Economic policy vets see risks in AI overreliance

Jared Bernstein, the CEA chair under President Biden, noted that policy work is "highly bespoke as you're explaining the data flow (and much else) to very busy people who don't want only market-shop analysis," highlighting the human element in economic advising.

This sentiment comes as Harvard economist Jason Furman suggested to Axios that he would utilize AI tools over staff economists for much of the day-to-day analytical tasks if he returned to his previous role as chairman of the White House Council of Economic Advisers.

This proposal generated discussion online, particularly within the community known as Econ Twitter, as many veterans of economic policy express concern over policymakers becoming too reliant on artificial intelligence.

Bernstein further elaborated that while market reactions to figures like CPI, jobs, and GDP are relevant, policymakers also seek to understand how reports fit into the administration's broader narrative.

He stressed that within the White House, "the human skill of reading-the-room is paramount," adding that his sense is AI struggles with this and likely always will, a conclusion informed by listening to its creators.

Claudia Sahm, a former Federal Reserve economist, raised a concern echoing those in consulting and investment banking: if junior roles disappear, where will future senior leaders originate? She pointed out that Furman's staff previously founded the Yale Budget Lab and served in later administrations.

Anna Wong, Bloomberg chief U.S. economist, countered that the ability for a senior person to operate independently is empowering, suggesting that AI itself, through accumulated training, could spawn ideas, moving beyond the need to replace aging experts.

However, Casey Mulligan, who previously served in President Trump's first-term CEA, argues the impact on labor isn't straightforward. He noted that building AI tools in two federal offices required more economists, not fewer, due to increased demand.

This phenomenon, where labor-saving tools boost demand for labor, is termed Jevons Paradox. Stanford economist Neale Mahoney concluded that the responsibility rests with older generations to foster organizations where people and AI work together, while preparing younger workers for advanced AI integration.