What The AI Boom Is Doing For Americans

TL;DR

Prime Big Deal Days · Oct 6–7Offer from Amazon

Get office and shipping supplies delivered free — and shop member deals

• Fast, free delivery on millions of items
• Access to Prime Big Deal Days deals on October 6–7
• Prime Video, Amazon Music and more included

Start your free Prime trial →Free trial for eligible customers · Cancel anytime

As an affiliate, we earn on qualifying purchases.Hide

In an Oct. 5 report, economist Noah Smith argues that the U.S. AI boom is already benefiting Americans through construction jobs, public revenue, economic activity and consumer use of AI. The figures he cites indicate possible gains, but the long-term effects, distribution of benefits and local costs remain unsettled.

Economist Noah Smith argued in an Oct. 5 report that the U.S. AI boom is already benefiting Americans through construction work, local tax revenue, economic demand and consumer use. His case challenges Paul Krugman’s view that AI has brought few gains to ordinary people, though the evidence Smith presents does not settle how benefits and costs will be distributed over time.

Smith points to construction employment and wages as signs that the data center buildout is supporting workers. He says construction employment and the sector’s share of the workforce have risen since the AI boom began in late 2022, and that real wages for construction workers have increased since mid-2022. He also cites estimates that data center projects have a significant positive effect on construction employment. The figures and estimates are presented in his report; they do not, on their own, establish how much of the change was caused by AI investment.

Smith also argues that data centers can generate property and sales taxes, corporate taxes and fees for state and local governments. Those revenues may support public services and infrastructure. He cites Loudoun County, Virginia, as an example of a community that has used data center-related growth to fund public amenities. Whether a project leaves a community better off depends on its local costs as well as the revenue it brings.

For the broader economy, Smith points to a figure shared by economist Jason Furman: investment in information-processing equipment and software accounted for 92% of U.S. GDP growth in the first half of 2025, while those categories made up 4% of GDP. That comparison concerns a specific half-year period and category of investment; it does not isolate AI’s contribution. Smith also cites research estimating $172 billion in annual U.S. consumer surplus from AI, a survey-based estimate of the value users receive beyond what they pay.

Where AI Spending Reaches Households

The argument shifts the debate about AI’s public value beyond whether the technology creates large numbers of permanent jobs. Data center projects may support construction employment during the buildout, while taxes and fees can contribute to local budgets. AI tools may also provide value directly to people who use them. These are distinct channels of benefit, and none guarantees that gains will reach every household or outweigh the costs in a particular place.

The economic stakes extend beyond the technology sector. If AI-related investment forms a sizable share of overall economic growth, a slowdown could affect suppliers, workers and communities that depend on continued construction and spending. Conversely, a sustained buildout could support demand for labor and services. Smith’s account frames these as current economic effects, while acknowledging that their net value depends on costs such as pressure on electricity systems and the possibility that investment displaces other activity.

For readers, the practical question is how the gains are shared. Local tax policy, project agreements, wages and energy costs can shape whether nearby residents benefit. Smith says governments can seek payments or other community benefits from developers, including support for infrastructure, schools and lower utility costs. The report does not establish that these arrangements are uniform or that they fully offset the burdens data centers may create.

From Internet Skepticism to AI

Smith places the current disagreement in a longer debate over whether new technology improves everyday economic life. He recalls Paul Krugman’s 1998 prediction that the internet’s economic impact might prove little greater than that of the fax machine, then argues that internet access and e-commerce later produced gains for businesses and consumers. Smith also notes Krugman’s 2011 observation that American kitchens had changed little since 1957, arguing that household appliances introduced in the intervening decades were relevant counterexamples.

That history informs Smith’s challenge to Krugman’s more recent claim that AI is largely a technology for corporate owners, with few benefits reaching regular Americans. Smith’s report focuses on four possible channels: jobs tied to construction, government revenues, broader economic demand and direct consumer use. It is an argument about current benefits, not a forecast that AI will distribute income fairly or produce lasting gains.

Smith says the data center buildout involves substantial labor even though construction accounts for only about a third of a data center’s cost, citing an estimate by economist Stijn Van Nieuwerburgh. He also relays a Goldman Sachs estimate that 500,000 additional construction and trades jobs may be needed by 2030 to sustain the buildout. That is a projection, not a count of jobs already created. Smith’s report attributes other figures to sources including Virginia’s Joint Legislative Audit and Review Commission, the Tax Foundation, Pew Research Center and academic research.

“This is a technology of, by and for oligarchs, with hardly any of the benefits trickling down to regular Americans.”

— Paul Krugman, as quoted in Smith’s report

Costs and Long-Term Effects

Smith says the long-term effects of AI on the distribution of income and wealth are not known. The employment and wage trends he cites are consistent with strong demand for construction labor, but they do not by themselves show how many jobs resulted specifically from data center projects or how durable those jobs will be. The cited estimate of 500,000 additional trades jobs by 2030 is a projection.

There is also no settled measure in the report of the net effect on local communities. Data centers can strain electricity systems, use water, create noise and compete with other projects for labor and financing. Tax receipts and community agreements may offset some costs, but the result will vary by location. It also remains unclear how much of the economy’s recent growth would have occurred through other investment if AI spending had been lower.

The $172 billion consumer-surplus estimate comes from survey-based research cited by Smith. It is an estimate of consumer value, not cash paid to households or government revenue. Public attitudes toward AI also remain negative, according to the polling Smith discusses, despite reported use of the tools and perceived productivity benefits. The report does not resolve concerns about job displacement, misuse or overreliance on AI.

Tracking Jobs, Bills and Benefits

The next evidence will come from labor, investment and public-finance data as data center construction continues. Employment and wage figures can show whether demand for construction workers persists, while state and local budgets can indicate how much tax revenue projects generate and how governments spend it. The projected need for additional trades workers by 2030 will be tested against actual hiring and construction activity.

For communities considering projects, the terms of local agreements and the effects on power, water and infrastructure will help determine whether residents share in the gains. For the wider economy, future growth data may clarify whether AI-related investment remains a major source of demand. The longer-term balance will depend on consumer value, business productivity, job changes and costs that are not yet fully measured.

Key Questions

What does Noah Smith say Americans are gaining from the AI boom?

Smith points to construction jobs and wages, local tax revenue, economic demand and consumer value from using AI tools. His report argues these benefits are already visible, while acknowledging that their long-term distribution is unknown.

Does the report show that AI created all the construction jobs cited?

No. Smith cites rising construction employment and wages and says estimates find a significant positive effect from data center construction. Those trends do not establish that AI projects caused every observed change.

What does the $172 billion figure measure?

It is a survey-based estimate of annual U.S. consumer surplus from AI: the value users receive beyond what they pay. It is not cash distributed to households or a measure of government revenue.

What costs of data centers does Smith acknowledge?

He notes that data centers can put pressure on electricity systems and create noise. The report also discusses the possibility that the buildout draws labor and financing away from other economic activity. The local balance of benefits and costs can vary.

Source: rss