday about those parallels, the role of scientists in public life, and yes, this weekend’s other blockbuster release.

What lesson does Oppenheimer’s experience offer right now, as Washington scrambles to figure out what to do about AI?

We need to be having informed discussion. And our politicians are going to need to start thinking and passing some legislation to regulate this technology.

What happened to Oppenheimer should be a reminder that the sooner we do that, the better.

Oppenheimer created the atomic bomb, and within months he was speaking out publicly about the dangers it posed and the need to regulate it and to create international guidelines and norms for ensuring that we don’t engage in an arms race. Well, he was ignored. Harry Truman ignored his advice and the national security establishment helped to bring him down precisely because he was threatening their budgets. And instead of regulating nuclear weapons and essentially banning them, which is what should have happened under international controls, we engaged in a very expensive arms race.

You just met with Ed Markey about his NDAA amendment. What do you make of the amendment, and its chances?

He’s introducing legislation to prohibit the use of AI in launching decisions for nuclear weapons, which would seem to be a wholly commonsensical thing to do.

But he explained to me that he’s encountering questions and reluctance from his Republican colleagues who say, “We don’t think we need to address this now,” and “Aren’t some of our adversaries already trying to use AI to determine when and how to use nuclear weapons? So if they do, don’t we have to do it?” And of course I disagree with that.

The film depicts Harry Truman’s disgust with Oppenheimer’s misgivings about the bomb during an Oval Office meeting. The president comes off as callous in that scene, but isn’t there a case that people in academic communities often have visions for society that tend toward the utopian or impractical, and that scientists should be subordinated to elected leaders when it comes to shaping public policy decisions?

No, because I don’t think our politicians, many of them are scientists. So we need to understand that scientists should be a part of a political debate.

And they themselves will disagree based on their own philosophical and religious and ethical values. But nevertheless, they are important parts of the debate because they have a knowledge about the technical and scientific issues.

Much of the movie revolves around the question of whether Oppenheimer’s advice to government officials — on the export of radioactive isotopes and the development of a hydrogen bomb — was purely technical, or was somehow tainted by communist sympathies. How do you draw a line between a scientist’s technical conclusions and their views about the world?

I think it’s impossible to separate your values. Oppenheimer was a great scientist precisely because he was a humanist as well. And he loved the poetry of T.S. Eliot and the novels of Ernest Hemingway.

The notion that scientists should only be giving quote “objective” scientific advice is precisely the problem. Oppenheimer was always trying to use his political values to inform us on the choices we faced with how to live with the bomb. And that was quite proper.

Do you plan to see the Barbie movie?

[Laughter] I’ll probably see it.

And so it begins: Twitter began its rebrand as “X” today, taking the first step toward turning Elon Musk’s company into his long-promised all-in-one “super-app” with streaming, text, and banking features.

But let’s take a closer look at his very first step, which was to replace the iconic blue bird logo with an Art Deco-style “X.” Andres Guadamuz, a law professor and blogger who focuses on copyright in the digital world, did so in (what else) a Twitter thread this morning, breaking down exactly where the logo came from and what kind of future it might (or might not) have as the company’s trademark.

At first, Guadamuz thought the X came from a commercial font called Special Alphabets 4 by font designer Monotype. But in an update later Monday, he pointed out the same font is also a character in Unicode, the universal standard for text and emoji that allows computer displays across the world to display the same characters. That raises major questions about whether the logo can be trademarked, something he says intellectual property experts are skeptical about. — Derek Robertson

Could AI-generated websites give resource-strapped candidates a leg up?

In today’s Morning Score newsletter, POLITICO’s Madison Fernandez reports on a new service called “Hey Victor,” which uses artificial intelligence to whip up professional-looking campaign websites from just a few basic candidate assets like their photo, slogan, and biography at a price of $300 and then $25 per month. Founded by two alums of Andrew Yang’s run for the 2020 Democratic presidential nomination, the service is allowing Democratic candidates to sign up for its beta starting today.

Madison writes that “Around 50 sites have been created using the program so far, including for Democrats running for the Georgia state House, New York state Assembly and New York City Council.” Despite how snappy they look, Patricia Nelson, Yang’s creative and social media director, said campaign staffers and designers need not fear Hey Victor putting them out of work: For cash-strapped down-ballot campaigns, “This role doesn’t exist, and they can’t afford to pay somebody to do it on the local level,” she said. “They don’t want to worry about building a website. They want to worry about raising money.” — Derek Robertson