A new study argues that artificial intelligence could become the most powerful engine of American prosperity in decades, directly challenging widespread warnings that the technology will cause mass job losses and economic upheaval. The report, titled «Boomsday Not Doomsday,» counters doomsayers who predict that AI will replace workers on a massive scale, including Anthropic CEO Dario Amodei's forecast of a white-collar «bloodbath» and Senator Bernie Sanders' warning that AI could wipe out «100 million American jobs.» Instead, the study contends that AI will lower the cost and accelerate the production of nearly everything, from food and housing to healthcare and education, creating new opportunities and raising living standards.
Fears of technological unemployment are nothing new, the authors note. In 1900 more than one in three Americans worked in agriculture; today fewer than 2 percent do, yet the country produces ten times more food on one-third as much farmland. The economy did not collapse but instead created hundreds of millions of jobs that were previously unimaginable. The pattern repeats with each major leap forward, from electricity to the internet, and AI is no exception, they argue.
Current data supports this optimistic view. The unemployment rate stands at a low 4.3 percent. Between 2022 and early 2025, workers in the most AI-exposed occupations actually experienced smaller increases in unemployment than those in the least exposed roles. Meanwhile, demand for AI skills is surging, with job postings seeking them jumping 111 percent in a single year. Americans already value AI tools at an estimated $172 billion annually, based on what they would need to be paid to give them up, according to the study.
The practical benefits of AI are already visible across the economy. The Mayo Clinic has used AI to identify pancreatic cancer up to three years earlier than conventional methods. Physicians at The Permanente Group using AI-powered note-taking save roughly one hour per day, allowing more time for patient care. A Kentucky master electrician with no coding background used AI to build a website that helps homeowners avoid unnecessary and expensive electrical panel upgrades. The study points to similar patterns in education, construction, and manufacturing, where AI helps workers turn know-how into ownership.
However, the study warns that the real risk is not that AI moves too fast but that the United States moves too slowly. While more than half of American adults have tried generative AI, only 28.3 percent use it regularly, ranking the country 24th out of 30 nations measured. France and the United Kingdom rank higher. Meanwhile, China is deploying AI aggressively. In 2024 China installed 295,045 industrial robots compared to America's 34,164. China also produces some of the most capable open-source AI models and enjoys strong public optimism, with three out of four Chinese citizens expecting AI to improve their economy.
State legislatures have introduced more than 1,700 AI-related bills this year. Some Democrats have called for government ownership of major AI companies and bans on new AI supercomputers. The study's authors argue that Washington faces a clear choice: it can tax, regulate, ration, or nationalize AI, reserving its benefits for the largest companies and risking ceding leadership to China, or it can follow the approach that succeeded during the internet revolution and foster innovation.
The authors emphasize that while some workers will need help adjusting and retraining, the overall effect of AI will be to make Americans more capable, not replace them. They predict that productivity improvements will raise the quality of life, especially for lower-income households, who could soon enjoy a standard of living that rivals the richest royalty of past eras. The idea of a four-day workweek, once a futuristic dream, becomes plausible as AI takes over mundane and repetitive tasks.
Prisons and reform schools are already using AI to train young people for jobs that are waiting for them, the study notes. The challenge, the authors conclude, is not to slow down AI but to ensure that Americans put it to work, rather than letting regulation hinder progress and hand the advantage to competitors abroad.
