Concern about artificial intelligence and the labor market is an interesting kind of horseshoe. On the one hand, all indications are that normal people have a lot of anxiety about the way this technology might impact their jobs and employment prospects. On the other hand, the leading figures in the A.I. industry all believe their work is going to lead to massive short-term labor-market disruption and, in the longer term, the total displacement of human workers.
But between those peaks of anxiety is a middle ground of economists and journalists saying this is overhyped.
Economists know that Luddites have worried about automation leading to permanent disemployment for hundreds of years, underrating the benefits of productivity improvements. And cynical journalists think A.I. companies are just talking their own books when they speak of sweeping technological transformation.
I mentioned offhand on an earlier episode that I am largely on the combined side of the savvy industry insiders and the ignorant masses — I think economists are peddling a misguided midwit take that overindexes on 250 years of Industrial Revolution relative to the full sweep of history. Timothy Lee of the great A.I.-focused Substack Understanding AI heard me offer this opinion and suggested coming on the show to debate it.
My Co-Host Jerusalem Demsas is largely on Tim’s side. They think A.I. will change labor markets but in a broadly normal way — that it will displace jobs but do so at a gradual pace that is similar to how work and the economy normally transform over time. They call attention to the fact that most jobs consist of a big bundle of tasks, so as some tasks get automated, people will shift the balance of their work rather than the job evaporating altogether. They’re also both believers in the value of the “human touch” — the idea that, in many contexts, people willingly pay a premium for the benefits of human interaction.
I think this is all a little too pessimistic about the pace of technological change and too rosy about the virtues of actual human beings. The human touch can be great, but lots of people are mediocre and half of them are below average. You by definition can’t replace a Lionel Messi or a Lamine Yamal with a robot, but the vast majority of people aren’t superstars at anything, much less in a job where using automation would break the rules.
Last but not least, I make the case that I am not a “doomer.” What I am actually expressing is extreme optimism about the ability of technology to make human toil unnecessary — I just think it’s important that we take this scenario seriously so it is managed for broad human benefit.
Jerusalem, Tim, and I get into all of it on this week’s episode. Watch or listen wherever you get your podcasts.
Paid subscribers can watch the ad-free embedded video at the top of this post and read the full transcript below.
New episodes post every Thursday.
Listen to this episode with a 7-day free trial
Subscribe to Slow Boring to listen to this post and get 7 days of free access to the full post archives.
