Lenny RachitskyHost of Lenny's Podcast | Author of Lenny's Newsletter

The 2 types of prompt engineering

2 Jul 2025 · 1 min

Watch on YouTube

0:47

Lenny Rachitsky breaks down prompt engineering into two modes: conversational, where you iterate through back-and-forth with the AI to refine outputs, and classical, where you perfect one prompt for repeated use in products.

Chapters

First of all, there's the conversational mode in which most people do prompt engineering. You're using Claude, you're using ChatGPT. You say, hey, you know, can you write me this email? Does kind of a poor job and you're like, oh no, like make it more formal or add a joke in there. And it adapts its output accordingly. I refer to that as conversational prompt engineering because you're getting it to improve its output over the course of a conversation. Notably, that is not where the classical concept of prompt engineering came from. It actually came a bit earlier from a more AI engineer perspective. You're like, I have this product I'm building, I have this one prompt that are super critical to this product. I'm running like millions of inputs through this prompt each day. I need this one prompt to be perfect. I just take this one prompt, improve it and then never change it. Those are the two modes.