Artificial intelligence has revived an old question: are machines coming for our jobs? In this week’s Grumpy Economist Weekly Rant, John Cochrane places AI in historical perspective, arguing that today’s anxieties echo past reactions to transformative innovations—from the printing press to tractors to the internet. Each wave of technological change eliminated specific tasks, yet expanded productivity, incomes, and new forms of employment.
Cochrane explains that AI is more likely to replace tasks than entire jobs, freeing workers from routine activities and enabling higher-value output. Like previous general-purpose technologies, AI will follow an S-shaped curve, requiring time for firms to adapt and reorganize. Rather than signaling economic decline, AI represents a new productivity-enhancing breakthrough—particularly in sectors such as health—at a moment when growth has slowed, and demographic headwinds are rising. The long-run economic lesson: innovation has historically made societies richer, not displaced.
Transcript
Hi, I’m John Cochrane, Senior Fellow here at the Hoover Institution, and welcome to The Grumpy Economist Weekly Rant. Today, we’re going to rant about artificial intelligence, economics, and jobs.
Here comes AI. We’re all going to be unemployed. We need a universal basic income. Now we’ll be sitting in the streets with nothing.
Relax. We’ve seen this play hundreds of times before: new innovations that come in and make us all profoundly better off than our ancestors were. Remember when Gutenberg invented the printing press? Oh no — the monks will all be out of a job. The steam engine, the mechanical loom, railroads, steamships, tractors, automobiles, trucks, airplanes, radio, TV, air conditioning, computers, word processors, copy machines, the PC, the internet. This is just one more, like all the others.
Are we really going to lose all our jobs? Yeah, AI will do many things humans can do, just like tractors did many things that humans could do, and we don’t see the 70% of Americans who used to work on the farm sitting in the gutter unemployed.
AI, in particular, is likely to replace tasks, not jobs. I can only think how much nicer my morning would have been if I had AI to sort through the spam in my email, to make that complicated, uh, plane reservation that I’ve been trying to make for a long trip, to get rid of the drudge work before I could get down to thinking hard about the Grumpy Economist Weekly Rant. That’s likely to be true everywhere.
Remember when ATM machines came in? Are all the bank tellers unemployed? No. Employment expanded in banking because now the banks were able to mechanize the, uh, boring tasks and move those people onto higher-value things. When things become cheaper, output expands. The business can expand. You can do more things with the same people.
In fact, my whole professional life, I’ve been worried about the growth slowdown. Growth in the US used to be 4% a year. Now it’s 2% a year, and that’s better than everywhere else, like Europe, where it’s zero.
There’s been a big debate raging among economists: are we out of ideas, or can we just not get the permits? I was kind of in the permits camp, but now that the debate is over, we have a great new idea, one more on this long step of productivity-enhancing innovations.
But hold on, like all past times, it’s going to take a while. It won’t be this spring. All of these technologies have followed an S-shaped. It took a few decades from inventing the automobile until everybody got cars and businesses adapted to it. Similarly, it’ll take a while for businesses to adapt to AI.
The internet was invented in 1990. You didn’t really see it improving things until the mid-2000s and 2010s, and it’s still hard to get the permits.
There’s plenty left to do in the American economy. So AI is great. It’s one more great innovation, like all the others that have made us so spectacularly well off.
In particular, I look forward to AI helping health. Once you’ve got a bunch of stuff, what is there but your health? And AI seems to be really poised to help with people’s health.
In fact, with declining birth rates, maybe artificial people will be useful for the economy, since we don’t seem to be making real people anymore.
Relax — it’s going be a great ride.
Thanks for listening. And if you like this, remember to click to subscribe.
Read more from John Cochrane on AI:
John Cochrane, “AI, Society, and Democracy: Just Relax,” The Digitalist Papers.
John H. Cochrane is the Rose-Marie and Jack Anderson Senior Fellow of the Hoover Institution at Stanford University. An economist specializing in financial economics and macroeconomics, he is the author of The Fiscal Theory of the Price Level. He also authors a popular Substack called The Grumpy Economist.
Powered by