In an essay published by The Brookings Institution through its Hamilton Project, three economists at MIT, Daron Acemoglu, David Autor, and Simon Johnson, make the case that AI does not have to be a force that displaces workers. Used well, they argue, it could be something far more hopeful: a technology that makes human skills more valuable, opens up entirely new kinds of work, and allows workers to genuinely thrive. They call this vision “pro-worker AI.”
Greg Rosalsky, writer and reporter for NPR’s Planet Money, picks up the thread in his March 2026 economics paper roundup, providing context for why this conversation is landing with particular force right now.
Below are key highlights from both pieces:
- The anxiety driving the conversation: Rosalsky opens by noting that surveys suggest the majority of American workers are experiencing what he describes as “AI anxiety,” with tech leaders citing AI as the reason for high-profile layoffs even as analysts question whether AI is the real cause. The Brookings essay arrives as a direct response to that climate.
- What “pro-worker AI” actually means: Acemoglu, Autor, and Johnson define pro-worker technologies as those that make human skills and expertise more valuable, and identify five categories of technological change, ranging from labor-augmenting to automating to new task-creating, with only the last being unambiguously good for workers and only automation being unambiguously bad.
- The de-skilling risk: Rosalsky illustrates one of the essay’s core concerns through the example of GPS and smartphone maps, which eroded the value of skilled taxi drivers’ knowledge of their cities. As ride-hailing apps lowered barriers to entry, a larger pool of less skilled drivers entered the market and wages fell. That is the kind of outcome the authors want to avoid repeating with AI.
- A more hopeful possibility: The economists write that AI’s potential to serve as a collaborator, extending human judgment, enabling new tasks, and accelerating skill acquisition, is equally transformative and currently underexploited. They envision a future where human workers collaborate with AI rather than compete against it.
- Two obstacles standing in the way: Acemoglu, Autor, and Johnson point to misaligned incentives, where leading firms see greater returns in automating expertise than in creating new tasks for workers, and to what they call pro-automation ideology, with the tech industry’s fixation on Artificial General Intelligence making it less likely that firms will invest in genuinely pro-worker tools.
- Nine policy levers for change: The authors outline nine recommendations, including focusing pro-worker AI development on health care and education, using DARPA-style competitive prizes to direct research, rebalancing the tax code so that investing in workers is not penalized relative to investing in capital, strengthening antitrust enforcement, establishing mechanisms for worker voice, and building legal protections that prevent human expertise from being appropriated by machines without consent.
Acemoglu and Johnson shared the 2024 Nobel Prize in Economics, and Autor is widely regarded as one of the leading labor economists of his generation. Their core message, as Rosalsky notes it, is that the current trajectory of AI development is not inevitable. It is in all our collective interest to ensure that the AI-intensive future works well for workers, and the policy tools to make that happen already exist.
Acknowledgement: The summary section of this post was generated using AI writing tools as part of my AI Explorations process. For more information, please refer to the original articles listed below.