How laughable.ai Works

Commercialize your cynicism. Get paid for reporting AI failures.

Our Mission

laughable.ai is a community-driven platform dedicated to improving AI systems by identifying and cataloging their failures. We believe that by collecting real examples of AI mistakes, we can:

  • Help AI researchers understand common failure modes
  • Create better training data for future AI models
  • Educate users about the limitations of current AI systems
  • Build a comprehensive database of AI failures for research

How It Works

1. Test AI Models
Interact with various AI models and test them with challenging prompts

Use our platform to test different AI models like GPT-4o, Claude, and Gemini. Try prompts that you think might cause the AI to make mistakes or generate incorrect information.

2. Identify Failures
Spot and report AI mistakes, hallucinations, and errors

When you find an AI response containing errors, categorize the type of failure (hallucination, logical error, factual mistake, etc.) and submit it to our database.

3. Verification Process
Our team reviews submissions for accuracy

Submitted failures are reviewed by our team to verify they are genuine AI mistakes. This ensures our database contains high-quality, verified examples of AI failures.

4. Earn Rewards
Get paid for valuable contributions

Contributors earn points and monetary rewards based on the quality and quantity of verified failures they submit. Top contributors can earn up to $1,000 per month.

User Levels

Level 1-2: Beginner

New users start here. Submit your first verified failures to start earning points and advance to higher levels.

Level 3-4: Advanced

Users with a track record of quality submissions. Earn higher rewards per verified failure.

Level 5: Expert

Our top contributors who consistently find valuable and unique AI failures. Highest reward tier and special recognition.

For AI Research Teams

AI research teams can access our verified database of AI failures to:

  • Improve model training with real-world failure examples
  • Identify common failure patterns across different models
  • Test and benchmark model improvements
  • Access categorized and verified training data

Ready to Get Started?

Join our community today and start finding AI failures, earning rewards, and helping improve AI systems for everyone.