I sent an AI into a tailspin recently.
Not intentionally. I just asked it to help me diagnose a technical problem, an image that wasn't showing up in a newsletter I was building. It read through the code, found something that looked suspicious, declared victory, and stopped looking. It was wrong. I pushed back. It found something else that looked suspicious, declared victory again, and stopped looking. Also wrong.
Third time, I just did the obvious thing, deleted the image and re-uploaded it, and it worked immediately.
When I pointed out what had happened, the AI said something that stopped me cold. It admitted it had done exactly what humans tend to do: read until it found a possible answer, assumed that was it, and never went further.
I laughed. Then I got a little uncomfortable. Because I do that too. We all do.
It has a name, of course. Confirmation bias. The tendency to stop searching for information once you've found something that supports what you already suspect. Psychologists have been writing about it since the 1960s. We've all nodded along to articles about it. We've all thought, yes, other people really need to work on that.
And then we've all gone right back to doing it.
Here's my personal version that drives me absolutely out of my mind: I send an email with three things in it. Three. Clearly written, numbered if I'm feeling generous. And someone reads the first one, handles it, and moves on with their day like they've just solved world hunger. The other two just float there. Unanswered. Existing in some administrative purgatory.
Is that laziness? Sometimes. But honestly, I think more often it's confirmation bias in disguise. They read until they found something actionable, their brain said got it, and that felt like enough. The rest of the email became invisible because they'd already decided they understood the situation.
Sound familiar? Good. Keep reading.
Now let's bring this home, because this isn't just about email management. This is about your funeral home. Your crematory. Your operation.
The client who hires a consultant.
I've watched this one from the front row, because I live on the consulting side of this industry. An owner brings someone in, gets the first recommendation, implements it, and mentally considers the engagement complete. The consultant had four more things to say. Critical things. But the owner found the first answer plausible, it confirmed what they already half-believed, and that felt like enough.
A month later they're wondering why things still aren't quite right. The answer was in the paragraphs they didn't read.
The operator in the training room.
I've seen it happen in certification courses too. Someone clicks into place early, gets that oh, I've got this feeling after module two, and from that point forward they're physically present and mentally somewhere else entirely. They leave with a certificate and half an education, fully convinced they have the whole thing.
Confidence is great. Premature confidence is a liability.
The crematory operator Performing Cremations.
Everything looks fine. The equipment is running. The cycles are completing. The numbers seem about right. And because the first thing checked confirmed that nothing was on fire, figuratively speaking, they moved on to something else.
But operational standards aren't a checkbox you stop at when things look okay. They're a complete read-through. Every time. Skimming a checklist until something confirms your assumption isn't due diligence. It's just a more official-looking version of guessing.
The owner reading industry reports.
Cremation rates are up. You knew that already. You read the headline, it confirmed what you believed, you closed the tab. But did you read far enough to catch the regional variance? The demographic shift moving faster in your market than the national average? The service expectation data quietly rewriting what families want from a cremation provider?
The answer that feels right after one paragraph might be completely wrong by paragraph four.
I'm not standing above any of this, by the way. I am fluent in confirmation bias. I've walked into consulting engagements already knowing what the problem was. I've looked at an operational report and seen what I expected to see. I've read enough of an email to feel informed.
The AI that helped me debug my newsletter was, in a strange way, a mirror. Watching it confidently present a wrong answer, not once but twice, because it found a plausible answer before it found the answer was genuinely humbling. That's me. That's you. That's every smart, experienced person who has ever let competence become a shortcut.
The cure isn't complicated. It's just disciplined.
Read the whole email. Ask one more question after you think you understand. Run the complete checklist even when you're sure everything is fine. Stay in the discomfort of not knowing yet just a little longer than feels necessary.
The right answer is usually a few paragraphs further than where you stopped.