AI’s biggest risk isn’t errors, it’s making thinking optional.

Generate ideas first; let AI refine, challenge, and expand them.

Use the Three Role Rule to keep AI as a partner, not a substitute.

This post continues the conversation we began in Part I, where I introduced the One Thought Rule: the simple but powerful practice of generating at least one original idea before turning to AI. That rule stands as the foundation of this entire system.

The core idea from Part I is straightforward: your brain must generate at least one original thought before you consult the machine.

Whether you’re a manager drafting a strategy memo or a student staring down a response paper on Robert Cialdini’s principle of reciprocity, the temptation is the same: open an AI tool, type a prompt, and let the machine do the heavy lifting. Within seconds, a clean, well-structured draft appears on your screen. Efficient? Absolutely. But efficiency has a shadow side, and it’s one we don’t talk about nearly enough.

If you’re not careful, your mind barely participates.

For readers of this Power & Influence blog, that’s not a small concern. Influence depends on clarity. Clarity depends on thinking. And thinking depends on effort.

The real risk of AI isn’t that it occasionally gets things wrong. It’s that it makes thinking optional.

Fortunately, there’s a way to use AI that sharpens your mind instead of dulling it. It’s simple, practical, and surprisingly easy to adopt. I call it the Three-Role Rule, and it’s designed to keep your brain in the driver’s seat while still taking full advantage of what AI can do.

Before we get to the roles themselves, it’s worth pausing on one key insight: Thinking is not a single act. It’s a sequence. You generate ideas, you refine them, you test them, and only then do you expand them. Skip the first step, i.e., the generative one, and the rest of the sequence collapses. That’s where cognitive diminishment begins.

AI is wonderful at accelerating the middle and later stages of thinking. But it should never be the first thinker in the room. That job belongs to you.

The Three-Role Rule rests on a simple premise: Once you’ve generated your own ideas (thanks to the One Thought Rule from Part I), AI can play three extremely helpful roles, i.e., roles that strengthen your thinking rather than replace it.

Each role has a distinct purpose. Each engages a different cognitive system. And each keeps your brain active in ways that protect your long-term intellectual health.

Role One: The Refiner

Once you’ve produced your own draft, however rough, AI becomes a kind of refiner. Not an editor, not a ghostwriter, and certainly not the primary author. A refiner. Something that helps you see what you can’t see on your own.

This is where AI can be extraordinarily helpful. It can point out where your argument is thin, where your logic jumps too quickly, where your explanation is muddy, or where you’ve buried the lede. It can help you clarify, strengthen, and sharpen your thinking. But it can only do that if you’ve given it something to refine. If you hand it a blank page and ask it to start, you’ve already surrendered the most important part of the process.

And there’s an added benefit: asking for critique, even from a machine, is a form of practice for receiving criticism from humans, which is an important skill to have. People with a growth mindset lean into this discomfort because they know it’s how improvement happens.

Role Two: The Challenger

After you’ve refined your ideas, you’re ready for the second role: the challenger. This is the part most people skip, and it’s the part that makes the biggest difference. Once your ideas are on the page and reasonably clear, AI can become an intellectual sparring partner. It can push back, poke holes, and force you to defend your thinking.

Ask it how someone who disagrees with you might respond. Ask it what assumptions you might be making without realizing it. Ask it what the strongest counterargument would be. This is where your reasoning deepens. This is where your blind spots surface. This is where your thinking becomes more resilient.

And cognitively, this kind of stress testing, revising, and improving (which engages key regions and networks of the brain) is what makes your thinking more resilient. In persuasion research, this is known as inoculation, i.e., exposing your ideas to challenge strengthens them. It’s the same principle that makes good trial lawyers formidable: They anticipate the counterarguments before they ever enter the courtroom.

Role Three: The Expander

Only after your ideas have been generated, refined, and challenged should you let AI play the final role: the expander. This is where it broadens your perspective, introduces related concepts, and helps you see beyond your initial frame. It might suggest research you hadn’t considered, frameworks you hadn’t applied, or examples that strengthen your point. This is where AI shines, not as a replacement for your thinking, but as a supplement to it.

When you put all of this together, you get a thinking sequence that looks like this: you generate first (Part I), then refine, then challenge, then expand. It’s a full cognitive cycle, and it keeps your brain active at every stage. Most AI workflows skip the first step entirely. That’s where the trouble begins.

Putting It Into Practice

If you want to try the Three-Role Rule right now, here’s a simple exercise. Before you finish this article, write two sentences about how AI is affecting your work or studies. Don’t overthink it. Then ask yourself how you would defend or expand that position. That’s your thinking muscle warming up.

The most powerful intelligence in the room should still be your own. AI can sharpen your thinking, challenge it, and broaden it, but it should never replace the act of thinking itself. Use AI as a partner. Keep your brain in the driver’s seat. And let your thinking, not your tools, be the source of your influence.

AI can help you think better, but it can’t do the mental work that’s critical to your overall success.

There was a problem adding your email address. Please try again.

By submitting your information you agree to the Psychology Today Terms & Conditions and Privacy Policy

© Psychology Today