A sharp disagreement has emerged over whether artificial intelligence can unlock unprecedented macroeconomic expansion. While Anthropic chief executive Dario Amodei has suggested the technology could drive yearly growth between 10 and 15 percent, academic researchers Ben Moll of the London School of Economics and Alex Imas from the University of Chicago consider such figures highly improbable over the next decade to fifteen years. Preferring a modest baseline of 4 to 5 percent, Moll and Imas have put money on the line through a public wager that American per-capita real GDP expansion will not touch 15 percent in any single year through 2033. They contend that physical limitations, long construction timelines for hardware, and shifting consumer demand toward scarce human effort will prevent immediate runaway financial yields.
The academic debate comes as Anthropic released an interactive simulation tool on September 10, grounded in a research document titled “Economic Scenarios for Transformative AI” (Korinek et al., 2026). Instead of offering precise predictions, the framework breaks occupations into discrete task components to measure how varying levels of machine capability and adoption speed might alter financial output. To benchmark expectations, the organization surveyed more than 10,000 American citizens during August, revealing that
