As AI expands, Americans have doubts

Summary of As AI expands, Americans have doubts

by Marketplace

25m•January 26, 2026
Listen to Episode

Overview of As AI expands, Americans have doubts

Marketplace explores two macroeconomic threads: (1) a surge in gold and what central-bank buying may signal about confidence in the U.S. dollar, and (2) the rapid expansion of AI infrastructure — how it’s powering economic growth, provoking public unease, and stressing local energy grids as data centers multiply.

Key takeaways

  • Gold has rallied to record highs (futures topped $5,000/oz), and central banks are increasing gold holdings. Gold now roughly outvalues U.S. Treasury bonds in those holdings; dollar-denominated assets fell below half of global reserves in absolute terms.
  • Momentum and geopolitical/economic anxiety (tariffs, debt, conflicts) are driving demand for gold and other metals, but momentum rallies are unpredictable.
  • Durable goods orders jumped in November (~5% headline), driven largely by airplane orders; excluding transportation, orders rose ~0.5%. AI-related spending (data centers, hardware, software, R&D) is a meaningful component of recent business investment.
  • The Federal Reserve Bank of St. Louis estimates AI-related components contributed just under 1 percentage point to U.S. GDP in the first three quarters of last year — a material amount in a $31 trillion economy.
  • Public sentiment toward AI is skewed toward concern rather than enthusiasm (Pew Research). Local opposition to AI data-center projects has shut down or slowed multiple developments.
  • Data centers already consumed ~4% of U.S. electricity in 2024; projected growth (+133% by 2030) could push that to ~10% of national power consumption within five years, raising local grid, cost, and environmental concerns.

Topics discussed

Gold and central-bank behavior

  • Why gold is rising: investor angst (geopolitics, trade, debt), perceived safe-haven status.
  • Central banks increasing gold allocations (gold ~28% of their holdings); diversification away from dollar assets into other currencies and gold is happening but not necessarily a wholesale sell-off of U.S. assets.
  • Momentum effect in commodities and difficulty in timing reversals.

Durable goods & business investment

  • November durable-goods surge largely driven by Boeing/aircraft orders; machinery orders (including for data centers and computing) are up.
  • Factors boosting equipment investment: AI demand, older installed capital, favorable tax deductions.
  • Uncertainty (tariffs, trade) still weighs on long-term investment decisions.

AI’s economic impact and public doubt

  • AI’s GDP contribution is measurable but small as a share (just under 1 point) — nonetheless significant in dollar terms.
  • Universities and public institutions are integrating AI (example: California State’s $17M OpenAI deal covering ~500k users), causing classroom and ethical debates: cheating, distraction, mental-health effects.
  • Local community resistance to data centers: concerns over noise, traffic, pollution, property values, and increased electricity bills (example: Hobart, Indiana; residents saw ~26% electricity price rises; city approved Amazon project with $47M payment to city).

Data centers: types, function, and impacts

  • Hyperscale vs. colocation: hyperscale (single-tenant, big cloud providers — training models; located where power/real estate costs are low). Colocation (multi-tenant “apartment building” for compute) serves banks, hospitals, apps, and local low-latency needs — often in urban areas.
  • Data-center realities: noisy, hot, energy- and water-intensive; use hot/cold aisle designs, large battery arrays, specialized cooling (e.g., for NVIDIA chips), and substantial power capacity (comparable to small towns).
  • Co-location operators argue they deliver better energy efficiency at scale versus many enterprises running private on-prem data centers.

Notable quotes and soundbites

  • “Things go up, then they go up some more.” — Jay Hatfield, on momentum rallies.
  • Professor Enid Baxter Rice: “I’m concerned. I’m really scared.” — on student wellbeing and AI in classrooms.
  • Local resident on inescapable AI integration: “What do you do? Just stop using your phone? No.” — capturing everyday frustration.
  • Spencer Kaplan: “I’m rooting for natural intelligence.” — reflecting anthropological skepticism of AI culture/inevitability.
  • Volkswagen CEO: “Given an unchanged tariff burden... large additional investment cannot be funded.” — highlighting how tariffs influence capital decisions.

Data & figures (from episode)

  • Gold futures exceeded $5,000 per troy ounce (record).
  • Gold makes up ~28% of central-bank holdings; U.S. dollar assets under 50% in absolute terms (IMF data referenced).
  • Durable goods orders: +5% (November headline); +0.5% excluding transportation.
  • AI contribution: just under 1 percentage point to GDP (first three quarters last year).
  • Data centers electricity use: ~4% of U.S. power in 2024; forecast to increase ~133% by 2030 (~10% of power).
  • Cal State–OpenAI deal: $17 million covering ~500,000 students and faculty.
  • Hobart, Indiana: area electricity prices rose ~26%; city approved an Amazon data center with a $47 million payment.

Implications and recommended follow-ups

For policymakers:

  • Accelerate grid expansion and resilience planning to accommodate major data-center demand.
  • Improve community engagement and transparent benefit-sharing for local projects (jobs, infrastructure support, energy mitigation).

For businesses and institutions:

  • Consider co-location for efficiency rather than bespoke on-prem data centers.
  • Invest in worker/student training and clear policies on AI use to manage academic integrity and mental-health impacts.

For citizens and local governments:

  • Scrutinize data-center proposals’ real economic benefits versus costs (infrastructure strain, property impacts, energy prices).
  • Demand clarity on environmental mitigation (water usage, emissions, power sourcing).

Bottom line

AI is a measurable driver of economic activity (notably via data-center and equipment spending), but its infrastructure expansion exposes tensions: real economic gains vs. local environmental, social, and energy-system stresses. At the same time, global financial nervousness has fueled gold’s rise and a modest rebalancing of central-bank reserves — a signal worth watching for shifts in global confidence and risk appetite.