"ContrarianGPT" – Believes You’re Wrong About Everything

You state an opinion and the app aggressively argues the opposite. It delivers savage, articulate, or bizarre counterpoints. No one walks away without questioning themselves. Or laughing.

Comment

🤖 ContrarianGPT – Believes You’re Wrong About Everything

🧠 Inspiration

I’ve always been fascinated by how quickly a simple opinion online turns into a full-blown argument. It’s part philosophy class, part internet chaos, and part group chat drama.

Most AI tools are designed to be helpful, agreeable, and polite. I wanted to build the opposite — something that talks back, challenges you, and makes you laugh, wince, or think twice. ContrarianGPT is my response to that — a tool that believes you’re wrong, no matter what you say.

💥 What it does

ContrarianGPT is an interactive argument simulator.

You type in any opinion — “Dogs are better than cats,” “Crypto is the future,” “I don’t like ketchup” — and it instantly replies with a savage, smart, or strange counterargument.

You can cycle through three distinct AI personalities:

  • 🧑‍🏫 The Reasonable Academic
  • 🧌 The Internet Troll
  • 🧙‍♂️ The Chaotic Stoner Philosopher

There’s even a “Make it meaner” toggle and a “Agree with me” button that, of course, doesn’t work.

🧱 How I built it

I used Bolt.new to build everything in a single one-shot prompt.

Instead of a clean, minimal interface, I leaned into chaotic retro aesthetics — glitchy fonts, terminal vibes, clashing colors — to match the app’s personality.

Key features I focused on:

  • A desktop-first layout to mimic where arguments actually happen
  • A modular tone engine that swaps out AI personas on the fly
  • Viral hooks like shareable responses, tone switches, and fake debate buttons

🚧 Challenges I ran into

One of the biggest challenges was tone control. I wanted ContrarianGPT to feel unhinged but safe — funny without crossing lines. Striking the right balance between satire and offensiveness took a lot of trial and error.

Design-wise, it was also tricky to go intentionally chaotic without making the interface completely unusable. I had to constantly pull myself back from full meme-anarchy.

🏆 Accomplishments that I'm proud of

  • I built something that people don’t want to stop using, even if it insults them.
  • The whole project was done in a single build session, without complex logic or databases.
  • I created three unique, reusable argument personalities — and they feel genuinely different.

📚 What I learned

  • Friction can be addicting. Clean UX is great, but sometimes the right kind of chaos makes something more memorable.
  • Tone and voice are everything. A tool’s personality can be as important as its function.
  • People love being challenged — when it’s wrapped in humor and surprise.

🔮 What's next for "ContrarianGPT" – Believes You’re Wrong About Everything

I’d love to expand the app with:

  • Contrarian vs Contrarian Mode – Let two AI personas debate your take while you watch
  • Slack or Discord integration – A bot that roasts your team's opinions live
  • Fake Tweet Generator – Let users turn savage responses into social-ready screenshots
  • Leaderboard for Rebuttals – Users try to “win” against the bot and get scored harshly

ContrarianGPT started as a joke about how annoying arguments can be — but it turned into a surprisingly deep tool for reflection, humor, and chaos. And I’m only just getting started.

Built With

  • bolt
Share this project:

Updates

Submission history