Artificial intelligence is advancing faster than the political institutions expected to govern its consequences.
More than 200 economists and researchers, including 15 Nobel laureates and specialists from OpenAI, Anthropic and Google, issued a joint statement calling for immediate preparation for AI’s economic impact. The signatories warned that the technology could produce a transformation larger than the Industrial Revolution, but within a far shorter period.
Anton Korinek, a University of Virginia professor who helped organize the initiative, summarized the urgency clearly.
“Steam, electricity, and computers each gave societies decades to adapt. AI may give us only a few years,” Korinek said.
He added, “We cannot improvise our strategy and institutions in the middle of the transformation; waiting for certainty means arriving too late.”
Those warnings should not be dismissed as speculative alarmism. They come from people actively developing, financing and studying the systems at the center of the transformation.
For years, the public discussion about AI employment has been trapped between two oversimplified predictions. One side claims automation will destroy work on an unprecedented scale. The other insists technology has always created more jobs than it eliminated.
Both arguments avoid the central policy question.
Even when new technology eventually creates employment, the transition can be deeply disruptive. New jobs may emerge in different cities, require different qualifications or offer lower wages. A displaced administrative worker cannot immediately become a machine-learning engineer. A regional economy built around customer service, accounting or routine analysis cannot reinvent itself overnight.
The United States should therefore prepare for disruption without assuming that mass unemployment is inevitable.
Congress should begin by funding continuous research on which occupations and communities are most exposed. Traditional labor statistics are too slow for a technology that changes capabilities every few months. Policymakers need current data on job displacement, wage effects, productivity, hiring and geographic concentration.
Education policy must also move beyond the familiar demand for more science and technology graduates. AI will affect legal services, finance, media, healthcare, logistics, government administration and creative industries. Workers across the economy need practical training in how to use automated systems, evaluate their output and retain responsibility for consequential decisions.
The tax system deserves scrutiny as well. If AI allows companies to generate far greater revenue with fewer employees, a tax structure heavily dependent on individual labor income may become less sustainable. This does not automatically justify a new “robot tax,” but it does require serious analysis of how automation could affect public revenue and social insurance programs.
Companies developing AI should participate in financing the transition. They will receive enormous economic benefits from public education systems, government research, energy infrastructure and the courts that protect intellectual property. Asking them to contribute to worker training and economic adjustment is not hostility to innovation. It is a condition for durable public legitimacy.
The goal should not be to freeze technology. Attempts to halt AI development would likely fail and could weaken the United States in global competition. The objective should be to ensure that productivity gains translate into broader prosperity rather than extreme concentration of income and economic power.
AI policy has focused heavily on safety testing, misinformation and national security. Those issues matter, but the daily economic consequences may become visible much sooner.
Washington still has time to act before disruption reaches its full scale. The experts’ statement is valuable because it rejects the most dangerous option, waiting until displaced workers and struggling communities force the government to respond in crisis mode.
