A five-minute exercise to see whether your next AI check adds insight—or just another step.

The Bridge | Be A Decisive Leader

Remember taking an open-book exam?

You knew an answer. But the book was right there, so you checked anyway.

Sometimes you caught a mistake. Sometimes you confirmed what you already knew and used time you needed for the harder questions.

AI makes that same habit easy to bring into the workday.

You know what a routine email needs to say. You understand the trade-off. You have enough information for a small, reversible decision.

Then: Let me just ask AI.

That can be useful. It can also turn one decision into several: which version to use, which caveat matters, whether to ask again.

AI makes information easier to access and synthesize. The underlying quality still matters. Garbage in, garbage out has not disappeared because the answer reads well.

And when we are already tired of making decisions, another round of options may add friction.

I’m interested in a specific risk here: habitual checking may make us less comfortable acting on our own judgment. That is a possibility to watch in our behavior, not a universal claim about what AI does to people.

Try this: the five-minute check

Choose one familiar, low-stakes, reversible decision this week.

Before opening AI, write:

1. My decision: What would I do?
2. My reason: What evidence supports it?
3. My uncertainty: What specific information could change my mind?

If you identify a meaningful gap, use AI to help investigate it and check consequential claims against a reliable source. If you cannot identify one, make the decision.

Afterward, take a minute to record whether the choice worked and what you learned.

That small practice follows the Decisive EDGE rhythm: context, decision, action, takeaway. It also helps interrupt one route into Decision Debt—keeping a choice open after further analysis has stopped serving a clear purpose.

This week’s article explores the open-book trap and offers a practical guide to when to automate, augment, verify, and retain human ownership.

Read “The Open-Book Trap: When AI Adds Work to Decisions We Already Know How to Make” →

For your next AI interaction, ask:

What do I need to learn before I can act?

Give the tool that job. Then use your judgment.

— Matthew