Anthropic Lost $8 Billion Last Year — and It May Not Matter Who Wins AI

Anthropic lost more than $8 billion last year on $4.6 billion of revenue, and its investors still believe the company could float at a valuation above $2 trillion. That gap — eight figures of red ink priced at two trillion of paper — sits at the center of Wall Street's preparation for what could become the three biggest initial public offerings in history: SpaceX, already public, plus Anthropic and OpenAI, both expected to follow.

The numbers behind the hype are staggering. Anthropic has committed more than $500 billion to the computing power and AI infrastructure it will need to keep growing. OpenAI expects revenue to rise almost tenfold, from $36 billion this year to $350 billion by 2030 — yet still projects a cumulative cash burn of nearly $280 billion over the same period.

On the face of it, the economics are ugly. Even so, investors are scrambling to get in, while doubts grow over whether frontier AI labs can ever generate enough value to justify their astronomical spending.

Who actually captures the value of a breakthrough?

The companies that invent breakthrough technologies rarely capture the biggest chunk of the value those technologies generate. That value usually accrues to those who build the complementary assets — the products, platforms, and services built around them.

History is not on the labs' side. In the early twentieth century, electricity transformed nearly every industry, but the biggest productivity gains went not to power generators. They went to manufacturing companies like Ford, which used electricity to revolutionize production. The risk for the labs: ending up like utility companies, providing a shared foundation that others profit from.

The internet ran the same script. Telecoms companies spent heavily building the fiber networks and connectivity that made the web possible. But much of the long-term value flowed to the businesses built on top: search engines, e-commerce, social media platforms, and cloud software.

Frontier AI is now following that pattern, becoming infrastructure rather than software. The leading labs are plowing staggering sums into AI chips, data centers, electricity, and cloud infrastructure — and the models themselves are becoming a bedrock that others capitalize on, much like electricity or the cloud.

Why is technological leadership getting harder to defend?

AI models are rapidly becoming commoditized. Performance is converging, with Chinese labs in particular closing what once looked like an unassailable lead held by their U.S. rivals. That convergence makes technological leadership far harder to defend.

Meanwhile, the companies that already control today's complementary assets are enormous. The hyperscalers — Amazon, Alphabet, Meta, Microsoft, and Oracle — own much of the cloud infrastructure, software, and customer relationships through which businesses actually use AI. Microsoft, for example, can plug different models into its Copilot AI assistant and sell them through software already used by millions of businesses.

Where is the new moat?

One thing looks certain: the low-hanging fruit has been plucked. Frontier labs have already scraped much of what is publicly available on the internet, and that well is running dry. Competitive advantage now shifts to those holding proprietary data — especially data held by private companies and those with unique or niche expertise.

That knowledge becomes the moat. For business leaders, that is reassuringly good news. For years, many fretted that tech giants would swallow every industry whole. That has not always happened, and specialist businesses that know their customers better than anyone else have continued to thrive.

Take Sweetwater. The U.S. music retailer sells many of the same guitars and microphones as Amazon, yet it still pulled in $1.86 billion in sales last year. Its edge comes partly from specialist sales engineers who know the gear and their customers inside out — knowledge AI cannot simply scrape from the internet.

AI is unlikely to change that. If anything, it raises the value of knowing your customers, your products, and your market better than anyone else. The companies that do that may have far more to gain from AI than the companies building the models.

What does history say about the richest AI fortunes?

Some value is already flowing to the companies selling the picks and shovels of the AI boom. Chipmaker Nvidia's meteoric rise to become the world's most valuable company shows just how lucrative that can be.

But history suggests the richest fortunes may still lie ahead, and may be minted by companies we haven't even identified yet. They are most likely to be the ones that figure out how to reinvent themselves around the technology, rather than simply build it. As the source analysis puts it: inventing technology is rarely the same as making money from it.