A simple way to write better prompts

Improve your results with "pre-prompting"

I hate the phrase prompt engineering.

It turns the simple act of creating a prompt into a scientific process reserved for folks with IQs above 135.

It overcomplicates an activity (i.e. prompting) that is actually quite straightforward.

But I want to help you simplify. Without sacrificing your results.

Here’s a simple tactic that you can use to rapidly improve your prompts — and the resulting outputs from your LLMs.

Introducing “pre-prompting”

The first conversation I ever had with ChatGPT was about the voice in my head. Specifically, I wanted to know:

How can I quiet my inner critic?

ChatGPT’s response was serviceable. But it wasn’t better than an SEO-optimized blog post.

I was so young and innocent back then — I didn’t know that a bad prompt would lead to a bad answer.

And with that first query, I failed to recognize that ChatGPT knew nothing about me.

ChatGPT didn’t know:

• My age

• The typ…