In short

Phil Bell interviews Oxford's Carl Benedikt Frey on the political economy of technology. Frey argues progress is contingent, not inevitable: decentralised institutions suit exploration of new technologies while centralised coordination suits scaling mature ones, citing the Soviet computing failure. Europe caught up in post-war manufacturing but failed in digital, hampered by fragmented service markets the IMF likens to a 110% internal tariff. Whether AI raises wages depends on whether it automates or enables new tasks.

“Progress is not inevitable” could be the motto of this whole publication, and Carl Frey has spent a career showing why.

In this conversation he lays out a deceptively simple framework: new technologies need decentralised institutions to explore and discover them, but centralised coordination to scale them, and societies fail when they apply the wrong one at the wrong moment. He uses it to explain everything from why the USSR could build steel but not computers, to why Europe caught up on post-war manufacturing yet fell behind on digital.

For Europe specifically, his diagnosis is uncomfortable: the biggest barrier isn’t a lack of talent but self-imposed fragmentation of its own service markets. And on the question everyone asks, will AI lift or flatten wages?, his answer is that it’s a choice between automating and enabling, not a forecast.

Key takeaways

  • Different institutional settings suit different lifecycle stages: decentralisation for exploration, coordination for scaling.
  • Centralised systems (the USSR) excelled at static heavy industry but failed at dynamic computing.
  • Europe's digital lag stems from self-imposed service-market barriers, the IMF estimates them equivalent to a 110% internal tariff.
  • AI's wage impact hinges on automation (stagnation) vs enabling new tasks (growth).
  • Business dynamism is declining, and heavy regulation disproportionately burdens small firms and solo inventors.

Listen to the full episode

This is a summary. Listen to the full conversation on your platform of choice, or read the companion essay on Substack.

Read on Substack → Watch on YouTube Listen on Spotify

Frequently asked questions

Can Europe catch up on AI?
It faces structural barriers, especially fragmented service markets the IMF likens to a 110% internal tariff, which, unlike its post-war manufacturing success, have left it lagging in digital and AI.
Is technological progress inevitable?
No. Frey argues progress is contingent on institutions: decentralised systems enable exploration of new technologies while centralised coordination suits scaling mature ones.
Will AI raise or stagnate wages?
It depends on use. Automation that replaces labour tends to stagnate wages (the First Industrial Revolution), while enabling technologies that create new tasks drive wage growth (the Second).

People & ideas in this piece

Carl Benedikt FreyOxford Internet InstituteDaron AcemogluPascual RestrepoJoel MokyrAdam ToozeThe Technology TrapHow Progress EndsAutomation vs enabling technologyTechnology lifecycle

Topics: AI & Geopolitics , AI, Work & the Economy