How to Refine ChatGPT Outputs Using the Narrowing Technique

Here's how to use ChatGPT properly Part 6 Once you've run your ChatGPT prompt for the first time and you've received the initial result, this is where we can really refine and make our output a lot better. So we do this through a process called narrowing. You've been following along with this series and you've downloaded the free guide. Then you'll know that the framework that I teach is called risen, I.e. roll instructions, steps, end goal, and narrowing. Narrowing is the final step, and it's one of the most important. Basically, once you have an output, we can tell what's wrong with it. So we use iteration to improve the initial prompt. For example, let's say that I've asked ChatGPT to write a blog article for me and when it comes out it's full of language that sounds very robotic. It doesn't sound human at all. In the narrowing stage, I'm going to add an instruction like make sure that it sounds human. Or I've asked it to write a social media post and it's filled it with emojis, which I hate. I'm going to add a narrowing step, which is no emojis. Importantly, we do this at the end of the process, once we have tested our prompt, to see what the output's like.