Economist Noah Smith is challenging the prevailing narrative that the AI boom is solely benefiting tech oligarchs. In a recent post on his blog, Smith argues that Paul Krugman’s assertion that AI "does nothing" for regular Americans ignores substantial, measurable economic benefits. Smith points to a surge in construction employment, significant tax revenue for local governments, and a massive increase in consumer utility as proof that the infrastructure buildout is trickling down in meaningful ways.

Construction Jobs and Wage Growth

Smith highlights that construction employment has risen since the AI boom began in late 2022, even amidst strict immigration policies that have removed a significant portion of the workforce. He notes that real wages for construction workers have increased, with data showing a jump in wages relative to the national average in 2025. This suggests healthy labor demand driven by the physical buildout of data centers, which Goldman Sachs estimates will require 500,000 new construction and trades jobs by 2030.

Local Tax Windfalls and Public Goods

The article details how data centers generate substantial property, sales, and corporate taxes that fund local public goods. Smith cites Loudoun County, Virginia, as a case study, noting that the proliferation of data centers has helped transform the area into a tech destination with improved schools, libraries, and roads. He argues that local governments can leverage their power to demand community benefit agreements, ensuring that infrastructure costs are covered and residents see direct improvements in services and amenities.

The Macro-Economic Buffer

Smith posits that the AI boom has acted as a critical positive demand shock, offsetting the negative economic effects of recent political instability and tariff uncertainty. He references data from Jason Furman indicating that investment in information processing equipment was responsible for 92% of GDP growth in the first half of 2025. Without this AI-driven investment, Smith argues, the broader economy might have suffered a severe downturn, leading to significant job losses for ordinary workers.

Consumer Surplus Beyond Infrastructure

Beyond the physical buildout, Smith emphasizes the direct utility Americans are deriving from AI tools. He cites a study by Brynjolfsson et al. from April 2025, which estimates a total annual consumer surplus of $172 billion in the United States. This figure exceeds the combined run-rate revenue of major AI labs like Anthropic and OpenAI, demonstrating that the technology is already providing immense value to households through medical advice, productivity gains, and creative assistance.

Key Takeaways

  • Construction employment and wages are rising due to the data center buildout, despite labor supply constraints.
  • Local governments are using data center taxes to fund schools, roads, and public services, with Loudoun County serving as a prime example.
  • AI investment accounted for 92% of U.S. GDP growth in the first half of 2025, acting as a buffer against other economic headwinds.

The Bottom Line

Krugman’s skepticism overlooks the tangible, immediate benefits of the AI infrastructure boom, from blue-collar job growth to local tax revenue. For builders and policymakers, the focus should shift from debating abstract inequality to capturing the real-world value already being created.