Major cloud providers like Amazon are making multi-billion dollar investments in AI startups like Anthropic, which then commit to spending that money back on the provider's cloud services. This "circular" financial arrangement locks in future revenue and inflates growth metrics with non-organic activity.

Data reveals an extreme power law where model labs OpenAI and Anthropic capture nearly all AI startup revenue, and their share is growing. This indicates value is accruing to the foundational layer, posing an existential threat to the long-term viability of application-focused startups.

A massive portion of cloud providers' growth comes from just two AI companies, OpenAI and Anthropic. Since these same providers (e.g., Microsoft, Google) are also major investors in those startups, it creates a circular economy where investment capital flows directly back as revenue for compute.

Frontier models can raise more capital than the entire application layer built upon them. This unique financial power allows them to systematically expand and absorb the value of their ecosystem, a dynamic not seen in previous platforms like cloud computing.

The AI market has cleared its first ROI hurdle: model revenue has justified massive infrastructure investment. Now it faces a second, harder test. Enterprises spending billions on AI tokens must demonstrate tangible financial benefits, like higher margins or revenue, to sustain the flywheel.

Despite a booming AI startup ecosystem, revenue is intensely concentrated. Foundational model providers OpenAI and Anthropic capture nearly 90% of the market, and their share is growing, squeezing out application-layer companies.

The business model for AI is pivoting away from SaaS-style subscriptions. Enterprise-focused labs like Anthropic see massive revenue not from adding users, but from the immense token consumption of API power users. A single developer can be 100x more valuable than a subscriber, forcing a shift to consumption-based pricing.

Unlike software bottlenecked by engineering headcount, AI models scale with capital. A frontier model company can raise more than its entire app ecosystem combined, then use that capital to launch competitive first-party apps and subsume third-party developers.

The AI ecosystem has a systemic revenue recognition problem. A single compute token's value can be recognized as ARR multiple times as it's resold down the value chain (e.g., from OpenAI to an application wrapper). This creates inflated, non-durable revenue figures across the industry.