The implicit question
Every prompt carries a hidden question back at the AI: who is asking? A 600-word blog post for a first-year high school teacher is a different piece of writing than a 600-word blog post for a 30-year classroom veteran. Same task. Different reader, different author.
Most prompts leave this implicit. The AI guesses — and it tends to guess a generic, educated, middle-of-the-road user. That's where generic AI outputs come from.
Three self-facts that change everything
1. Who you are
Not a resume. A sentence:
I'm a 14-year high school English teacher in a mid-sized district. I'm skeptical about technology fads but curious about what's real.
2. What you're optimizing for
Also one sentence:
I want feedback that helps me ship a lesson tomorrow, not research for a PhD.
3. What your constraints actually are
The honest ones:
I have 45 minutes between now and end of day.
My school is cautious about AI, so anything I build has to be defensible to a skeptical principal.
These three together turn a generic reply into your reply.
Templates across roles
Steal these, tune them:
Student:
I'm a sophomore studying cognitive science. I learn best by arguing. I want to understand this topic well enough to explain it to a first-year. I have 20 minutes.
Engineer:
I'm a mid-level backend engineer newish to Go. I care more about clarity than cleverness. I want to ship this by EOD.
Nonprofit operator:
I run programs at a 12-person nonprofit. Our funders are cautious. I need to write for both a warm donor and a skeptical board member in the same email.
Write your own. Save it somewhere. Paste it into prompts for serious work.
Homework
Write the three-sentence self-description that fits you, today. Not a branded one. A working one. Save it. We'll use it in Discernment next module.
Inspired by Anthropic's "AI Fluency: Framework & Foundations".