Josh Weiner
Daron Acemoglu, the MIT economist who shared the 2024 Nobel Prize in economics, used a recent Bloomberg Weekend interview to deliver one of his starkest warnings yet: liberal democracy, he argues, is at the cusp of losing its footing, undone by a failure to adapt to the postindustrial economy and a retreat from the promises that made it work in the first place.
Acemoglu’s diagnosis centers on a specific failure of the political center-left. In his latest book, What Happened to Liberal Democracy: Remaking a Politics of Shared Prosperity, he argues that liberals failed the working class by not confronting the economic dislocation that automation and globalization produced, leaving an opening that populist politics on both ends of the spectrum has since exploited.
He extends that argument directly into the AI era, and his critique there is pointed. Acemoglu has said current AI investment is predominantly focused on replacing human tasks rather than augmenting workers’ capabilities, a design choice he frames as a continuation of the same pattern that hollowed out middle-class employment during earlier waves of automation, rather than a technology being deployed to make workers more productive and valuable.
His language about the pace and oversight of that deployment is blunt: this is a massive social experiment being run without any guardrails, he has said, a characterization that puts AI development in the same category as the kind of unregulated technological and economic shifts he blames for democracy's current fragility.
Acemoglu also takes aim at social media algorithms as part of the same underlying crisis, arguing that AI and liberal democracy share a common failure mode: a tendency to escalate everything, where systems optimized for engagement or efficiency push outcomes toward extremes rather than toward the kind of broad, shared prosperity that stabilized democratic politics in the postwar era.
The through-line across his AI commentary and his democracy argument is consistency rather than two separate critiques. Both, in his telling, stem from the same mistake: treating technological and economic disruption as something markets and algorithms will sort out on their own, rather than as a political problem requiring deliberate guardrails and a shared-prosperity agenda to prevent the kind of backlash currently reshaping democratic politics across multiple countries.
For a Nobel laureate whose prize-winning work examined why some nations succeed economically while others fail, the warning carries particular weight: Acemoglu isn't arguing that AI or algorithmic social media will inevitably destroy democratic institutions, but that the current trajectory, unregulated, task-replacing, and engagement-maximizing, is running the same experiment that already pushed liberal democracy to the edge once before.
