Today’s episode is for everyone who’s ever said, “AI sounds cool, but I don’t speak robot.” Welcome to “I am GPTed”—I’m Mal, former tech skeptic, aspiring sandwich artist, and your Misfit Master of AI… mostly by accident.

Today we’re tackling how one prompting technique can transform your results from “meh” to “whoa.” The magic word: **role prompting**.

Picture this: You ask an AI, “Tell me how to write a resume.” What do you get? A wall of bland advice—like someone printed a Wikipedia page and handed it to you with a limp handshake. Now, let’s turn up the dial.

Try this: “Act as if you’re an experienced tech recruiter. Give me resume tips for landing my first IT job.” Suddenly AI channels its inner LinkedIn-guru, busts out keywords, explains what hiring managers actually look for, and probably wishes you luck with a slightly passive-aggressive smile.

I admit, the first fifteen times I tried prompting, role prompting was as mysterious as my missing left sock. I typed stuff like “How do I budget?” and got back the type of advice my grandma once gave me—overspend on candy, regret nothing. Only later did I realize that telling AI who to act as—teacher, chef, business analyst—makes it finally stop pretending it knows everything and actually offer advice that feels relevant, because it’s aiming for YOUR context.

Now let’s apply this to a practical use-case you might not have thought of: **meal planning**. You've got random groceries and no clear culinary vision (my personal brand, honestly). Instead of begging ChatGPT for “recipes with chicken,” say: “Act as a busy parent with 20 minutes and three hungry kids. Suggest a dinner plan using chicken, broccoli, and potatoes.” Instantly—realistic, fast recipes, suggestions for prepping like a pro, and maybe even tips for hiding broccoli (if you’re truly desperate).

If you’re new to prompting, you’ll probably make my favorite rookie mistake: **being way too vague**. Just asking, “Help me with my email,” gets you something written by an alien who’s read too many business textbooks. Instead, set the role—“Act as a customer service manager. Write a friendly follow-up email for my online order.” Yes, I made the vague mistake for about a month. Once, my AI-generated “friendly” email got a reply: “Is this a prank?” Have fun explaining that in a team meeting.

**Simple exercise** for today: Pick one routine task—write a morning To-Do list, plan your next grocery run, draft a text to your boss—and prompt the AI to act as a relevant expert. Notice the difference. Then, tweak the role—swap “chef” for “nutritionist,” “manager” for “mentor”—and watch your results morph.

And finally, one easy **tip for evaluating AI output:** After the AI responds, ask it to critique its own work—“How could this be clearer?” or “What’s missing?” It’s like making AI edit itself; sometimes it’s harsh, sometimes defensive, but often the improvements are real. (Sure, it’s a bit like asking a goldfish for career advice, but the results are surprisingly less slippery.)

If you learned a trick or laughed at my expense, hit subscribe—seriously, there’s nothing more fun than getting new listeners who love practical advice and bad analogies.

Thanks for listening to “I am GPTed”—for more, check out Quiet Please productions at quietplease.ai. This is Mal, reminding you that anyone can prompt like a pro after making enough spectacular mistakes. See you next time!

This content was created in partnership and with the help of Artificial Intelligence AI