“Most of the benefits of technological change are passed on to consumers.”
— William Nordhaus

Friday charts: Welcome, new billionaires!

In addition to thousands of millionaires and hundreds of centimillionaires, the stock market minted seven new billionaires on Thursday. Each of Anthropic’s co-founders is now worth roughly $22 billion.

(One non-founder, CFO Krishna Rao, was a billionaire at yesterday’s opening price, but, sadly, only a centimillionaire by the close.)

Collectively, Anthropic’s 5,300 employees are now worth something like $400 billion.

That’s pre-tax, of course. And mostly on paper, although some of the wealth has been realized. The founders each sold $280 million of stock at the IPO price. 

(That made it a good day for the government, too. The IRS will take 23.8% of the founders’ realized gains.)

It’s perhaps worth stating the obvious here: All this newfound wealth came from investors. No one is poorer because the Anthropic Seven are richer.

In fact, we’re all a tiny bit richer: We get to use Claude for free.

Still, billionaires are existentially unpopular. “I don’t think we should have billionaires,” Zohran Mamdani has said. 

We do, though. Lots of them. The US was home to 1,265 billionaires in 2025, the most in the world by far. (Communist China is second with 363.)

Should we welcome seven more?

William Nordhaus would say we should.

In a 2004 paper, the Nobel-winning economist modelled what fraction of the benefits from technological advances is kept by the people and companies responsible for them. He called the number he arrived at “minuscule.”

“I estimate that innovators are able to capture about 2.2 percent of the total social surplus from innovation.”

2.2%!

The rest — 97.8%! — is captured by consumers.

Nordhaus attributes this happy arrangement to how quickly “Schumpeterian profits” — the portion of a business’s profits attributable to innovation — get competed away.

“US capitalism grinds Schumpeterian profits into such a fine powder that they can barely be detected in the macroeconomic [data],” he concludes.

Anthropic may be the best possible example of this.

Nordhaus estimates that innovators initially capture 7% of the value of their innovations, but that their share of the profits depreciates by 20% a year as competitors respond (which works out to a net present value of 2.2%).

In the case of Anthropic, the competition responds even faster. As soon as they release one billion-dollar model, it’s a flat-out race to release the next one.

They don’t always win it.

On the same day Anthropic made history with the largest-ever IPO, it lost its place at the top of the LLM food chain. Google’s new Neon and Carbon models are widely said to be better than Anthropic's Fable — in Anthropic’s strong suit of coding included.

Before this week, Google hadn’t even been considered a serious competitor.

The stock market created seven new billionaires yesterday because investors believe Anthropic will eventually capture a substantial share of the value it's creating. History suggests they might be disappointed.

Nordhaus, for example, attributes the dotcom bust to mistaken beliefs about Schumpeterian profits. He estimates that, at their 1999 peak, stock prices implied that tech companies would capture 50% of the social benefits of the technology they were building.

The stock market crashed in 2000, Nordhaus says, because the reality was closer to 2.2%.

If the AI boom follows the same pattern, investors might come to regret creating seven new billionaires this week.

Consumers, on the other hand, should welcome them.

Let’s check the charts.

More billionaires v. growth:

Political economist Blair Fix finds a positive correlation between the number of billionaires in a country and GDP growth: The more billionaires a country has, the faster it grows. Mamdani’s country of no billionaires might therefore have no growth, either.

Top 400:

Not all billionaires are equally useful, of course. But the US has increasingly more of the productive kind. 71% of the Forbes list of the 400 richest Americans for 2026 is self-made, up from 40% in 1982. 

Billionaires by city:

New York City has the most, according to the 2026 Billionaire Census published by Altrata. Adding this week's seven new billionaires would put San Francisco just ahead of Hong Kong.

They may not be there for long.

A study finds that turnover among the elite is high: About 70% (the decimal on the right above) of wealthiest Americans drop out of the top 1% within 10 years. Most of this is new money, too. Only 7% of people with a grandfather in the top 1% of wealth are themselves in the 1%.

What hath Anthropic wrought?

It’s also possible that the innovations AI labs like Anthropic have created will have negative social benefits, like the current explosion of cyberattacks, for example. Or the risk that “misaligned AI systems could corrupt institutional decision-making or cause large-scale harm,” as Anthropic warns in its S-1.

Schumpeterian competition:

The rapidly deflating cost of AI tokens suggests that Anthropic has even less time than usual to capture the value created by its frontier models.

The bull case:

A new index from Stripe suggests that the median AI company is growing 40% per annum and that AI companies in the aggregate are growing over 100%. (AI labs like Anthropic are excluded, I think.)

Dotcom redux?

Fidelity notes that, as measured by both price (the lines at the top) and valuation (the lines at the bottom), the AI boom (in black) has closely followed the dotcom boom (in blue and orange).

Investors may regret making the Anthropic Seven so wealthy this week.

With luck, though, it will be because consumers captured all the value.

Have a great weekend, Schumpeterian readers.

Raydium is Solana’s longest-running AMM and now its leading venue for tokenized stocks.

Raydium generated $8.7M in net revenue in Q3 2026, more than triple Q2, and has bought back $206M of RAY since launch. LaunchLab revenue helped drive the rebound.

Raydium handled 51% of Solana’s tokenized asset spot volume. Those pairs generated 40% of Raydium’s swap fees while accounting for less than a fifth of its trading volume.

Virtually all blockchains make money from blockspace, while most stablecoin issuers make money from reserves. Very few businesses have exposure to both. With USDG, Arbitrum becomes one of them, retaining its upside to fee-based revenue should onchain activity recover, while adding a stablecoin revenue line that does not depend on it.

Read the full report from Blockworks to see how we built the model and what the bull case could mean for Arbitrum.