Ever had that moment where you discover a simple trick that makes you wonder why everyone isn’t using it? This is one of those moments.

Most people approach AI like they’re ordering from a drive-through menu. “Give me advice.” “Write me an email.” “Help me with this problem.” And what do they get? Generic, vanilla responses that could apply to anyone, anywhere. But there’s a hidden lever in AI that most people never pull, and it’s ridiculously simple: give it a specific role to play.

Instead of asking AI to “review my business plan,” try “pretend you’re a venture capitalist who’s seen 500 startup pitches and review my business plan.” Watch what happens. The feedback suddenly has teeth. It asks about burn rates, market validation, and competitive moats. It’s like switching from a foggy pair of glasses to HD vision.

The Psychology Behind the Magic

Think about what happens when someone asks you to “think like a detective” versus just “solve this problem.” Your entire mental framework shifts. You start looking for clues, questioning assumptions, connecting dots differently. You’re accessing a completely different mental model.

AI works the same way, except it has thousands of mental models from analyzing millions of conversations, books, and expert communications. When you activate a specific role, you’re essentially saying, “Hey, use THAT specific filing cabinet of knowledge, not the general one.”

It’s like the difference between asking a Swiss Army knife to “cut something” versus specifically pulling out the scissors. Same tool, vastly different results.

Let’s Test This Out

Instead of: “How can I improve my presentation skills?”

Try: “You’re a TED talk coach who’s prepared speakers for main stage events. How can I improve my presentation skills?”

The difference isn’t subtle. Where the first might give you bullet points about eye contact and practice, the second dives into story arcs, the power of pauses, and how to create what speakers call “goosebump moments.”

Roles That Work Surprisingly Well

The Skeptical Investor:
“You’re a skeptical angel investor. Poke holes in my business idea.” This one’s brutal but invaluable. It’ll find weaknesses you’ve been blind to.

The Five-Year-Old:
“Explain this concept like you’re talking to a curious five-year-old.” Perfect for when you need to simplify complex ideas without losing their essence.

The Time Traveler from 2034:
“You’re visiting from 10 years in the future. What advice would you give me about my career choices today?” This creates fascinating perspective shifts.

The Devil’s Advocate Lawyer:
“You’re a lawyer whose job is to find every possible flaw in this contract/agreement/plan.” Watch how quickly it spots issues.

The Experienced Therapist:
“You’re a therapist with 20 years of experience in workplace dynamics. Help me navigate this situation with my colleague.” The emotional intelligence in responses shifts dramatically.

The Michelin-Star Chef:
“You’re a Michelin-star chef. Transform this basic recipe into something extraordinary.” Even if you’re just trying to jazz up weeknight pasta, the suggestions get creative.

The Minimalist Designer:
“You’re a minimalist designer inspired by Dieter Rams. How would you redesign my workspace?” Suddenly you’re getting advice about negative space and functional beauty.

Here’s where it gets interesting: the more specific you make the role, the more nuanced the response. “You’re a therapist” gives different answers than “You’re a therapist who specializes in perfectionism and imposter syndrome.” AI picks up these subtleties and runs with them.

The Secret Sauce: Combining Roles

Want to really level up? Combine perspectives. “First, analyze this as a cautious CFO. Then, look at it as an innovative startup founder.” The contrast between viewpoints often reveals insights neither role would catch alone.

Or try sequential roles: “Start as a brainstorming creative director, then switch to a practical project manager to make it actionable.” It’s like having a mini advisory board in your pocket.

Making It Work for You

The key is matching the role to your actual need. Need brutal honesty? Call in the skeptical venture capitalist. Need encouragement while learning? Bring in the patient mentor who’s been teaching for decades. Need to see blind spots? The devil’s advocate is your friend.

This isn’t about tricking AI or playing games. It’s about accessing the right knowledge in the right way. Just like you wouldn’t ask a pastry chef for advice on car repairs, you shouldn’t ask generic AI for specialized insights when you could activate expert mode instead.

Next time you’re about to type a question into AI, pause. Ask yourself: who would be the perfect person to answer this? Then tell AI to be that person. The transformation in response quality might just make you wonder why this isn’t taught as AI 101.

Because really, why would you settle for generic when specific is just one sentence away?

Want to level up your Gen AI skills or need help developing an AI strategy that actually works? Let’s connect on LinkedIn.