This is an edition of The Atlantic Daily, a newsletter that guides you through the biggest stories of the day, helps you discover new ideas, and recommends the best in culture. Sign up for it here.
Many people are understandably focused on the question of how AI will transform our civilization. Will it cure cancer? Will it end world hunger? Will it hack everyone’s bank accounts and incubate the next pandemic? But those who are already profiting from this technological revolution also have a more immediate concern. Their problem is—to paraphrase one of my favorite lines from Scarface—what to do with all the cash.
Over the past few days, two of America’s most prominent AI leaders have offered proposals for returning some of the spoils of the AI boom to the public. On Sunday, Anthropic announced that it would set aside 5 percent of the company’s stock in a new LLC—the Anthropic Public Benefit Fund—to be spent on causes it deems to be in the public’s interest. More stock may be added as the company becomes more valuable. If the company goes public at its current projected valuation of $1.5 trillion, the fund could hold $75 billion, making it one of the largest charities on the planet. (For context, the Gates Foundation Trust had about $89 billion in it at the end of last year.) Elon Musk, a centibillionaire and a longtime skeptic of traditional philanthropy, criticized the move and tried to one-up it, writing on X that he’d be giving one share of SpaceX to “every kid in America.” According to The New York Times’ back-of-the-napkin math, this could end up being a donation of more than $10 billion.
What AI’s biggest winners decide to do with their money could have profound consequences for how Americans live their lives. The birth of modern AI may turn out to be the biggest wealth-creation event in human history, yet most Americans are feeling those effects indirectly. AI is making some people’s jobs easier and driving innovation, but it isn’t hitting everyone’s personal finances equally, and almost half of U.S. adults think the technology will widen the gap between the rich and the poor. Musk and Anthropic are now laying out two different visions for AI-driven philanthropy—raising thorny questions about the level of control that the investor class will exert over our shared future.
The first and most obvious question for Anthropic: What if the company’s understanding of the “public good” doesn’t line up with the public’s wants or needs? Anthropic’s leaders, Dario and Daniela Amodei, have already committed to giving away most of their own personal wealth, and Dario has written on his blog about the kinds of philanthropy he admires. The Gates Foundation and PEPFAR, President George W. Bush’s public initiative for global AIDS relief, are both on his list. But how the firm’s leaders will actually decide to apportion these gifts remains somewhat nebulous. A press release for Anthropic’s fund lists “large donations to traditional nonprofits” and “direct distributions to specific populations” as examples of possible projects. Another stated goal is to remediate the effects of a massive, AI-driven labor-market transformation—an ironic instance of a company trying to solve a problem it helped create.
As many politicians and even AI corporations themselves have argued, this tech is powerful enough, and potentially such an important part of the global economy, that the wealth it creates may merit some degree of public control. Last October, as one possible plan for reckoning with the economic consequences of AI, Anthropic floated the idea of using “sovereign wealth funds to give citizens stakes in AI revenues.” (OpenAI’s leaders have also reportedly considered giving 5 percent of the company to the government.) At least for now, though, the Anthropic Public Benefit Fund isn’t giving anyone outside of the company a stake. And the fact that it won’t be a 501(c)(3), like the Gates Foundation, means that it won’t have to deal with the kinds of mandatory financial disclosures that are meant to keep charitable organizations maximally transparent. A democratic way to approach this project would be for the government to aggressively tax the company and decide for itself how to spend the money. Not every American agrees with their government representatives’ approach to resource allocation, but at least in the public sector, the people have a voice.
Although Musk’s path is also billionaire-led, it gives Americans a much more direct stake. Donald Trump clarified yesterday that Musk’s donation, if it happens, would use the country’s new child-savings accounts (known as “Trump Accounts”) as a vehicle. Rather than earmarking shares for a self-sustaining charitable foundation, Musk would be effectively cutting out the middleman, giving American children a piece of the company that houses his biggest AI bets. It would likely be a smaller total donation than Anthropic’s; one share per child may not turn out to be all that much money in the long run. And Trump Accounts aren’t a perfect dispersal mechanism; parents can’t reject donated shares, and they can’t immediately sell them. Once account holders turn 18, though, they can spend Musk’s gift on whatever they want.
Musk, like other tech billionaires, has questioned traditional modes of philanthropy. He knows that the public knows that giving can be an expression of power and control—a way to rehabilitate a donor’s image or provide political insulation for the donor and his enterprises. Musk may be getting something out of this too. If more people own a little chunk of the AI economy, will they start to become more receptive to data centers and other Big Tech interventions? The donation could also give recipients a reason to be less critical of Musk and his businesses; people with a stake in the company might be more inclined to tolerate unscrupulous behavior. Still, if he can follow through on his approach (and especially if he continues to donate AI stock, as he hinted in his post yesterday), his gift could have a real impact
When SpaceX was about to go public earlier this year, my colleague Matteo Wong observed that “by weaving itself into the finances of most businesses and households in America, the AI industry is quickly making itself too big to fail.” AI is already deeply embedded in our life: in the businesses we interact with, in the pension funds and 401(k)s around which we plan our future. These massive donations, which represent a deviation from tech-industry norms, could reshape society for the better. But if the public comes to rely on Anthropic’s investments or on direct payouts from SpaceX, it’ll need to start caring even more about how well these companies perform. Public welfare needn’t be contingent on the whims of a few powerful firms; so much else already is.
Related:
Here are four new stories from The Atlantic:
- Exclusive: Trump Jr. privately breaks with his dad.
- Why is strategy so difficult?
- Helen Lewis on all the president’s hotties
- France will never be enough for the people desecrating it.
Today’s News
- Former New York City Mayor Michael Bloomberg is donating roughly $80 million to support Democrats in the midterm elections, and is focused on helping the party win control of the House. Meanwhile, the Republican National Committee is transferring $50 million to groups supporting Republican House and Senate candidates.
- The United Kingdom announced that it will deploy Royal Air Force fighter jets to Saudi Arabia to defend against missile and drone attacks by the Iran-backed Houthis in Yemen. The jets will protect Saudi cities and infrastructure but will not conduct offensive strikes inside Yemen, British officials said.
- Elon Musk ordered restrictions on Ukraine’s military-grade Starlink service last week, disrupting drone operations and air defenses, according to The Wall Street Journal. Musk reportedly discussed the decision, and President Trump said he did not know who was responsible for the move.
Dispatches
- The Weekly Planet: The Trump administration has mandated that federal firefighters pursue a policy of “full suppression” on every wildfire, Joshua Partlow reports. He explores the sometimes fatal dangers of this aggressive approach.
Explore all of our newsletters here.
More From The Atlantic
- Radio Atlantic: Trump’s “staggering one-two punch” to Ukraine
- Charlie Warzel: You can believe everything.
- Donald Trump really wants this golf course.
- The cynicism of Trump’s climate policy
- Elizabeth Bruenig on the woman who has prevented 200 executions
Evening Read
Theranos-Style Fraud Has Become the Norm
By Idrees Kahloon
Perhaps Elizabeth Holmes was simply ahead of her time: Today, the Theranos founder’s crimes look less notorious amid a pervasive culture of grift and graft.
Her biotechnology company infamously promised to revolutionize medicine by conducting hundreds of tests on mere pinpricks of blood. A decade after it fell into disgrace, and three years after she began serving her 11-year federal-prison sentence for defrauding investors of hundreds of millions of dollars, Holmes is the subject of a new documentary. You Can See Everything, directed by Lance Oppenheim and Nathan Fielder, came about because Holmes and her partner, the hotel heir Billy Evans, invited Fielder, a surrealist sketch comedian, to live with them just before she went to prison. They allowed him to film whatever he saw. The result is unsettling, because Holmes’s placid indifference toward the truth has become more familiar than it was when Theranos unraveled.
Culture Break
Reflect. Lego’s overt outreach to an adult market has changed the nature of its products, Sophia Nguyen reports. She explores what the company lost when it grew up.
Read. Twenty years since Michael Pollan’s book The Omnivore’s Dilemma was published, America’s food system is still a mess—but there are reasons for optimism, he writes.
Rafaela Jinich contributed to this newsletter.
When you buy a book using a link in this newsletter, we receive a commission. Thank you for supporting The Atlantic.