Simple, short and correct

About: Every Saturday, I share my thoughts on working with AI. Subscribe for free.

Simple, short, and correct.

That's the kind of answer I'm looking for. Always.

No matter if it comes from AI or a human, including myself.

Before AI chats, the hard part was finding someone you could trust and getting answers from them. Once you had that, you could assume they were correct and accept complex answers.

But now, with AI chats, it's different. The trust part is gone.

Sometimes they're right. Sometimes they're wrong. And it happens all the time. You can't assume correctness anymore. Fact-checking is now part of the interaction. It's part of the game.

So what? Do you avoid these magic answers generated at the speed of light?

No. But I try to force them to be simple and short. That makes checking easier. And when an answer can't meet those two traits, I almost always assume it's wrong and dismiss it.

This way, I get the most out of AI chats while protecting my time and energy.

What about you? How do you handle this trust part of AI chats?

Recent posts
featuredConfidence vs. faith
Start building confidence in your ability to guide AI
Logging requests when the OpenAI API errors in Python
Add context to OpenAI API error messages for easier debugging
code
Improve your docs by giving your AI assistant the project's issues
See why a virtual keyboard bug calls for issue-aware docs
misc
Curious about the tools I use?