Nearly three-quarters of people with “white collar” jobs have four year college degrees or higher. Keep that in mind as you read this article by Chip Cutter and Haley Zimmerman at the Wall Street Journal:
CEOs are no longer dodging the question of whether AI takes jobs. Now they are giving predictions of how deep those cuts could go.
“Artificial intelligence is going to replace literally half of all white-collar workers in the U.S.,” Ford Motor Chief Executive Jim Farley said in an interview last week with author Walter Isaacson at the Aspen Ideas Festival. “AI will leave a lot of white-collar people behind.”
At JPMorgan Chase Marianne Lake, CEO of the bank’s massive consumer and community business, told investors in May that she could see its operations head count falling by 10% in the coming years as the company uses new AI tools.
The comments echo recent job warnings from executives at Amazon Anthropic and other companies.
Amazon CEO Andy Jassy wrote in a note to employees in June that he expected the company’s overall corporate workforce to be smaller in the coming years because of the “once-in-a-lifetime” AI technology.
“We will need fewer people doing some of the jobs that are being done today, and more people doing other types of jobs,” Jassy said.
Anthropic CEO Dario Amodei said in May that half of all entry-level jobs could disappear in one to five years, resulting in U.S. unemployment of 10% to 20%, according to an interview with Axios. He urged company executives and government officials to stop “sugarcoating” the situation.
Slight more than thirty years ago policy makers of both parties whole-heartedly embraced the notion that higher education was the key to a bright future. That has guided policy ever since. I have always thought it was short-sighted for relatively simple mathematical, psychological, and pragmatic reasons.
Those who benefit from a college education have IQs between one and two standard deviations above normal (less than 15% of the population). Individuals with IQs much below that struggle in college. Many will drop out having incurred considerable debt. Individuals with IQs much above that don’t actually need college. They’ll learn well enough without it.
In the olden days when I was in school about 10% of those of age went on to graduate from college. They were mostly pre-professional or pre-managerial. Nowadays nearly 40% of young people graduate from college and nearly 60% of young people have at least some college, degree or not. Many of them are doing jobs that would previously have been done without college educations.
No “white collar” jobs for them will not make them happy or erase their educational debts.
I suspect that “half” is understating the problem. I suspect that every “white collar” job that is not legally protected or that lacks the power to protect itself is at risk whether it makes sense or not. I doubt that many Fortune 500 CEOs will lose their jobs due to AI (even if AI could do those jobs better and cheaper).
That in turn will create a new problem. Why go to college? I daresay that most of those going to college do it as job preparation. What if college doesn’t prepare you for any jobs? Indeed, I suspect that problem is already here. Almost 90% of college students are already using AI. How much sense does it actually make to pay upwards of $50,000 per year to have AI do your work for you?
Oxford and the University of Bologna were founded in the 11th century. Cambridge and the Sorbonne were founded in the 13th century. In my view artificial intelligence presents an existential challenge to higher education greater than any they’ve faced in the more than a millennium during which higher education has existed.
I have seen any number of AI pieces on various subjects. I have been struck by the superficiality. A notch above Siri.
My impression so far is that it’s the mother of all data survey and statistical analysis tools. I saw operations cited in your post. I can see that. But the Luddite overreaction seems unwarranted.
I’ve thought that functions requiring creativity, hands on or face to face, and especially judgment, will remain. (I’m sure there are more) That’s a lot.
As for college vs the trades, that’s a well worn subject, and follows under hands on. But it makes the point of predicting, which is hard, especially about the future…..
I think AI replaces a lot of the paper pushing that a lot of college grads have to do. This is a good thing. I hate the paper pushing part of my job.
As you said, college students are of above average intelligence. Many will start new businesses. This is a good thing.
The total amount of time required to do work and create things will go down. This will free up people’s time to do better things. Unfortunately, most people will waste this new time watching AI cat videos on the internet.
Drew:
I don’t think that “Luddite” quite captures my views. “Fatalistic” would probably be better.
I don’t think there’s anything that can be done to mitigate the adverse effects of rashly-applied AI. And rest assured there will be adverse effects and it will be rashly applied, especially initially. I think that Charlie Musick is right. Properly applied AI will take over a lot of drudgery. Unfortunately, that won’t be the reason that big company managers are interested in it. They will inevitably want to save costs, i.e. reduce head count. My secondary prediction is that AI won’t work out that way but that’s what they’ll be seeking.
Meh. I think it may eventually replace a lot of jobs but it will take a while. AI can write papers for college kids but I don’t think that many jobs resemble writing a paper. It will mostly be a productivity aid at first, which will garner some job loss but lead to new ones also.
Steve
Dave – Sloppy writing. I didn’t mean you. Rather, a generalized Luddite reaction to AI.
When AI can write and/or explain Finnegans Wake, I will be impressed. Otherwise, it is a glorified spell/grammar checker.
Google search AI summary does not even attempt it. Since no human can fully understand it, AI has nothing to plagiarize, and it is not remotely grammatically correct.
Musings.
It seems there is a muted reaction by commenters here. I thought about Steve. I understand AI can survey a bazillion treatment protocols…..but can it replace ER judgment and experience on sight, and right now? Not buying it.
I could go through my profession, but I don’t want to bore. It’s not formulaic.
I continue to think – and it’s not to minimize – that operations is in the crosshairs. Real decision making not so much.
I ran across this article, Superintelligence Will Never Arrive.