
A few months ago I spent three days going around the same decision. A project had come up that I wasn’t sure I wanted. The scope looked right but something felt wrong and I couldn’t name it. I’d written out pros and cons. Twice. The second list looked almost identical to the first, and I was just circling.
So I tried Copilot differently than I usually do. Instead of asking for an answer, I described the whole situation and asked it to skip the advice and tell me what I should be thinking about. What questions did I need to answer before I could decide?
What came back was four questions I hadn’t asked myself. The third one stopped me. I realized the hesitation wasn’t about the project. It was about a pattern I’ve been letting slide that the project would have asked more of me.
What I’d needed all along was someone to help me see what I was actually weighing.
What this teaches you
Most people use Copilot like Google. They type a question, they want an answer back. That’s one way to use it, and it works. But it’s often more useful as a thinking partner. You describe a problem you’re working through, and you ask it to think with you: what should I be considering, and what am I missing?
The decision stays yours. What Copilot adds is the thing you hadn’t thought of, the question you forgot to ask yourself. You usually walk away with a clearer picture than when you started, even if the answer is the same one you were leaning toward.
Try this
Pick a decision you’re turning over right now. Something you’ve been thinking about in the background, something you keep meaning to sort out.
- Open Copilot (copilot.microsoft.com or your Copilot-enabled browser).
- Describe the situation the way you’d explain it to a trusted friend who doesn’t know your life: be specific about what you’re weighing. Like this: “I’ve been trying to decide whether to keep my shop open on Sundays. It’s my slowest day, maybe six customers. But a few of them come specifically on Sundays and I’d hate to lose them. I’m not sure if staying open is worth it, or if I’m just afraid to disappoint people.”
- Ask for help thinking, in so many words. Try one of these: “What should I be considering that I might not be thinking about?” Or: “What questions would help me figure this out?” Or: “What usually matters in a decision like this?”
- Read what it says. Some of it you already know. Some of it you don’t. The parts that make you think “hm, I hadn’t thought about that,” those are the ones that matter.
- If something it raises is worth exploring, ask it to go deeper: “Can you say more about that?” or “What would I need to know to answer that?”
You’re using it to organize your own thinking. If it offers a recommendation along the way, that’s the least useful part of the reply.
What to watch for
Copilot doesn’t know your situation beyond what you’ve told it. Its suggestions are generalities. It might raise something that doesn’t apply to you at all, so treat the response as a checklist to think against, and cross off whatever doesn’t fit.
And sometimes it will confidently suggest something that’s just wrong for your situation. It has no way to know. That’s why you’re the one in the conversation. You’re the expert on your life, and the tool is just helping you look at it from new angles.
The output matters less than the thinking that happens when you try to answer the questions it raises.
I took the project. Copilot never told me to, and I’d have been suspicious if it had. But by the time I’d written out my answers to those four questions, I knew what I’d been hesitating about, and it had nothing to do with the project. I’d have gotten there eventually on my own. I don’t think I’d have gotten there that afternoon.
Next in this series: how to tell when Copilot is actually right, and when it’s just sounding right. That skill will save you more time than anything else you’ve learned so far.
Making Sense of Artificial intelligence (AI) is a six-part series walking first-time users through the basics, one skill at a time.