Governments should prepare for multiple Artificial Intelligence (AI) adoption scenarios to avoid policy missteps as the technology evolves, deputy managing director at the IMF, Bo Li said, warning against both overreacting to hype and underestimating potential risks.
Speaking on the topic of AI and the Resilience Gap: Diffusion, Dependency, and the Policy Agenda, the deputy director said authorities need to build flexibility into economic planning to manage uncertainty around the pace and scale of AI diffusion.
“We think that’s very important. Not to lock your policy into one view about the speed of diffusion, about the capability of AI,” Li said.
Li outlined two broad scenarios developed in collaboration with Massachusetts Institute of Technology and other partners: a gradual and uneven adoption of AI, and a faster, more widespread rollout that could require significant upfront economic adjustments.
“One scenario is what we call gradual, uneven adoption of AI. Another is a faster, broader diffusion that require probably a larger front-loaded adjustment cost,” he said. The differing scenarios could have wide-ranging implications for labor markets, as well as fiscal and monetary policy, depending on how quickly businesses and governments integrate AI technologies.
“And under these two scenarios, their implications for policy, for labor market, for monetary policy, fiscal policy, can be very different,” Li said. He cautioned policymakers against anchoring decisions to a single outlook, noting that both baseline expectations and less likely but high-impact outcomes should be considered.
“On the one hand, you don’t want to chase the hype. But on the other hand, you don’t want to miss some of the important risks,” Li said. “Even though they may be tail for now, but it could prove materialized down the road.”
Scenario planning, he added, would help policymakers design strategies with what he described as “high option value,” allowing economies and institutions to remain resilient to potential technology shocks.
“So those kind of scenario planning can really help authorities to think about policies… so that our institutions, our economy can be resilient to this kind of technology shock,” Li said.
