Dario Amodei stood on stage at Anthropic's Code with Claude developer conference in San Francisco on Wednesday and dropped a number that makes the rest of the AI industry's growth look pedestrian. Revenue and usage grew 80x year-over-year in Q1 2026. Not 80 percent. Eighty times.

Anthropic had planned for growth "ranging from a little to 10x," Amodei said during a fireside chat with his sister and co-founder Daniela Amodei. The gap between 10x and 80x is why Claude has been running into capacity walls all year. It is also why Amodei, half-joking, told the audience that he hopes the growth rate slows down because 80x is "too hard to handle."

He is not joking. Earlier reporting placed Claude Code's annualized revenue run rate at roughly $2.5 billion. If the 80x trajectory holds, Anthropic is on pace to become one of the highest-revenue software companies in the world by year-end, barely three years after its first public product launch.

The Bubble Question Just Got Complicated

The Atlantic published a long analysis this week titled "Maybe AI Isn't a Bubble After All," built around the observation that Anthropic's revenue is growing faster than Zoom during the pandemic, Google during the early 2000s, and Standard Oil during the Gilded Age. The writer, Roge Karma, notes that if the current growth rate continued, Anthropic would take in more money than any other company in the world by early next year.

Six months ago, The Atlantic's same writer wrote that AI looked like a bubble. What changed is Claude Code. When Anthropic released its November update, AI crossed what Karma calls "some invisible threshold between interesting gadget and life-changing technology." Teams of autonomous AI agents could take over a developer's computer and complete in minutes what previously took days or weeks.

The evidence is piling up. Goldman Sachs interviewed 40 software companies in mid-April and found many were "overrunning their initial budgets" for AI tools "by orders of magnitude." Researchers who ran an experiment last year showing developers were 20% slower with AI re-ran it with the latest tools. This time, developers were almost 20% faster. Over half of American businesses now have a paid AI subscription, up from a quarter in early 2025.

The Infrastructure Problem

Anthropic's Code with Claude conference was not just about revenue numbers. The company simultaneously announced a deal with SpaceX to lease the entire Colossus 1 data center in Memphis, adding over 300 megawatts and 220,000 Nvidia GPUs of capacity. Claude Code rate limits were doubled for Pro, Max, and Team plans effective immediately. Peak hour throttling was eliminated.

The fact that Anthropic needed to lease an entire supercomputer from Elon Musk's rocket company to keep up with demand tells you everything about where the AI industry is right now. Demand is not the problem. Supply is the problem. And supply is a problem because the demand turned out to be 8 times what anyone projected.

The Broader Picture

Amodei also addressed how AI adoption is spreading. Software engineers are the first adopters because they can immediately measure the productivity gain. But he framed coding as a foreshadowing of how things will work across the entire economy. Meta just announced it will lay off 10% of its workforce. Mark Zuckerberg told investors that "projects that used to require big teams can now be accomplished by a single very talented person."

Nvidia's old 2022 chips now cost more today than when they launched because demand for AI compute has overwhelmed every generation of supply. Semiconductors are in such short supply that OpenAI has scrapped its video-generation app entirely to free up computing power for its core products.

The question is no longer whether AI works. It works. The question is whether the infrastructure can scale fast enough to meet the demand that actual, measurable, budget-busting productivity gains are creating. Anthropic's 80x number suggests the answer, for now, is no.

Amodei's comments reported by Business Insider and multiple outlets. The Atlantic analysis by Roge Karma. Goldman Sachs enterprise AI survey cited via The Atlantic. SpaceX/Colossus deal reported by Ars Technica.