In 1945, elevator operators in New York went on strike, and the city seized up. Nobody would ride an automatic elevator. The technology had existed for years. People just didn't trust a box that moved on its own.
Within a couple of decades, the operator was gone and nobody thought twice. We'd learned to trust the box.
What's the pattern?
We've been handing tasks to machines and learning to trust them for two centuries. The clock. The thermostat. The autopilot, which flies most of every commercial flight you take. Each time, the same arc: refusal, then reluctance, then forgetting there was ever a person in the loop.
That arc is fine for tasks. An elevator has one job.
Where it goes wrong is judgment. In my next book I write about how the mind gets hijacked: it trusts what's confident, consistent, and familiar. AI is all three. It never hesitates, never has a bad week, and never says "I'm not sure" unless it's told to. Which makes it feel more trustworthy than a person, on exactly the questions where it's less.
A colleague who says "I don't know, let me think" is showing you judgment. A model that answers instantly is showing you probability. We're wired to prefer the second. That's the trap.
What should leaders do now?
So, three things.
Sort your decisions into tasks and calls. The task, the machine can own. The call, a human owns, even if the machine advises.
Reward hesitation. The person who says "something feels off" is doing the thing the machine can't. If your culture punishes that as slow, you'll get a company that trusts the confident box every time.
And keep practicing trust in each other. It's a skill. It atrophies. A team that only checks with the machine forgets how to check with the person next to them.
We learned to trust the elevator. Good. Let's not learn to trust it for where we should be going.
What did you last take from a machine without a second thought, and would you have taken it from a colleague?
Related: Is perception the new battlefield?
More like this on LinkedIn. Follow @samradofficial
