The numbers nobody is talking about loud enough: Anthropic went from $1B in annual recurring revenue to $30 billion in 15 months. OpenAI, the household name, is on track to lose $14 billion in 2026 and just pushed its breakeven target to 2030.
That gap between the two stories is where the next wave of AI wealth is being created — and most retail investors are looking in the wrong place.
Why this is the most important AI chart of the year
In the entire history of enterprise software, no company has scaled revenue this fast. SaaS unicorns used to take a decade to get from $1B to $10B ARR. Anthropic did 30x in less time than it takes to ship a baby twice.
What changed? Two things:
- The cost of running serious AI fell off a cliff. Inference costs are down roughly 76% year-over-year for frontier-class models. Customers who previously couldn’t afford to embed AI into their workflows now can.
- Enterprise budgets are reshuffling, not expanding. When Klarna replaces a chunk of its customer support stack with AI, that money used to go to BPO firms. Now it goes to model providers. That’s not “new spend.” That’s the largest budget transfer in software history.
Anthropic’s customer base reportedly went from ~500 enterprises spending over $1M annually in February to over 1,000 by April. That’s doubling in less than two months.
The wealth angle most people miss
Here’s the contrarian take: the biggest winners of this AI buildout aren’t the model companies themselves. They’re the picks-and-shovels.
NVIDIA’s Q4 FY2026 Data Center revenue hit $62.31B — up 75% year-over-year. Hyperscalers are spending hundreds of billions on AI infrastructure to support exactly the kind of usage Anthropic and OpenAI are driving. And they have to keep spending whether the model layer eventually consolidates to one winner or three.
That’s a different bet than picking the right AI chatbot. It’s a bet on the underlying compute economy continuing to grow regardless of which logo wins the consumer war.
Where the real opportunity is for individuals
Most people read these AI revenue stories and think one thing: “I should buy NVDA.” That’s the obvious play, and it might still work — but it’s already priced in by people who saw this coming two years ago.
The non-obvious wealth plays right now:
1. Use AI to compress your own income timeline. A solo developer with Claude or GPT-5.5 can now ship in two weeks what took a five-person team six months in 2022. If you’re not using these tools to build something — a product, a service, a content business, an audience — you’re voluntarily playing on hard mode while your competitors play on easy mode.
2. Position around the second-order winners. Power generation, grid infrastructure, cooling, advanced packaging, specialty chemicals for chip manufacturing. Every dollar spent at the model layer flows down to physical infrastructure. That part of the market is still mispriced because it isn’t sexy.
3. Be skeptical of pure consumer AI plays at current valuations. ChatGPT-style apps are real businesses, but the moats are thinner than the multiples imply. The market is still figuring out who has actual lock-in vs. who is renting attention.
The Fed week wrinkle
The Fed kicks off its April meeting today and announces tomorrow. No one expects a rate change, but the language matters more than the decision. If the Fed signals they’re comfortable with the current path, AI infrastructure stocks (already at record highs) probably extend higher because cheap capital is the rocket fuel for the buildout.
If they hint at staying restrictive longer, expect a sharp short-term pullback in the same names. That pullback is your buying opportunity, not a reason to panic.
The structural story — AI revenue growing faster than any prior software wave — doesn’t change because of one Fed meeting. The cycle pattern is what reshuffles entry points.
What you should actually do this week
Skip the rate-decision day-trading. It’s a coin flip dressed up as analysis. Instead:
- Audit which AI tools you’re personally paying for and what they’re actually saving you. If you’re spending $100/month on AI subscriptions and getting back less than 2 hours of weekly time savings, you’re not using them right.
- Look at your portfolio’s exposure to the “second-order” infrastructure plays vs. the obvious AI logos. Most retail portfolios are dramatically underweight the picks-and-shovels.
- Set up a small recurring buy into the broad tech infrastructure complex. The single best decision an investor can make right now is to not try to time the perfect entry — and just be exposed.
The bigger picture
Anthropic going from $1B to $30B in 15 months isn’t just an investment story. It’s evidence that the productivity gap between people who use AI and people who don’t is widening fast — and that gap will compound into a wealth gap over the next decade.
This is the part that should keep you up at night if you’re sitting on the sidelines: the people who learned how to compound capital in the last cycle won. The people who learn how to compound output with AI will win this one. The two groups are not necessarily the same.
Want to track your own path to financial independence — including how your portfolio benefits from this AI buildout? The Wealtharian Wealth Tracker lets you monitor your net worth, FU money progress, and investment milestones in one place. Try it free →