And how my prompt fixes that

By author (one image from pexels)

Not a member? Click here

I was reading an article the other day. It talked about how ChatGPT was filled with biases; it would lie if it meant getting an answer that looked better or that agreed with the user.

Most people, including me, have already noticed this, but I’ve never seen how you can make ChatGPT actually tell you the truth.

Basically, it lies to you to confirm what you wanted to hear.

Sure, you can add:

“and be completely honest, don’t say things just to be nice”

But it’s never made much of a difference, and felt oddly vague.

So I decided to test things out, and after almost an hour of testing, I was finally able to scientifically (fancy, I know) engineer the best ChatGPT prompt for getting an honest answer.

Why You Probably Need This

Let me show you some examples of how ChatGPT usually answers, just so you understand why you need this prompt:

Leave a Reply