Can you vote for or against technological progress? In most elections, it’s not that simple. But this November, Californians will have as close an opportunity as any group of voters has ever had to do so. That opportunity comes in the form of a ballot initiative proposing a wealth tax on the state’s billionaires. Prop 40, sponsored by SEIU-United Healthcare Workers West, one of California’s largest labor unions, is summarized as follows: “billionaires who were residents of California on January 1, 2026 would have to pay a one-time state tax equal to 5 percent of their net worth. The tax would be due in 2027.”

You may be asking yourself whether anyone — for or against it — could actually believe that, if passed, the wealth tax would be a “one-time” measure rather than the first of many. California has no plans to slow its rate of spending, trending to almost $400 billion annually. The measure’s official summary tells voters that “Ninety percent of the money would have to be spent on health care services for the public. The rest would have to be spent on education, food assistance, and administration of the wealth tax.”

California is infamous for its high taxes, but young people head west anyway, because that’s where the gold is. These two economic facts have coexisted for decades. Because California has supported so much of American technocapitalism in the last half century, it would have been hard to imagine a measure as extreme as a confiscatory wealth tax gaining serious traction. But Prop 40 has a real chance of passing in a few months. Something has changed in how the public views technology and the wealthy. That something is AI. And so, while Prop 40 presents itself as a tax on the ultrarich to prop up healthcare and education spending in California, it really should be thought of as a referendum on AI.

Of California’s 200 or so billionaires, most did not make their wealth by innovating in AI. A few did — like Cursor’s Michael Truell, Anthropic’s Dario Amodei, or Scale AI’s Alexandr Wang. Most made their wealth in the infrastructure of the technology boom that began with personal computers and exploded with the Internet. Sergey Brin, Eric Schmidt, and Jensen Huang were all rich before the AI wave (though they are considerably richer because of it). The state’s official ballot analysis, from the Legislative Analyst’s Office, tells us as much. “California Is Home to Many Billionaires,” the document says. “Many of these billionaires gained their wealth as executives or investors in California technology companies.”

Today, in the midst of an AI revolution that promises to, well, be a revolution, the rhetoric around billionaires has changed into a specific anti-AI reaction, even counter-revolution. And as a result, all the billionaires are, whether they like it or not, “AI billionaires.” An “AI billionaire,” in this essay’s usage, is not someone who became fabulously wealthy by creating a successful AI product. It is someone whose fortune is now bundled up in assets that have benefitted handsomely from AI, and stand to gain even more. Because AI promises to affect the entire economy, no billionaire is exempt from this definition in the public’s eyes: land billionaires, energy billionaires, real estate billionaires… They’re all AI billionaires now.

Billionaires

Like most American children, growing up I knew vaguely that there were billionaires out there.

Warren Buffett was the richest man in the world, and he was friends with Bill Gates. I knew little about either man. Buffett was a good man and from Nebraska; Gates was a computer nerd. Then, in sixth grade, there was a printed-out picture of the Mexican businessman Carlos Slim in the back of my classroom. Slim held the title of the wealthiest man in the world from 2010 through 2013 and a place on the bulletin board of 2013’s “big names” next to tennis player Li Na and whistleblower Edward Snowden. He made his money on Mexico’s phone monopoly; his Wikipedia page calls him an “oligarch.”

When I was in high school in the late 2010s, Mark Zuckerberg was the ultimate “evil billionaire.” After the 2016 Presidential election, in which Donald Trump made great use of social media, Facebook and its CEO became scapegoats for a media and political establishment looking for reasons why Trump had won. With growing public anxiety around the effects of social media on teenagers, Zuckerberg was in the hot seat.

A decade later, the principal cause of Zuckerberg’s public villain status is very different: he and his renamed company, Meta Platforms, are spending billions on building data centers. The 2017 Zuckerberg, dragged before Congress to answer questions about Russian advertising purchases, might be relieved to learn that within a decade no one would be asking him about elections at all, or asking him to censor political material; his relief would stop if he learned how much complaining there’d be about his company’s investments in technology infrastructure (“data centers”) throughout the United States.

And that is essentially the difference between Zuckerberg as technology billionaire and Zuckerberg as AI billionaire.

Elon Musk has been through the same transformation. His life was easier when he was hated because Tesla wasn’t unionized, or when SpaceX made messes on the beach in Texas. Now, he’s merged SpaceX with xAI (itself merged with X), built massive data centers, and deployed Tesla’s self-driving cars around the United States. The complaints are considerably different.

The Wealth Tax

One of California’s quirks is our ballot initiative system, a fixture of Western populism, championed in 1911 by Governor Hiram Johnson. As a result, for over a hundred years, California voters have had the power to impose sweeping changes (such as on legal weed or the three-strikes law) via popular referendum. While the United States was built on representative democracy, California experimented with direct democracy. California’s referendum system is a political technology for taking a direct measurement of public opinion, without the scaffolding of representative democracy watering down the truth.

Come November, the referendum system will put the billionaires on trial.

As a coastal teenager, I was well-versed in anti-billionaire rhetoric. It used to be about income inequality — the fact that the top one percent of the population, at least on paper, earns a fifth of the pre-tax national income. This seems unfair if you believe that no person can be 100,000 times more productive than the median worker, and that “wealth” measures not only what its owner built, but how much of the value around him he was able to capture. A wealth tax was the type of proposal we’d talk about and research in my high school debate class, in the same realm as all countries should eliminate their nuclear arsenals and plea bargaining ought to be abolished in the US criminal justice system.

Take some of the billionaires’ wealth away, the argument goes, and the government can redistribute it to make our lives better. Zohran Mamdani’s socialism is not confiscatory for the sake of it; it promises to use this confiscated wealth to improve the lives of ordinary working people through free daycare and free groceries, and so on. Earlier campaigns to confiscate American fortunes failed to break into the mainstream — even when oil billionaires were blamed for climate change and environmental damage, even when finance billionaires walked away from the wreckage of ‘08 with government bailouts and, aside from some sacrificial lambs, with their bank accounts intact.

In March 2009, the House of Representatives approved — 328 to 93 — a 90% tax on bonuses paid by any company that received a major federal bailout; it died in the Senate. The most successful attempt to put a wealth tax in front of American voters was Elizabeth Warren’s “two-cent tax,” the centerpiece of her 2020 presidential campaign: a two percent annual tax on households with a net worth between $50 million and $1 billion, and three percent over a billion. One poll found that 61% of registered voters approved of the proposal, but Warren lost the Democratic nomination without winning a single state. Nothing made a wealth tax urgent enough to vote on.

The rise of the “centibillionaires” is the conventional explanation. The first to achieve this milestone was Jeff Bezos in 2018, and he passed the $200 billion mark in 2020; Elon Musk leapfrogged him to $300 billion in November 2021. But that explanation is five years too soon. Warren’s two-cent tax polled through those same years but died without a vote. ChatGPT launched in November 2022, by which point the centibillionaire club already had a dozen members. Three years later, SEIU-UHW filed Prop 40.

One champion of this wealth tax comes from a district where all billionaires are most likely to be AI billionaires. Of California’s 52 representatives in the House of Representatives, none has pushed harder for the wealth tax than Ro Khanna, who represents the geographical cradle of Silicon Valley — district CA-17 — which encompasses Sunnyvale, Cupertino, Santa Clara, as well as much of Fremont and San Jose. Nowhere in the country has more to lose from a wealth tax than Silicon Valley.

Just eight companies in CA-17 have a combined market cap north of $10 trillion, which is larger than the GDP of every non-US, non-China country. Mostly-rational economic actors decided that this district creates more value than the vast majority of countries. The reward for this value creation, given to entrepreneurs and even early employees, is the title and financial position of ‘billionaire.’

Elected to Congress in 2016 and endorsed by Silicon Valley royalty Eric Schmidt, Sheryl Sandberg, and John Doerr, Khanna originally promised to be Silicon Valley’s voice in Washington: securing funding for domestic semiconductor manufacturing, pushing for high-skilled immigration reform. His campaign chairman called Khanna “the first true Silicon Valley candidate.” And he was, for a time. In his first term, he sponsored two bipartisan bills, both signed by President Trump. Khanna was “Silicon Valley’s ambassador to Middle America,” a symbol of overcoming the divides between red and blue, rural and urban, tech and the rest.

But now, tech is unpopular with the American public, and Khanna wants to build a national profile. He spoke earlier this year at Stanford with Bernie Sanders in a town hall entitled “Who Controls the Future of AI: THE OLIGARCHS or THE PEOPLE?” (In his op-ed describing the event, Khanna boasted of his popularity among Stanford students: “On February 20th, I was at Stanford University with Sen. Bernie Sanders to speak to over 1,600 students about the defining issues of our time: inequality and AI. We had the largest turnout since President Barack Obama visited the campus in 2015.”)

Ro Khanna has become the de facto spokesman of California’s Prop 40. This is puzzling both because he did not file the proposition (that honor went to two healthcare-industry officials, one from SEIU-UHW, who did so on October 22, 2025) and because his membership in the House of Representatives has no bearing on the ballot proposition’s results. We ought to understand Khanna’s support of the wealth tax as the enforcement of his call for AI to be “reclaimed from billionaires.” And we ought to understand Prop 40 as the first of what will probably be many electoral events that hinge on AI.

The proposed “tax on billionaires,” as articulated in Prop 40, does not just apply to Californians with one billion dollars of cash in their bank account. It also applies to Californians who hold one billion in illiquid startup equity, shares a founder cannot sell without tanking his own company. And the proposal includes a provision on super-voting shares that presumes a founder’s taxable wealth is represented by his voting control, not his actual economic stake in the company (the initiative’s drafters dispute this, but it’s supported by the plain text of the measure). A large percentage of startup founders literally can’t sell their shares for (nearly) as much as they’re worth. They don’t have the money.

It’s hard to be a Congressman vying for national popularity when your district produces the most hated product in the American economy. The new guard of Big Tech companies, frontier labs like OpenAI and Anthropic, is increasingly villainized at national scale. As AI sets off the latest wave of creative destruction — some layoffs, some discomfort — it polls as more unpopular than ICE, per a March NBC News survey. The most negative demographic was the youngest, 18-34, with a net AI favorability rating of minus 44.

But try to imagine Houston’s representative in Congress calling for oil production to be handed over to “the people.”

Silicon Valley’s Wealth

Most of Silicon Valley’s wealth is speculative. Stock prices are a wager on the future. The $10 trillion market cap in CA-17 is a measure of what investors believe they will produce, a belief that depends on conditions staying favorable to capital and industry.

Much of Silicon Valley’s wealth is owned by employees. It’s not unusual for an early employee at a startup to have a “tiny” ownership percentage in a business worth tens or hundreds of millions of dollars. Most people in tech who live in the Bay Area know someone who was a single-digit employee at a unicorn or a decacorn. Silicon Valley may be the only place in the world where employees routinely become millionaires many times over: when founders win, they share the prize money. Startups also distribute risk, and illiquid paper wealth, not just cash profits. All of those wealthy former startup employees were once holding highly unstable, illiquid shares, working to make them worth a lot more.

While the so-called “billionaire tax” may not (currently) apply to Employee #7 at Mercor (though it does to the startup’s young founders), there is still a massive sea of engineers worth tens of millions or even hundreds of millions who work at big tech companies who are next on the chopping block: support for a wealth tax targeting “centimillionaires” is now growing. What was a one-time tax on 200 people in California may soon be expanded into a nationwide, annual tax on America’s thousand-or-so billionaires. “California is America, only sooner.”

As the ballot proposition moved from a fringe proposal to the realm of actual political possibility in late 2025, several notable departures took place. (The tax determines its targets as of January 1, 2026.) Peter Thiel and David Sacks both announced new offices — in Miami and Austin, respectively — on New Year’s Eve. Google founders Sergey Brin and Larry Page relocated most of their activities to Nevada. The Valley’s favorite estimate of the wealth lost is Chamath Palihapitiya’s: a trillion dollars of personal wealth — a thousand billions — has left California since December, and that’s before counting what the fleeing billionaires’ companies produce. By the California Tax Foundation’s accounting, publicly reported departures total $777 billion; Page, Brin, and Zuckerberg’s departures alone represent nearly 38 percent of the state’s billionaire wealth.

The “billionaire exodus” is not finalized — billionaires are mobile, so they can return to the Golden State if the proposition fails. As of writing, Polymarket puts a 34% chance that Prop 40 passes this November, down from a high of 61% in January, when the moving trucks were running around the clock; an August UC Berkeley IGS poll found that 48% of likely voters back Prop 40, with 41% opposed. If a founder moves, their company does not necessarily need to move with them. But for convenience and ideology’s sake, many founders may relocate their companies after they move. Soon after Elon Musk decided he was fed up with California, he took his companies with him. I suspect many more will follow if Prop 40 passes.

For now, though, most of the state’s billionaires — surely with access to the best tax lawyers in the world — are staying, moving some assets out of state, and praying that the retroactive provision of the measure dies in court. Jensen Huang, for example, has announced he’s not leaving, arguing that by living in California, he has agreed to follow all of California’s rules. I wonder if he’d still be on board at 10 percent, or 20.

Silicon Valley’s Future

I live in San Francisco, where I recently heard someone say, “I automated away someone’s job today,” and he wasn’t joking. Just as tractors reduced the need for farmers, and the mechanized loom reduced the need for weavers, AI has begun reducing the need to hire for particular lines of work, although it has not produced mass unemployment. The first to be hit are some of the jobs of the 20th century — consultants, paralegals, copy editors — and some of the jobs of the 21st — Uber drivers, run-of-the-mill software engineers. (No matter what jobs it is creating; more on this later.) Young people feel economic anger; taking the right path and studying computer science or economics is no longer a guarantee of economic security.

If regular people can’t have economic security, why should the AI billionaires?

Socialism is most popular among 20-somethings, people roughly my age. We were not red diaper babies; our parents were capitalists who raised us on friendly billionaires. A leaked email written by Peter Thiel in 2020 asserts that 70% of American Millennials have socialist leanings; he urges his audience — including Mark Zuckerberg — to take this political reality seriously. Thiel blamed student debt and expensive housing, but he didn’t foresee that technology itself would be the impetus for socialism. Socialism has not become more popular in the last 15 years. Gallup has asked Americans about socialism and capitalism since 2010, and young adults’ approval of socialism has barely moved, hovering near half throughout the years. But what moved was what young adults think about capitalism: falling from 68 percent approval among 18-to-29 year olds in 2010 to 45 percent by 2018. Last year, capitalism’s rating among all Americans hit the lowest point Gallup had recorded, 54 percent.

My generation did not convert to socialism, but defected from capitalism. Progressive politicians like Alexandria Ocasio-Cortez, Bernie Sanders, Zohran Mamdani, and Ro Khanna are winning influence among coastal 20-somethings — at least the ones who aren’t becoming the youngest self-made billionaires in history.

This isn’t a new radicalism. The generation’s leftmost cohort is roughly the same size it’s been since 2010. When wokeness was the state religion, the evil of capitalism was table stakes. So calling yourself a socialist is less of a leftist lurch and more of a rebrand. When identity politics — “woke” — collapsed as an organizing frame, the standing resentment and anger needed a new object. Enter the AI billionaire.

AI labs and startups have rushed to hire “storytellers,” brand consultants, and other creatives to explain why their products are good, not dystopian: why artificial intelligence is not just another product but a positive good that will positively impact your life. An industry confident in its popularity would not need the hired help. (I work for a publication meant to spread the gospel of technology and capitalism — I’m no exception — though Arena is not developing AGI anytime soon.)

These same companies have spent the last five years prophesying an artificial general intelligence “machine god.” That prophecy may have been more aimed at investors and talent the labs were competing for, but the public overheard the pitch. The average American balks at a machine god that takes away their job and a stable middle-class lifestyle. There is nothing wrong with advertising. But this is now a new phenomenon: an industry advertising to people whose jobs it is threatening to automate. In July of this year, Anthropic ran an ad starting with a frame of a burning house that asks “who’s going to hit the brakes if we need to?” over a photo of Arlington National Cemetery. The ad, titled “There’s Hope in Hard Questions,” pivots to a nicer tone, suggesting that AI can cure diseases and help people with their jobs, rather than replacing them. I suspect most people will remember the first half of the ad more than the second.

For young people whose entry-level jobs are the first to be automated (consultants, early-level software engineers), and who hear tales of how hard it is to get a job right now, tech is an easy villain. Though the great man theory of history has fallen out of fashion, it’s easier to blame Sam Altman or Dario Amodei than the tech industry, abstractly.

Neither man is helping his case. The labs have coalesced around the idea of some sort of wealth redistribution of their own design. Sam Altman spent $14 million of his own money studying universal basic income (another debate class proposal) before announcing, in an April 2026 interview with The Atlantic, “I no longer believe in universal basic income as much as I once did,” preferring instead “collective ownership that could be in compute or in equities or something else.” His OpenAI has proposed a national Public Wealth Fund, seeded with contributions from AI companies, to give every citizen “a stake in AI-driven economic growth.” The shift from an AI-funded UBI to public shared equity is the exact parallel of the shift from “income inequality” to “wealth tax.” Income may be fake; assets are what matter.

Dario Amodei, in a June 2026 essay, wrote that if AI-driven labor displacement becomes large and permanent, “universal basic income could be financed through taxes on relevant companies or raising the capital gains tax.” The premise of Prop 40 — that AI wealth should be distributed — was conceded first by the AI billionaires themselves, who would, naturally, prefer to design the machinery of wealth redistribution rather than having it designed for them.

The proposition’s smartest defenders can point at Altman and Amodei. If the two most knowledgeable people about the AI labs say that AI will change the economy as we know it, we should take them at their word — positioning redistribution before the change arrives so that the AI revolution is closer to the Industrial Revolution than the French Revolution. Their argument is that the wealth tax is the price of keeping things peaceful when jobs eventually (in five years? a dozen?) are automated away. The real contest is about who gets to redistribute the AI wealth: AI companies themselves, or the voters.

The punch line is this: AI is not actually producing mass unemployment. The unemployment rate sits near four percent, the historic floor. What AI has taken, so far, are entry-level jobs. Junior postings are now down a quarter from their peak; the “stable path” of toughing out a computer science degree now has a seven percent unemployment rate, even as the field keeps growing. It’s a narrow slice of the labor market, but elite aspirants were promised a stable, well-paying job at the end of their computer science degree; the broken promise has made them angry and vocal. Prop 40 doesn’t need mass unemployment to pass, but the perception that mass unemployment is years, months, days away. And the expectation that AI will replace, not complement, human work is one that the AI labs have sold for years.

Prop 40 is a bet on a future where AI produces everything and only a few AI oligarchs capture the wealth. This November, Californians are being asked to vote on that bet. Silicon Valley wanted humankind, its customer base, to believe in a revolutionary future. Prop 40’s success or failure will tell us if Californians believe it.