Everyone thinks the secret to getting insane results out of Fable 5 is writing a better prompt.

@ai.samaritan

12.4K followers · usually 3.8K views a post

Views

53.4K

Likes

3.3K

Comments

556

SoundTikTok sound · original sound - ai.samaritan

Promoted product

Claude Fable 5practicaly.ai ↗

Why it worked

The video provides actionable, specific prompts for using AI tools, framed within a relatable problem (getting better results). The creator's personal experience and the clear, numbered structure make the advice easy to follow and implement.

Summary

This video explains three prompting techniques for AI models like Fable 5 to achieve better results. It emphasizes identifying blind spots, having the AI interview you to uncover ambiguities, and prototyping before building to understand desired outcomes.

Structure

1Introduce the misconception about AI prompts
2Explain the 'blind spot pass' prompt
3Explain the 'make it interview you' prompt
4Explain the 'prototype before you build' prompt
5Conclude with the importance of finding unknowns

Product placement

Claude Fable 5 ACTS - it is an AI model that generates text and code based on prompts. Removing it would change what happens in the video. Fable 5 is the subject of the video. Practicaly AI APPEARS - it is the creator's website, but it is not operated or used to generate anything in the video. Removing it would not change what happens in the video.

Original caption

Day 7 of making you 1% better at using AI every day Everyone thinks the secret to getting insane results out of Fable 5 is writing a better prompt. It's not. Here are the 3 prompts I now run before starting anything real 👇 1️⃣ THE BLIND SPOT PASS Before touching the code, say: "I'm new to this part of the codebase. Do a blind spot pass and show me my unknown unknowns so I can prompt you better." → Surfaces the potholes you didn't even know to ask about. 2️⃣ MAKE IT INTERVIEW YOU Say: "Interview me one question at a time about anything ambiguous, and prioritize the questions where my answer would change the architecture." → Catches decisions you didn't realize you needed to make. 3️⃣ PROTOTYPE BEFORE YOU BUILD Instead of describing what you want, say: "Make me one HTML page with four wildly different directions so I can react." → Half the time we don't know what we want until we see it. Your prompt is a map. The real work happens in the territory. You can't prompt your way around what you can't see. That's the real skill now. 🎯 #claude #aitips #aitools #learnai #practicalyai

More from @ai.samaritan

On-screen text

#1- The blindspot pass Blind Spot Pass When starting work, one of the most useful things you can do is understand your blindspots. For example, if you're writing a feature in a new part of the codebase or using Claude to help you with unfamiliar work like iterating on a design, you're likely to have a lot of unknown unknowns. To do this, you can ask Claude to help you find your unknown unknowns and explain them to you. I like to use the literal words "blindspot pass" and "unknown unknowns" Giving it context on who you are and what you know is usually important for #2- Make it interview you Interviews Once I've done sufficient brainstorming, I likely still have unknowns. In this case, I ask Claude to interview me about any unknowns or ambiguities. When asking Claude to interview you, try and give it context about your problem to guide its questions. Here are some examples. "Interview me one question at a time about anything ambiguous, prioritize questions where my answer would change the architecture." #3- Prototype Brainstorms and prototypes When I'm working in an area with a lot of unknown knowns, involving criteria I only know to define when I see it, I like to ask Claude to brainstorm and prototype with me. It's extremely valuable to identify and verbalize unknown knowns early during prototyping, because finding them out during implementation can be (relatively) expensive. Small changes in a feature or spec can cause drastically different implementations in code and it can be more difficult for your agent to revert previous changes. For example, you may just want to see how a button added to a frame looks without having to wire up a backend route or maintaining additional state in the frontend. Known knowns what's in your prompt What you tell the agent you want. Unknown unknowns "I'll know it when I see it" Too obvious to write down, but you'd recognize it. Known unknowns questions you know to ask What you haven't figured out yet - and know it. Unknown Unknowns what you never considered The potholes you didn't know the road could have.