Research Report Released on State of AI Economy and Compute Growth Trends
Research publication ● Confirmed 72% confidence first seen
Researchers released a comprehensive report measuring the generative AI economy at $110 billion in revenue over the past 12 months with a $175 billion annualized run rate, using deduplicated end-customer spending metrics. The report identified a significant acceleration in global compute growth since 2020, breaking a 50-year trend of steady 66% compounded annual growth, driven by AI infrastructure demands and accelerators. The analysis also noted concerns about enterprise customer spending cuts from major players like OpenAI and Anthropic, raising questions about revenue sustainability.
Decision brief
- What changed
- A new State of the AI Economy report measured generative AI revenue at $110 billion over the past 12 months ($175 billion annualized run rate) using a deduplicated end-customer spending methodology, and separately documented a break in the 50-year trend of 66% compounded annual growth in global compute, with a sharp acceleration since 2020/2023 driven by AI accelerators.
- Why it matters
- The revenue figures suggest AI adoption is outpacing prior tech waves (roughly 3x faster than mobile/internet), but concurrent reports of enterprise customers (Microsoft, Uber, Amazon) cutting spend on OpenAI and Anthropic services raise real questions about whether current run-rate revenue is durable or a temporary 'honeymoon' effect. With OpenAI and Anthropic capturing ~90% of startup-sector AI revenue, any pullback in enterprise spending has outsized market-wide implications, and the compute growth break signals infrastructure investment commitments that assume continued demand growth—a risky bet if revenue growth stalls.
- Evidence
- The revenue and compute figures come from a single 'State of AI Economy' report covered independently by two Exponential View newsletters and referenced by The Algorithmic Bridge; the enterprise spending-cut claims and 'honeymoon' characterization are specific to The Algorithmic Bridge's interpretive commentary and are not independently corroborated in the other two sources.
- What remains uncertain
- It is unclear how robust the deduplication methodology is, whether the cited enterprise spending cuts (Microsoft, Uber, Amazon) are widespread or isolated anecdotes, and whether hyperscaler revenues merely covering depreciation implies negative unit economics or is a normal phase of infrastructure buildout. The claim that compute acceleration will 'sustain for years or potentially decades' is speculative and not substantiated with specific evidence in the coverage provided.
- Monitor next
- Watch upcoming quarterly enterprise spending and usage data from major AI vendors (OpenAI, Anthropic) and hyperscaler capex/depreciation disclosures to see if the reported enterprise pullback deepens or reverses.
Analytical support, not advice — assumptions and open questions stated above.