NGE · Investment Letter · Issue 26 · June 2026

Nobody Monetizes
Like the United States.

The instinct behind "one quadrillion by 2040" is correct even though the number itself doesn't hold up under verification. The real figure is more striking and far better documented: US private AI investment now runs 23 times China's, and the rest of the world barely registers on the same chart. This letter corrects the number, finds the real one, and asks the harder question — why can only one country monetize a technology the entire planet is racing to build.

Not investment advice. Not a recommendation to buy or sell. Research and long-horizon thinking only. Consult a qualified financial advisor before making any investment decision. Figures cited are sourced from Stanford HAI's AI Index, McKinsey Global Institute, PwC, and Gartner, current as of writing.

A Correction First

There is no credible
"$1 quadrillion by 2040" forecast.

Before building on a thesis, it deserves to be checked — and this one does not survive the check. No major institution — not McKinsey, not PwC, not Goldman Sachs, not the IMF — has published a projection of the global economy, or the AI economy specifically, reaching $1 quadrillion by 2040. The actual $1 quadrillion figure that exists in serious research is entirely different: an IMF working paper found global cross-border payments, traditional and crypto combined, approached roughly $1 quadrillion in total transaction value in 2024 — a payments-flow statistic, not a GDP or market-size forecast, and not specific to AI at all.

The real, credible numbers for AI's economic potential by 2040 are large but an order of magnitude smaller than a quadrillion: McKinsey's most comprehensive estimate puts AI software and services at $15.5 to $22.9 trillion in annual economic value by 2040, while its narrower generative AI estimate is $2.6 to $4.4 trillion annually, also by 2040. PwC separately estimates AI could add $15.7 trillion to global GDP by 2030. These are genuinely enormous numbers — McKinsey's high estimate alone is roughly equivalent to the entire current US economy — but they describe trillions, not quadrillions. Getting this number right matters, because the more interesting and better-documented story is what these dollars reveal about who actually captures them.

"The US held 97% of global generative AI deal value in H1 2025, with EMEA at just 2%." — AI Investment By Country, 2026 Statistics

23x
US-to-China private AI investment ratio, 2025
$23T
McKinsey's high-end AI software/services value by 2040
97%
US share of global generative AI deal value, H1 2025
The Real Number

The monetization gap
is wider than your instinct suggested.

Your underlying instinct — that the US captures a wildly disproportionate share of AI's economic value while the rest of the world barely competes — is not just correct, it understates reality. The actual concentration is far more extreme than a 65/25/10 split implies.

United States · Private AI Investment
  • $285.9B in 2025 · up 162% from $109.1B in 2024
  • 40 notable AI models produced in 2024 alone
  • 97% of global generative AI deal value, H1 2025
  • $470.9B cumulative private investment, 2013–2024
China · Private AI Investment
  • $12.4B in 2025 · 23x smaller than the US
  • 15 notable AI models produced in 2024
  • Closed the model-performance gap to 2.7 points on benchmarks
  • $119B cumulative private investment, 2013–2024

The single most important data point in this comparison is not the spending gap itself, but what happened to it in just one year: the US-to-China investment multiple widened from 11.7x in 2024 to 23x in 2025 — nearly doubling in twelve months, even as China's own AI investment grew a healthy 33% over the same period. The US simply grew faster from a much larger base, at 162% year-on-year. This is not a gap that is closing. By the rawest measure of capital commitment, it is widening at an accelerating rate.

The Genuine Counter-Signal

China is closing the performance gap while spending a fraction of the capital

Capital concentration and capability concentration are not the same story

The honest complication in this data is that China's underinvestment relative to the US has not translated into a comparable performance gap. Stanford HAI's research found the benchmark performance gap between leading US and Chinese AI models narrowed from double digits in 2023 to near parity by 2024, and separately to just 2.7 percentage points by 2025 — despite China spending roughly 23 times less in private capital. China also leads decisively in AI-related academic publications and patent filings, generating more than six times the number of generative AI patent inventions the US produced between 2014 and 2023.

This suggests the US advantage is specifically a monetization and capital-deployment advantage, not a pure capability advantage. China appears able to produce competitive AI research and models with dramatically less private capital — through different mechanisms, including substantial direct state investment (China's government alone contributed an estimated ¥345 billion, roughly 39% of the country's total AI investment, in 2026) and a different commercialization model. The question your original framing raises — why can't anyone monetize like the US — has a sharper answer once you see China closing the capability gap without closing the capital gap at all.

11.7x → 23x
US/China investment multiple, 2024 to 2025
6x
China's lead in generative AI patent filings vs the US
2.7pts
Remaining US-China AI benchmark performance gap
Why the US Monetizes Differently

Four structural reasons,
not one simple explanation.

🏦 Deep, Liquid Capital Markets

US venture and private equity markets can deploy hundred-billion-dollar single rounds — OpenAI alone raised $40 billion at a $300 billion valuation in 2025 — at a speed and scale no other capital market currently matches, turning research breakthroughs into funded companies almost immediately.

☁️ Hyperscaler Concentration

A handful of US hyperscalers — Microsoft, Google, Amazon, Meta — are projected to spend over $345 billion on AI infrastructure in 2026 alone, vertically integrating compute, model development, and distribution in a way that converts capex directly into monetizable product at unmatched scale.

🌐 Global Distribution Reach

US AI products reach a global customer base from day one through existing platform distribution — cloud infrastructure, app stores, enterprise software relationships — letting American companies monetize internationally in ways research-strong but distribution-limited ecosystems elsewhere cannot easily replicate.

📊 Revenue Realization Speed

Anthropic's annualized revenue reached $30 billion by April 2026 and OpenAI's roughly $25 billion — both achieved within a few years of widescale product launch, a monetization speed that reflects mature enterprise software sales infrastructure most other AI ecosystems are still building.

The Concentration Risk This Creates
A global technology wave this dependent on a single country's capital markets and hyperscaler infrastructure carries genuine systemic risk. If US AI capital expenditure cools — whether from a funding bubble correction, regulatory shift, or simple normalization after 2025's 162% growth spike — there is no comparably deep alternative capital pool currently positioned to absorb the resulting slowdown globally. The rest of the world's AI ecosystems are largely complements to, not substitutes for, US-led capital deployment.
What to Watch
Whether China's capability-without-capital model — strong model performance and patent output relative to its much smaller private investment base — eventually translates into comparable monetization, particularly in markets where US platforms have limited reach. If China begins closing the deal-value gap, not just the benchmark gap, that would be the more significant structural shift than any single model release.
The Honest Read

The original "$1 quadrillion by 2040" framing should be retired, but the instinct behind it — that monetization concentration is the defining feature of this AI cycle — was directionally right and arguably understated. A 65/25/10 US/China/rest-of-world split implied a contest with three real participants. The actual private investment data shows something closer to a single dominant capital pool (the US), a competitive but vastly under-capitalized second player (China) that is nonetheless closing the capability gap through other means, and a "rest of world" category that, at 2% of H1 2025 generative AI deal value, is not really a meaningful third participant in capital terms at all.

This data also complicates any simple "US wins" conclusion. If capital intensity alone determined outcomes, China's near-parity benchmark performance on roughly 4% of US spending would be impossible to explain. The more accurate read is that the US has built the deepest monetization machine for AI specifically — capital markets, hyperscaler infrastructure, and global distribution working together — while genuine research and model-building capability is more evenly distributed than the investment figures alone suggest.

The NGE View

The verdict.

What We Believe
Verify big numbers before building a thesis on them. The $1 quadrillion figure doesn't exist in credible forecasting, but the real, verified numbers — a 23x investment gap, 97% of generative AI deal value — make a stronger and more defensible case than the unverified figure would have.
The US monetization advantage is structural, not accidental — deep capital markets, hyperscaler vertical integration, and existing global distribution infrastructure compound together in a way that is genuinely difficult for any other ecosystem to replicate quickly, regardless of research talent.
Don't mistake capital concentration for capability concentration. China's near-parity benchmark performance on a fraction of US spending is the most important counter-signal in this data, and suggests the monetization gap may be more fragile than the investment numbers alone imply.
This concentration is itself a risk worth pricing, not just an opportunity to chase. A technology wave this dependent on one country's capital markets has limited shock absorption if that specific capital pool cools, which is a different risk profile than a genuinely globally distributed industry would carry.

The instinct that started this letter — that nobody monetizes like the United States — turns out to be correct, better documented, and more extreme than the original framing suggested. But the more durable insight sits underneath the headline number: monetization dominance and capability dominance have quietly become two separate races, and the US is decisively winning only one of them. Whether that gap closes, widens, or simply persists through the rest of this decade is probably the single most consequential open question in global technology economics right now — far more consequential than any single forecast number, real or invented.

NGE · A Futuristic Investment Letter

Long-horizon thinking on capital, technology, and the forces shaping the next decade of wealth creation. Written from first principles. Not consensus. Not noise.

— Pawan Bhatia · NextGen Economics · Bangalore, India