MIT economist David Autor, known for the 'China shock,' is warning that AI may help workers produce better output without building the deeper judgment that drives long-term earning power.
His latest research looked at patent lawyers in a three-month experiment. The study enrolled 133 lawyers at 11 U.S. intellectual-property firms with ongoing patent-drafting relationships with Google. Researchers randomly assigned access to a custom AI patent-drafting assistant, a then-unreleased Google Labs tool. Two-thirds of the lawyers got access. The rest waited until the study ended.
After 90 days, AI access raised drafting scores by 0.38 standard deviations, according to the paper. In a Google blog post, Autor and co-author Tanya Rodchenko described that as an 11-percentile-point gain relative to control scores. The paper says the improvement came from less weak work, not more excellent work.
But the harder test came when the tool was removed. Researchers asked lawyers to mark up a hypothetical patent that contained substantive and stylistic errors. On that independent judgment task, only the experienced lawyers showed an average improvement. The juniors did not.
Autor said in written responses to Fortune that the results 'challenge the idea of AI being an automatic skill equalizer.' He said the data suggest AI is 'a performance equalizer, but a skill-disequalizer,' because only practitioners with foundational mental models were able to level up their underlying skill sets.
The study sorted lawyers by experience, not age. 'Senior' meant seven or more years in practice. It did not study Gen Z as a group. Still, the findings bear most directly on early-career workers, the people most exposed to a labor market where speed is now cheap and judgment remains scarce.
Autor pushed back on the notion that AI is just another China shock. On the Possible podcast hosted by LinkedIn co-founder Reid Hoffman, he said AI 'will not be, in any sense, a repeat of the China trade shock' because that was a pure negative competitive shock to U.S. firms, while AI has a different texture in boosting productivity.
He also said the danger is not necessarily mass job loss. 'The danger that we may face with AI is not that it eliminates large numbers of jobs in net, but that it eliminates some areas of specialty while creating new ones simultaneously,' he said. The workers losing those old careers, he added, are generally not the ones who can take advantage of the new roles.
The market has already voted on AI as a productivity tool. What this research suggests is more sobering: capital can buy software, but it cannot guarantee wisdom. That matters in law, and it matters across the broader economy, where the most valuable asset is still judgment built over time.
For the taxpayer and for free enterprise, the principle is clear. Technology that raises output is welcome. But a system that accelerates performance while hollowing out skill leaves the next generation less prepared to compete when the software is turned off.
