The author argues the fear that AI will wipe out work misreads history: unemployment has stayed flat for decades despite automation, and no broad job disruption shows up in the data yet.
Seventy-one percent of American adults expect AI to mean fewer jobs over the next twenty years. The CEO of Anthropic has even predicted unemployment could hit 10 to 20 percent within a few years.
But follow that logic and unemployment should have been climbing steadily since the Industrial Revolution. In reality, the US unemployment rate has sat roughly where it was eighty years ago.
The author calls this the 'lump of labor' fallacy: the belief that the work to be done is a single list that only ever gets shorter.
Desires have no ceiling, so the list never shrinks
In If You Give a Mouse a Cookie, each wish the mouse gets fulfilled brings up the next one, and no list could be written in advance. The author thinks human desires work the same way.
A futurist a century ago might have predicted air traffic controllers, but hardly professional screamers. The author calls this restless pattern the hedonic treadmill: once you get something, you quickly want more.
The AI shock has not yet shown up in the data
The fear of AI rests on the premise that this time is different: unlike past technologies that pushed people into new niches, it will fill every crevice.
Economists' latest judgement is more cautious: AI may already be affecting junior white-collar hiring, but the broader impact has not yet appeared in the data.
Indian IT outsourcing firms have seen their share prices fall, yet their revenues and headcounts have held up. Radiologists, supposedly the first doctors to be replaced, now earn record salaries.
Humans keep believing in breaks
The author argues people prefer stories in which the future suddenly breaks from the past, because humans are gifted storytellers. Being able to imagine a flying pig makes it easy to mistake imagination for prediction.
In the end, nobody can calculate AI's future precisely. Don't rush to pick a side. But for two hundred years, people have always found new work after being replaced.
Why it matters
If AI really hits jobs, the first sign should show up in junior white-collar hiring numbers, not in unemployment headlines. Whoever tracks that figure will spot the change earlier than whoever watches layoff announcements.
Curated from high-quality sources, with concise summaries and key takeaways.
