“We’re not trying to schedule our workers more efficiently. We’re trying to replace them altogether.”
— An executive of a fast-food company
Andrew Yang, a 2020 presidential candidate, reports that quote about automation in his book The War on Normal People.
And though I think the executive is doing harm, I understand his reasoning.
Last week, I was paying for my lunch at a fast-food restaurant. The bill was $5.44. Paying with a $10 bill would have given me change of four one-dollar bills and 56 cents. I didn’t want the one-dollar bills, so I gave $11 to the woman at the register. Easy-peasy: $11 minus $5.44 was a five dollar bill and 56 cents.
She looked at the $11, puzzled. She tried to give the one-dollar bill back to me. “It’s $5.44,” she said. “Trust me,” I replied. I explained that $11 minus $5.44 was $5.56.
So she put the $11 into the cash drawer, keyed in $10, and the computer told her that my change was $4.56. That’s what she tried to give me. I explained again: $11.00 minus $5.44 was $5.56. After one more try, she finally understood and gave me the correct change.
I’ve had that experience several times at retail establishments. The people at the cash registers depend on the cash registers to tell them how much change to give. If anything happens that is slightly out of the ordinary, they don’t know what to do. They can’t do simple arithmetic.
Those people are created in the image of God. Their lives matter. Their happiness matters. But they are not smart enough to make change, even with the assistance of computerized cash registers. A humane society must somehow find a place for them.
It’s not just them. Even if employees can make change, it’s cheaper to replace them with automated kiosks. The kiosks can handle most transactions, don’t need benefits, never get sick, never complain, and never need vacations. The few remaining human workers will handle any unusual situations that the kiosks can’t.
The same logic increasingly applies to white-collar occupations. Artificial intelligence programs are getting good enough to handle routine legal tasks, interpret medical tests, and to do other “knowledge worker” jobs. Companies will replace human employees by automation if:
- It’s cheaper, and
- The quality is at least as good or (often) companies don’t care much about it, and
- There are no legal, institutional, economic, or social constraints that discourage companies from replacing people with automation.
Some of it is going to happen no matter what we do. Yang seems to take it as a fait accompli, offering a “universal basic income” as a remedy. He thinks that it’s a mistake to connect work with self-respect and personal dignity. I’m uneasy about those ideas. Yes, they’re one way to address the problem: but are they the best way? Stay tuned.
Check out my new book Why Sane People Believe Crazy Things: How Belief Can Help or Hurt Social Peace. Kirkus Reviews called it an “impressively nuanced analysis.”
The Thousand-Year View 
I get your point about simple arithmetic at cash registers. But I wonder what percentage of those cashiers could do it if they were shown.
I had a job at a Dairy Queen in 1985, when the newest cash registers calculated change. The store owner would have none of it and required us to manually count change. We were even to count it out loud when giving it back to the customer: coins into the hand as we announce the dollar this brought them to, then the running total for every bill we pressed into their hand.
I had no idea what this was when I was first asked to do it. I guess I’d never paid attention when, as a customer elsewhere, it was done for me. But when the owner showed me that one time I got it and did it.
Some people won’t get it the first time they’re showed, or the second, or the fifth. But I don’t think we know what percentage of cashiers *will* get it because it’s not shown to anyone anymore.
LikeLike
I had pretty much the same experience as you did, and there’s no doubt that training and practice are a part of it. That said, our economy has changed. A hundred years ago, there were plenty of jobs whose main demand was essentially to “take those bags off the cart and put them in the barn.” Those jobs still exist, but they are few and far between. People who would have easily held jobs in the 1900s are now having trouble.
And the trend is now reaching more complex jobs. One reason I got out of software engineering is that I could see what was coming: AI will do a first draft, and a small number of human software engineers will fix the problems. That’s essentially what happens now in companies where low-skill offshore developers do a first draft and a few high-skill Americans fix it; the main difference is substituting AI for offshore developers.
LikeLike
I’m trying to stay ahead of that trend by staying on the management side and moving from testing to engineering. Here’s hoping it gives me enough time to get through to retirement.
LikeLike
I learned to count change much as Jim Grey did. I think this is a skill that almost anyone can master. The problem has been low-level automation that made the skill unnecessary 99% of the time. The skill is unneeded and goes untaught now, so young workers don’t know it.
This is bound to happen higher and higher up the food chain. I see it in insurance claims. Young adjuters never leave their desks and therefore never visit an accident scene or meet a witness in person. The quality of the files I see is terrible compared to thirty years ago. But the insurance company can process so many more claims with fewer people now and most of them probably don’t need the full old-fashioned treatment. The cost is that these young adjusters don’t learn skills that older ones took for granted.
My fear is that if through some technical catastrophe we need to dust off an older skill set it may not be there to dust off. I acknowledge that this may not be a realistic fear.
LikeLike
I agree on all points. One other thing is that the cognitive demands of our jobs have been increasing for a long time. In the 19th century, there were still plenty of jobs on the level of “take those bags of grain off the wagon and put them in the barn.” However, the increase has been accelerating, so that more and more people are in danger of getting “left behind.” I really dislike the idea of just giving people a check, because that amounts to saying, “you have nothing to offer, but here’s some money so you can stay alive.” That’s part of the problem with our current welfare state (in addition to the moral hazard it creates for illegal immigration).
LikeLike
Agreed. Another problem, to which I think Jim also might testify, is that companies don’t care all that much about quality. Jim got out of the software documentation profession because he saw that most technology companies don’t want to bother with it at all, and if they *must* bother with it, they don’t care if it’s any good.
LikeLike