Most eyes were elsewhere in late August when the Bureau of Economic Analysis quietly updated its second-quarter data. However, the updated GDP number, which increased from a conservative estimate to 3.3% annualized growth, came as a shock to a dimly lit room. It was a different story, not just a larger number.
There was more to that story than just strong consumer demand. Beneath the spreadsheets, something more profound was taking place. Investment in businesses, particularly in intellectual property, increased at a rate that is uncommon outside of boom cycles. The chauffeur? Artificial intelligence was a technology that many people had prematurely dismissed as hype.
| Key Insight | Detail |
|---|---|
| Revised U.S. GDP Growth (Q2 2025) | 3.3% annualized |
| Main Growth Drivers | Consumer spending, AI-led business investment |
| AI Investment Growth | 12.8% annual rate in intellectual property spending |
| Labor Productivity Trend | U.S. lead widening over UK, Germany, and Japan |
| Core CPI Inflation | 3.1% annually |
| Projected Productivity Growth | 4% annually over the next five years |
| NASDAQ Q3 Performance | +22% rebound after Q2 drawdown |
| Fed Policy Outlook | Potential 50–75 bps rate cuts by end of 2025 |
| Strategic Market Response | Shift to high-quality growth stocks and short-duration bonds |
| Source References | Capital Group, BEA, AInvest, Bureau of Labor Statistics, JPMorgan |
AI infrastructure has become the focus of a new investment rhythm that has emerged in recent months. The money is going into data centers, energy supply agreements, and compute infrastructure rather than frolicking IPOs or virtual reality tricks. Innovation is no longer abstract. The economic landscape is now shaped by copper wires, cooling systems, and extremely efficient code.
Modine Manufacturing, a little-known business that creates temperature control systems for data centers, provided one especially instructive example. In just a few weeks, their stock increased by more than 22%. Constellation Energy, which provides Microsoft and Meta with nuclear power, saw an even greater increase. These are indicators of something structural rather than merely passing trends.
We were dialing up to access email the last time technology received this much financial attention. Today’s businesses, however, are not using borrowed funds like they did during the dot-com boom. Their surplus capital, which has been significantly enhanced by years of leaner operations and stronger balance sheets, is being deployed. Additionally, they’re aiming for efficiency rather than merely eye-catching branding.
Economists frequently follow productivity trends like ghost trails through fog, and by most accounts, we are about to enter a rare upswing. We are witnessing the beginnings of a productivity boom that may be especially advantageous for long-term growth for the first time in more than ten years. If it continues, annual productivity gains in the US could get close to 4%, which would be incredibly effective at reducing inflation and raising real wages.
I was amazed to see how many colleagues, who are frequently dubious of hype, were actively reorganizing workflows around AI tools in addition to experimenting with them at this summer’s economics conference in Dublin. The technology had quietly, almost without fanfare, transitioned from curiosity to backbone.
I couldn’t get that moment out of my head.
It is supported by the data. Over the last two years, the cost of implementing AI has drastically decreased. With the help of simplified tools, a small team can now handle tasks that previously required server farms and months of engineering. Additionally, these tools are becoming increasingly flexible, being used for everything from legal analysis to logistics.
At the same time, the American labor market is rapidly changing. Yes, there have been layoffs at big tech companies, but not randomly. These layoffs frequently accompany a shift toward a leaner, AI-enhanced workforce and away from the bloated hiring practices of the pandemic era. The trend is remarkably similar to the revolution in personal computing, when millions of new clerical jobs subtly emerged while some clerical roles vanished.
According to a McKinsey study, between 1970 and 2015, PCs created over 19 million new jobs while displacing 3.5 million existing ones. Tasks may be replaced by AI as well, but roles will probably be enhanced rather than completely replaced. Coding, legal research, and even customer service tasks are being streamlined to free up employees to perform intricate, creative, or social tasks.
Economists are now creating increasingly sophisticated models of labor impact by emphasizing tasks rather than job titles. Additionally, this opens up a particularly creative frontier for progressive employers: retraining teams, redesigning roles, and reallocating time toward high-value outcomes.
Even though trade policy and tariffs still occasionally create problems, especially in the manufacturing and logistics sectors, the general trend is still strong. The bond market has not panicked even though inflation has remained at 3.1%. The yield on the 10-year Treasury is currently around 4.2%, indicating that investors see room for rate cuts—but not desperately.
The same is true of equity markets. Following a rough period in Q2, the NASDAQ recovered 22% in Q3. Though the tone has changed, analysts are still cautious and have slightly trimmed their full-year earnings projections. We’re evaluating how much AI can change profitability rather than speculating about its significance.
Additionally, investment portfolios are changing. High-quality growth—businesses with strong cash flow, remarkably transparent strategies, and the pricing power to weather macro or policy shocks—is the trend. In the meantime, inflation-linked securities and short-duration bonds are becoming more popular as low-risk hedges.
The duality of preparation and resilience holds the secret. The economy is adapting swiftly, but it is not overburdened. The demand from consumers is still strong. The rate of productivity is increasing. Monetary policy is accommodating but cautious. Beneath it all, AI still behaves like a swarm of bees, with each agent moving quickly and small but collectively changing the entire environment.
Additional data points will surface in the upcoming months. Some might indicate turbulence. This change may be further validated by others. However, for the time being, one updated figure has reinterpreted resilience as an adaptive leap forward rather than a return to prior strengths.
With concentration, not with fireworks. With guidance, not with proclamations.
