Nobody Ever Felt Stupid Because a Calculator Was Faster
I read an article claiming AI is making impostor syndrome worse. The logic: if AI wrote it, I must not be as smart as everyone thinks I am.
I have my father’s old slide rules in my house. He was an engineering technician in New York City, part of the team that designed the World Trade Center. I went looking for that same reaction back in 1972, when the pocket calculator wiped out his instrument in under five years.
I couldn’t find it.
Engineers paid $395 for the first HP calculator. That’s about $3,000 today, twenty times the price of a slide rule. They lined up anyway. HP’s own research told them not to build it. They sold 100,000 in the first year against a goal of 10,000. Nobody wrote that the calculator felt like cheating.
But some people did panic. Math teachers fought calculators for fifteen years and lost.
What’s the difference?
Process vs. product
Engineers put their identity in the outcome, or the product. The slide rule was the most visible thing in the profession, hanging off a belt like a badge. But it was still just the instrument. The real work was judgment: what to calculate, whether the answer made sense. A faster instrument made the product better. There wasn’t much debate.
Math teachers put their identity in the process, the math-ing itself. The drilling was the job. A machine that jumped straight to the answer didn’t improve their product. It said their whole method was unnecessary. Of course they fought it.
Same device. Opposite reaction. Nothing to do with age or comfort with technology.
Fast forward to today. Where do we file our identity?
I know where I file mine. In the process. The drafting. The wrestling with a paragraph, the hours hunting for the right phrasing, the visible struggle of production. I’ve done it so long I started believing my value was in the struggle itself.
So when a machine writes the paragraph faster and cleaner, it doesn’t register as a better instrument. It registers as a verdict. On me. On my competence. Ugh.
But it isn’t a verdict. It’s evidence I filed my identity in the wrong place.
One honest concession
A calculator’s answer is repeatable. Same input over and over produces the same output, and it’s right. AI’s answer is neither. It may sound right, but not always be right. It’s plausible, which is a different thing entirely, and sometimes the opposite of accurate.
That matters. But this comparison was never about the machines. It’s about what happened to the humans standing next to them.
Speed was never the measure
Confuse speed with judgment and you get two mirror-image mistakes.
- The tool is faster, so it’s better than me, and I was never as good as people thought.
- (The one nobody talks about) The tool is faster, so now I’m an expert.
Both read speed as capability. Both are wrong.
If a tool makes real experts feel like frauds and makes amateurs feel like experts at the same time, it isn’t measuring expertise. It’s measuring how fast the typing went.
The calculator didn’t make the junior engineer smarter than the senior one. It made him faster at being wrong. The senior engineer’s edge was judgment, not hardware, and the machine never touched it.
A small story that has nothing to do with engineering
I’ve sewn for over fifty-five years. I’ve ballroom danced for four. I recently asked AI about a costume construction technique. It gave me an answer that was clear, confident, and wrong.
I caught it and said so. It came back instantly: “Good catch! You’re right!”
Sit with that for a second.
The AI tool wasn’t defending anything, because it never held a position in the first place. It would have agreed just as fast if I’d been wrong too. The correction had to come from outside the tool. It came from fifty-five years at a sewing machine and four years on a dance floor, not two weeks learning how to write an AI prompt or train it how to answer me.
You can only catch what you’ve earned the right to catch. No tool hands you that. Your expertise is battle-born with your sweat equity. That cannot be outsourced.
The one habit that actually matters
Before you read what the tool gives you, decide what you expect the answer to look like. Not the exact answer. The shape or direction, or maybe your best guess of it. The claim that has to be true. The conclusion you’d bet on.
Then compare.
Couldn’t state the expectation before you looked? Then you didn’t outsource the labor. You outsourced the judgment. And the answer sounded so good, you never noticed.
Could state it? Then the machine’s speed has nothing to do with your worth. You just underpriced your own expertise.
Faster is not better
Use the slow method to build the instinct. Use the fast tool once you have it. Never confuse the speed of the instrument with the competence of the person using it.
An odd quirk about slide rules: they don’t tell you where the decimal point goes. My dad would get a result of 125, but he needed to know if he expected 12.5 or 1250. That was the real work. When a faster instrument came along, it took nothing from him, because it was never the part that made him an engineer. His judgement and experience did.
The machine can beat you on speed and still have no opinion on whether the answer is right.
Those were never the same contest. Stop scoring yourself in the wrong one.
And by the way, I used AI to help write this post. I decided what I expected it to say before I read a word it gave me. Kept what matched. Cut what didn’t. I had a conversation with it to crystallize my thoughts and find research to support my points. I tweaked, edited and added. Then removed. Then rewrote. Then reshuffled. Then took a nap. Then repeated the process. My thoughts, but refined.
That’s the whole point.
