Thank you to everyone who joined our live session yesterday! For those who couldn't make it, here's an update on what you missed and what's coming next.

Click this link to watch the replay of the workshop!

Yesterday, we hosted a workshop on how to test AI products. Maria led the session with help from Scott, covering why testing matters for AI tools and how to do it effectively. Here's what we learned about testing your AI products, including Chipp.

Key Points from the Workshop

  • The Testing Mindset: Break often like waves, not porcelain - making testing a regular habit rather than waiting for perfection.
  • AI vs. Traditional Products: AI gives different outputs even with the same input, requiring more diverse testing approaches than traditional products.
  • Different Approaches for Different Goals: Use role-based prompts for creative outputs and specific examples for structured results.
  • The Dynamic Nature of AI: Unlike static products, AI tools change as models update, making testing an ongoing process rather than a one-time event.
  • The Magic Number: Test with about eight diverse people per round for the most efficient feedback before insights start to repeat.
  • Who to Test With: Include your ideal users, people outside your target audience, and those who might use your product in unexpected ways.
  • WhatsApp as a Testing Tool: WhatsApp provides a low-barrier platform for AI testing with the added benefit of preserving chat logs for analysis.
  • Prioritizing Feedback: Not all feedback requires action - choose what to implement based on your product's core purpose.

Next Steps

The most important takeaway: start testing now. For those using tools like Chipp, WhatsApp works great for testing because it's easy to use and lets you see how people actually interact with your AI.

Ready to learn more? Join the Chipp Disco Community for more events like this.

The link has been copied!