Like you, I’ve been blown away by the projected capital investments in AI. The only way I can wrap my head around the numbers is to anchor them to something more tangible in recent history. I keep coming back to the Interstate Highway System, which, adjusted for inflation, cost over $600 billion. Spread over multiple decades (since 1956), its annualized impact was roughly 0.10% of U.S. GDP.
By contrast, moderate projections are at over $1 trillion (with a “T”) for AI investment deployed over the next 10 years, which clocks in at 0.40% of GDP annually. That’s four times the relative scale, compressed into a fraction of the time. This is bigger, faster, and more concentrated. The Interstate was a slow, deliberate public investment with obvious and predictable returns (in hindsight). To sense the scale of impact consider that Interstates make up about 1–2% of U.S. road miles but carry a disproportionate share of vehicle travel (roughly 20–30% of total vehicle‑miles traveled), and this number is substantially larger for heavy‑truck and freight miles. AI investment, by comparison, is a flash flood of private capital. Its sheer velocity demands a different kind of analysis, and no matter how I crunch it, it is difficult to see how all this investment pays off.
If AI investments deliver on its promise, we’re looking at a potential GDP uplift ranging from 0.20% to 0.80% annually. Those might sound like small numbers, but they represent a tectonic shift. Still, that multiplier isn’t automatic. It hinges on diffusion i.e. how quickly AI tools reach and reshape small firms, public services, and lagging sectors. That’s why you’re seeing companies push AI into every nook and crevice of their operations. It’s why executives are predictably shoehorning and measuring the use and value of AI-infused workflows in the most inorganic ways.
The United States is, by deliberate policy, a service-driven economy. Services account for the overwhelming majority of it’s GDP, and that structural reality raises the stakes for AI adoption. Unlike past waves of capital investment that primarily transformed manufacturing or physical infrastructure, AI is positioned as a direct force multiplier for our critical service work infrastructure. Gains in information processing, decision support, customer interaction, and back-office automation translate immediately into productivity where most of our economic value is created.
If the U.S. wants to remain the world’s leading service economy, it cannot afford to lag in AI adoption and diffusion. Other nations must see the same opportunity and are racing to embed AI into finance, healthcare, logistics, and professional services. Adopting AI has become the foundation of maintaining leadership in the very sectors that define the US economy.