By Amanda Caswell

I wasn't expecting a lesson in self-awareness

(Image credit: Gabby Jones/Bloomberg/Getty)

If you've spent any time on social media lately, you've probably seen people asking ChatGPT to "roast" them. With ChatGPT's more dynamic memory mode, I decided there's never been a better time to give this a try. And yes, the prompt is exactly what it sounds like. You tell the chatbot to be brutally honest and it responds with a sarcastic breakdown of your habits and personality quirks.

I was expecting to laugh, but what I got was an exercise in self-reflection. Like many people, I occasionally wrestle with self-doubt. It shows up in everyday life as the voice in my head that questions whether I'm falling behind or if everyone else but me somehow has things figured out. Here's what happened and how to try it yourself.

The exact prompt I used

I started with the following prompt: Roast me based on everything you know about my work habits, goals, personality and recurring frustrations. Be brutally honest and specific. Focus on the things that hold me back most often.

ChatGPT did exactly what I asked. It pointed out tendencies to overthink, moments where I seemed overly critical of myself and situations where I appeared to focus more on what wasn't working than what was. And while I usually include a screen shot here, I have to keep it private. The roast was uncomfortably accurate. Try it and you'll see what I mean.

Latest Videos From

Watch full video here: Best Tech Gifts Under $100

But after reading through the response, I noticed that the "roast" didn't reveal anything new. The AI put into words thoughts I already had about myself. It felt as though ChatGPT was actually echoing my self-doubt.

So, I followed up with a prompt: For each criticism you just made, tell me whether it is a fact, an assumption or a cognitive distortion. Then provide evidence for and against it.

That's where there real breakthrough came through for me. By having ChatGPT analyze its own criticism of me, the roast became an investigation to show me actual evidence behind it based on our conversations. It will deliver the information in a chart so you can see everything side-by-side.

You may like

• I used the ‘5 Whys’ AI prompt to find my professional blind spots — the results were uncomfortably accurate

• I asked ChatGPT to roleplay as Simon Sinek — the results rewired my brain for productivity

• I asked NotebookLM to coach me like Jay Shetty — it exposed what ChatGPT missed

And that's where things got interesting because facts and feelings aren't the same thing. As the chatbot worked through its criticisms, I began to see a pattern. Many of the statements that felt true at first turned out to be assumptions.

For example:

Want more AI news?

Sign up to the Tom's AI Guide weekly newsletter summing up all the biggest AI news you need to know. Plus, analysis from our AI editors and tips on how to use the latest AI tools!

Contact me with news and offers from other Future brandsReceive email from us on behalf of our trusted partners or sponsorsBy submitting your information you agree to the Terms & Conditions and Privacy Policy and are aged 16 or over.

Fact: A project didn't perform as well as expected.
Assumption: Everyone else is doing better than me.
Cognitive distortion: Believing one disappointing result is proof of long-term failure.

The distinction seems obvious when written down. Yet in everyday life, those thoughts often blend together. You may have seen them yourself. A setback becomes an identity or one mistake becomes a prediction about the future. The exercise forced me to separate what had actually happened from the story I was telling myself about what it meant.

The real source of my self-doubt

(Image credit: Getty Images)

What surprised me most was that the source of my self-doubt wasn't a lack of confidence, but a lack of evidence. When negative thoughts appeared, I rarely challenged them, I simply accepted them as facts.

What to read next

• I gave ChatGPT permission to disagree with me with this prompt — and its responses became dramatically better

• I use the ChatGPT ‘alpha' prompt to make stronger arguments — here’s how it works

• I used ChatGPT to try the ‘Let Them’ rule for 24 hours — and I finally stopped overthinking

The roast prompt exposed a habit many of us share, which is treating our inner critic as a reliable narrator. But once the criticism had to defend itself with evidence, much of its power disappeared. The process reminded me that self-doubt often survives because we never ask it to prove its case.

One final prompt

Before ending the experiment, I tried one last prompt: Based on everything you know about me, create a fact-based case for why I am more capable than I currently believe. Use evidence only. No motivational language.

What I like about this prompt is that it doesn't give you a pep talk, it gives you proof. Meaning, rather than a generic sense of encouragment, ChatGPT points to actual accomplishments, skills and habits it has seen first-hand.

The AI reminded me that confidence doesn't always come from positive thinking but from seeing the evidence and reviewing it honestly.

From roast to realization

The viral roast prompt wasn't valuable because it insulted me. What was valuable is that it gave me real information about myself to analyze with it.

The roast surfaced the criticisms already living in my head. The follow-up prompts helped me determine which ones were true, which ones were assumptions and which ones were distortions. These prompts exposed how much of my thinking is built on conclusions I had never bothered to fact-check.

The next time your inner critic starts making bold claims about your abilities, give this prompt a try. You might just discover a new way to overcome self-doubt like I did.

Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok. Finally, you can visit our dedicated Tom's Guide Savings Squad hub for expert help on getting the best products for less.

More from Tom's Guide

• I tried Google Labs’ Dreambeans app — and it finally broke my infinite scrolling habit
• Reddit users aren’t picking one AI chatbot anymore — here’s how they stack their tools
• I thought Siri was finished. These 5 leaked WWDC 2026 features are its biggest update since 2011

Amanda Caswell

AI Editor

Amanda Caswell is the AI Editor at Tom's Guide and one of today’s leading voices in AI and technology.

A celebrated contributor to various news outlets, her sharp insights and relatable storytelling have earned her a loyal readership. Amanda’s work has been recognized with prestigious honors, including outstanding contribution to media.

Known for her ability to bring clarity to even the most complex topics, Amanda seamlessly blends innovation and creativity, inspiring readers to embrace the power of AI and emerging technologies.

As a certified prompt engineer, she continues to push the boundaries of how humans and AI can work together.

Beyond her journalism career, Amanda is a long-distance runner and mom of three. She lives in New Jersey.