The Economist, the 1843-founded house organ of free-trade liberalism, recently ran a piece calling Nobel laureate Daron Acemoglu “oddly unconvincing.” Alberto Garzón argues the real offense is political: Acemoglu’s work on technology and AI proposes putting the working class back at the center of economic policy — and that is heresy to the guardians of Big Tech’s untrammeled power.
The automation debate is two centuries old:
- Ricardo’s third edition of Principles (1821) broke with the cheerful orthodoxy that machines always create compensating jobs — machinery could be “very injurious to the interests of the class of labourers.”
- During the Industrial Revolution, handloom weavers’ real wages fell at least 25 percent; productivity rose, wages did not, and inequality exploded. Prosperity was not shared.
- Acemoglu’s Power and Progress (2023, with Simon Johnson): technology is not automatically good — whether gains are shared depends on institutions, power, and the direction society forces innovation.
- What fixed it in the nineteenth century was democratization, unions, and worker-protection law — counterpower that redirected technical change.
- His latest paper with David Autor and Simon Johnson is literally titled “Building Pro-Worker Artificial Intelligence.”
What makes Acemoglu dangerous: he reaches heterodox conclusions from inside the neoclassical mainstream, so he cannot be dismissed as a crank.
Garzón’s caveat is honest — Acemoglu’s “directing technology” pitch is vague on the how, and the only real answer is planning, the state, and social protection. That is where the next fight is. More shared prosperity means exactly the interventions the AI promoters do not want to hear about.