- Bloomberg senior editor Chris Anstey argued that today’s AI boom is missing the “secret sauce” that powered the 1990s internet-era economic expansion.
- The 1990s featured budget surpluses, declining tariffs, and immigration growth — fiscal conditions absent from today’s economy, which carries growing deficits and rising trade barriers.
- Economists dispute the Trump administration’s claim that AI-driven productivity justifies lower interest rates, noting that recent productivity gains predate AI deployment.
- Federal Reserve Governor Michael Barr countered that AI investments would increase borrowing demand, putting upward pressure on rates.
What Happened
Bloomberg News senior editor Chris Anstey published an analysis on March 27, 2026, titled “The AI Boom Is Missing the Secret Sauce of the 1990s.” The piece argues that today’s AI investment boom lacks the macroeconomic conditions that made the 1990s internet revolution economically productive. He expanded on the argument during a Bloomberg This Weekend segment with anchors David Gura and Christina Ruffini on March 29.
Anstey’s central argument is that massive AI capital expenditure alone cannot replicate the growth-without-inflation environment of the late 1990s. The fiscal and trade conditions that enabled that era’s expansion have fundamentally reversed.
Why It Matters
The comparison between AI and the 1990s internet boom has moved from academic debate to active policy dispute. President Donald Trump, Treasury Secretary Scott Bessent, and Federal Reserve chair nominee Kevin Warsh have each argued publicly that AI can deliver a productivity surge similar to the internet era, and that this justifies a looser monetary policy.
Bessent stated: “Our nation can see productivity boom like we did in the ’90s when we are not encumbered by a Federal Reserve which throws the brakes on.” The administration’s position is that keeping interest rates low would unlock AI’s full economic potential, just as low rates helped fuel the dot-com expansion.
Anstey’s analysis challenges that framing directly. He contends that the 1990s boom was not driven by technology investment alone — it required a specific set of macroeconomic conditions that no longer exist in the current environment.
Technical Details
During the mid-1990s, Federal Reserve Chair Alan Greenspan suspected that official productivity data was underestimating the impact of emerging information technologies. He persuaded his colleagues in September 1996 to hold off on raising rates despite conventional indicators suggesting the economy was overheating. The gamble paid off. Inflation remained below 2% for 17 consecutive months between 1997 and 1999, while unemployment dropped to 3.8% by April 2000.
The administration’s current argument references only this dovish phase of Greenspan’s tenure. Economist Dario Perkins identified a critical gap in the narrative: “Warsh and Bessent talk only about the dovish 1995/96 version of Greenspan; they overlook the hawkish 1999/2000 variant.” Greenspan eventually raised rates from 4.75% to 6.5% between mid-1999 and early 2000 as inflation risks materialized.
Joe Brusuelas, chief economist at RSM, challenged the premise that AI is already driving productivity gains. He stated that recent productivity improvements “are not because of artificial intelligence” but instead reflect pandemic-era automation investments that preceded the current AI wave by several years.
Federal Reserve Governor Michael Barr added another dimension to the debate, arguing that AI investments would actually increase borrowing demand across the economy. This would put upward pressure on interest rates, not downward — the opposite of what the administration’s thesis requires.
Who’s Affected
The debate directly affects monetary policy and, by extension, borrowing costs for businesses and consumers. If the Fed adopts the administration’s view and cuts rates to support AI-driven growth, it risks fueling inflation. If it maintains higher rates, AI-dependent companies face greater financing costs during a capital-intensive buildout phase.
Investors in AI infrastructure face a related question. Global AI capital expenditures plus venture capital investments have already exceeded $600 billion. Whether that spending delivers returns depends in part on macroeconomic conditions that sit outside any individual company’s control.
What’s Next
The fiscal environment presents the starkest contrast between the two eras. The 1990s featured budget surpluses, declining tariffs, and surging immigration — all of which supported non-inflationary growth. Today’s economy carries growing deficits projected to push national debt to 120% of GDP by 2035, rising trade barriers, and immigration restrictions. Whether AI productivity gains can overcome these structural headwinds will likely shape Federal Reserve rate decisions and broader economic policy through the second half of 2026.
Related Reading
- AI Boom Drives Q1 2026 Venture Funding to Record $300 Billion Globally
- BlackRock CEO Larry Fink Warns AI Boom Could Widen Wealth Inequality
- EFF: Publishers Blocking Internet Archive Threatens Decades of News History
- DeepMind Study: LLM Rewrites Game Theory Algorithms and Outperforms Human Experts
- Walmart Patents AI Pricing Systems That Use Customer ID Data for Markdowns
