The Federal Reserve is "obsessed" with AI but dares not take another "Greenspan-style gamble."

Federal Reserve officials are increasingly focused on the economic transformation potential of artificial intelligence, but they remain cautious about whether it has triggered a productivity revolution. Although AI investment is driving a significant portion of U.S. economic growth, central bank policymakers are still in the "it's too early to tell" phase, unwilling to make major policy bets like during the internet boom of the 1990s. At that time, then-Federal Reserve Chairman Alan Greenspan believed that innovation could promote faster economic growth without triggering inflation, and maintained low interest rates on that basis. U.S. Treasury Secretary Scott Bessent told CNBC last month that "the implementation of AI will really start to have an impact on productivity in the first half of 2026." He believes that the next Federal Reserve Chair should keep an "open mind" about the possibility of an AI productivity boom to avoid repeating the mistakes that stifled the internet boom. Among the five candidates for the Federal Reserve, four have recently expressed support for the AI productivity argument. Kevin Hassett, Director of the National Economic Council under Trump, stated that AI is enhancing worker productivity at an "astonishing rate," while BlackRock executive Rick Rieder noted, "We are in the midst of a productivity revolution." This debate comes at a critical moment for the Federal Reserve. Chair Jerome Powell's term will end in six months, and inflation remains above target levels, leading to divisions in the central bank's interest rate policy. The potential impact of AI on the labor market further complicates policymaking. Positive Signals in Productivity Data Recent research from the St. Louis Fed found that since the release of ChatGPT three years ago, generative AI may have increased labor productivity by 1.3%. Researchers discovered a clear correlation between high AI adoption rates in industries and productivity gains by regularly surveying workers about the time saved using AI. Alexander Bick, a co-author of the study, stated, "I am surprised that this signal has emerged so clearly at the industry level. The correlation is already there." Anna Wong, Chief U.S. Economist at Bloomberg Economics, pointed out that a productivity boom is a "dream come true" for central banks and a "unicorn" for macroeconomists, appearing only once every few decades. She believes that while macro-level evidence is still unclear, micro-level evidence is beginning to emerge. Corporate practices also support this view. Peter Capuciati, CEO of HVAC equipment AI service company Bluon Inc., estimates that its AI tools can save technicians up to 8 hours of work time per week. Currently, about 160,000 technicians use the free version, while 13,000 pay for the full service. Data Quality Limits Accurate Judgments Kristina McElheran, a scholar at the University of Toronto studying the future of AI and work, pointed out that the lack of "detailed, high-fidelity data on corporate AI usage" is a fundamental issue currently faced. Many compelling studies are based on "really problematic information." "We are blindly flying into this AI revolution," McElheran stated. "We do not have the statistical data needed for policymaking, nor do we have the statistical data needed for managers." "Modelers can only" use past trends and try to apply them to the rapidly unfolding events in front of us." This data dilemma has made Federal Reserve officials more cautious in formulating policies. While technological changes often take years to impact the economy and be reflected in data, central banks are under pressure to make judgments at critical moments. Current Federal Reserve Governor Christopher Waller has expressed relative caution, stating that he "has no doubt" that AI will drive economic growth and "hopes" for sustained productivity growth. Vice Chair Michelle Bowman, who is responsible for oversight, has discussed AI more in the context of regulatory work. Employment Shock Raises New Concerns The dual nature of AI technology adds complexity to policymaking. While technological advancements typically drive productivity improvements, they may also impact the labor market. The Federal Reserve's recent Beige Book survey indicates that AI is dragging down hiring demand, particularly for entry-level positions. Capital Economics research points out that the information technology sector, as an early adopter of AI, is contributing a larger share to U.S. economic growth while its employment numbers are shrinking—this is both evidence of productivity gains and reflects the risks of technological diffusion. Julia Coronado, founder of Macropolicy Perspectives LLC, notes that unlike the 1990s internet boom when companies used innovation to expand employment, businesses today are more likely to use AI to reduce headcount. Robert Gordon, a professor at Northwestern University and author of "The Rise and Fall of American Growth," states that he is willing to "shift from the usual pessimism about productivity." The 85-year-old scholar believes that AI could bring about faster growth than in the past few decades, "probably at a moderate level of around 2%, and may not reach 3% for a long time, perhaps one or two years." However, Gordon is also concerned that the upcoming era may have a darker side, as AI could create "a series of new social problems accompanying white-collar unemployment," and in "a society where white-collar jobs are the aspiration of every young person," this will pose severe challenges.