# Appendix G: The \"What Did I Miss?\" Question Ladder

From Let the AI Be Smart by Jason Santiago. lettheaibesmart.com/tools

Run these in order, after any answer that matters. The first rung is the one that does the heavy lifting, because it's the only one that doesn't point anywhere. Every other question aims at a hole you already half-suspect. The first one admits you don't know where the hole is, and lets the model go find the blind spot you had no idea existed.

- "What did I miss?" Hands the whole thing over and lets the model look from angles you physically cannot see from inside your own frame.

- "What are you assuming about my situation?" Drags the baked-in assumptions out where you can shoot at them, because the model filled every gap you left with a guess.

- "What question should I have asked before asking this?" Checks whether you're standing on the real problem or just decorating the symptom that happened to be annoying you today.

- "What would make this wrong?" A real claim can name its own undoing. If nothing would make it wrong, you're looking at a belief dressed up as a fact.

- "What would an expert challenge here?" Aims the pressure at the weak spot you already have a hunch about, and makes the model argue against its own answer.

- "What user did I forget about, and what breaks when this scales?" Surfaces the person and the load that weren't in the picture you drew, before reality surfaces them for you.

- "Is there a better solution than the one I walked in with?" Gives the model permission to tell you your whole plan is a smaller idea wearing the costume of a big one.

- "Restate what I'm asking for, and the why underneath it." The gap between what you said and what it heard is information you can't get any other way, and the missing piece is almost always the why you left in your own head.

Two things about running the ladder. First, mean it when you ask. These questions cost you the feeling of being the smartest one in the room, and that's the price of the big answer instead of the small one. Second, don't stop at the rung that flatters you. The rung that stings is the one doing the work.

Ask it what you missed, and let the AI be smart.