Happy Monday
One idea, one quote, one question. 90 seconds.
The Idea
In the 1970s, banks started installing a machine built to do a bank teller’s job. By the mid-90s there were hundreds of thousands of ATMs in America, and everyone assumed the teller was finished. The bank managers buying the machines assumed it too.
The opposite happened. Teller employment went up. Around 485,000 tellers in 1985, about 527,000 by 2002. The machine that automated the job created more of the jobs.
An economist named James Bessen figured out why. ATMs cut the number of tellers a branch needed from about 21 to 13, which made branches cheaper to run, so banks opened 43% more of them. And the work itself got better. The machine took the cash counting, the most repetitive part of the day, and the tellers who moved to what was left became relationship people. Advising customers, solving the weird problems, selling. Bessen said the ATM basically turned the teller into a marketing person.
This story has been stuck in my head since a conversation on the podcast last week. Ruben Hassid teaches millions of people how to use AI, he built the fastest-growing newsletter (How to AI) on Substack doing it, and halfway through the conversation he said something that took me right back to the tellers. AI is not making all of us better, he said, it’s making mediocre much better. He doesn’t pay his lawyer to write contracts anymore, AI does that in seconds. He pays his lawyer to find the gaps, the stuff that isn’t inside the AI. The contract is the cash counting. The judgment is everything else. Same story, fifty years later. The tellers who did fine were the ones who figured out early which part of the job the machine was going to take, and got good at the other part before they had to.
Quote
“The illiterate of the 21st century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn.” — Alvin Toffler
He wrote that in 1970, the same decade ATMs started showing up in banks. The tellers who survived were the ones who unlearned cash counting and relearned relationship building. Fifty years later, same test, different machine.
Question
What’s the cash counting in your job, and what would you do with your week if a machine took it off your plate tomorrow?
Listen
Ruben Hassid: The #1 AI Teacher (Billions of Views) Explains Who Survives and Who Doesn’t. The lawyer example is the part I built this newsletter around, but the conversation goes a lot further than that. Ruben has this idea of “clickers,” the people who open a new tool and just start hitting buttons and trying stuff, versus the people who sit there staring at the screen wondering if it’s safe to touch anything. He says that’s the whole gap right now. If you’ve been meaning to figure this stuff out, this is the episode to start with.
Read
Learning by Doing by James Bessen. This is the book the teller story comes from. Bessen studied every major technology wave back to the power loom and found that the money never showed up when the machine was invented. It showed up years later, once ordinary people learned to use it, and it went disproportionately to the ones who started learning early. He wrote it in 2015 and it reads like it was written for right now.
If you love this content (please share it), but also check out my Podcast, and connect with me on YouTube / Twitter.
— Scott