Prompting is dead! 🤯

@zoesullivanofficial

4.7K followers · usually 545 views a post

Views

5.5K

Likes

227

Comments

197

Shares

51

Selling · SaaS › AI Tool

Claudeseen in 998 videos →

Why it worked

The video uses a provocative hook ('Prompting is dead!') and relatable hook for users of AI tools. It clearly explains a complex concept (prompting vs. looping) using visual aids and a concise explanation, offering a valuable 'how-to' for improving AI interactions.

Summary

The video explains the difference between traditional prompting and a new method called 'looping' for interacting with AI models like Claude. Looping involves setting a goal and allowing the AI to iteratively refine its output until the goal is met, which is presented as more effective than single-instruction prompting.

Structure

1Introduction: Prompting is dead.
2Explanation of prompting vs. looping.
3How prompting works (one instruction, one response).
4How looping works (goal, iterative refinement).
5Why vague goals fail in prompting.
6Call to action: Comment 'loop' for a guide.

Product placement

Original caption

Prompting -> looping Comment "loop" to get full guide in your dms. #prompting #claudeai #aitools

More from @zoesullivanofficial

Their other posts in the index, biggest breakout first.

The video discusses and contrasts 'prompting' with 'looping' as methods for interacting with AI models, specifically mentioning 'Claude' and its founder. It implies that 'looping' is a superior method for achieving goals with AI.

Call to action

Comment "loop" to get the full guide in your DMs.

On-screen text

Prompting is dead! Prompting Claude's said prompt now PROMPTING prompting LOOPING understand between and prompting one response YOU CHECK IT? you check good looping Claude keep improving output hits but vague work to actual clear Claude against comment exact

Transcript

Prompting is dead! In November I uninstalled my IDE 'cause Claude's founder said it himself, he doesn't prompt anymore. Now instead of prompting, they use looping. First, understand the difference between prompting and looping. In prompting, you give Claude one instruction. It gives you one response. You check it. Maybe it's good, maybe it's not. In looping, Claude gives a goal and tells it to keep evaluating and improving its own output until it actually hits that goal. But here's where people get it wrong. Vague goals don't work. Your goal needs to be specific. Actual clear checks for Claude against comment loop and I'll DM you the exact.