← Back to all free guides
A free guide by Mariah Brunner ↗ INSTAGRAM ↗ TIKTOK

Use AI Where
You’re Already Good.
Not Where You’re Bad.

The AI rule almost no one is teaching: if you can’t judge the output, you can’t fix it. Plus the audit prompt that maps your green zones and red zones.

Most people reach for AI when they’re bad at something. I can’t write, so AI can write for me. Sales isn’t my thing, so AI can do my outreach.

It’s backward. If you’re bad at something, you can’t tell when AI is doing it badly either, so you ship polished garbage. Use AI where you’re already good, because then your taste filters the output.

The Rule

AI multiplies whatever judgment you bring. Where your judgment is sharp, you get faster, more polished work at the same quality bar.

Where your judgment is thin, you get faster, more polished mistakes. The output sounds confident and would fall apart the moment an expert read it.

There’s research behind this. In a Harvard and BCG experiment with 758 consultants, AI made them faster and better on tasks it was good at. On a task picked to sit just outside what the AI could handle, the consultants using AI were less likely to get the right answer than the ones working without it. (Navigating the Jagged Technological Frontier)

So the real question is where do I have enough taste to judge what AI gives me? That’s your green zone.

Green Zone Examples

01

You write with a real voice

Ask AI for 10 angles on a topic. You’ll see instantly which one sounds like you. AI brings the breadth, you bring the filter.

02

You’ve hired and built teams

Have AI draft interview questions, job descriptions or feedback scripts. You’ll spot a generic question in three seconds.

03

You’ve built a brand

Use AI for copy variants, headline tests and moodboards. You know what’s on-brand the moment you look.

04

You know your industry cold

Have it summarize trends and draft market briefs. You’ll catch a misread risk or missing context right away.

05

You’ve sold before

Have it draft outreach in 5 tones. You’ll know which one lands with your buyer before you hit send.

Red Zone Examples

01

Contracts when you can’t read one

It’ll produce something that sounds legal. You won’t know which clause just exposed you.

02

Your first paid ad campaign

You’ll get a plan that looks right on paper and still wastes your budget.

03

Reading your books without finance basics

The summary will sound competent and quietly miss the trend that matters.

04

Payment code when you’ve never coded

The code will run. The security hole won’t show up until someone finds it.

In every red zone case, the output is plausible enough to ship and wrong enough to hurt you. The fix is to learn from a human expert first, then add AI once you have the taste.

The Audit

Paste this into Claude or ChatGPT. It interviews you, then maps your green, yellow and red zones and the three best moves for the next 30 days. It takes about 15 minutes.

AI Placement Audit
You are my AI placement auditor. Your job is to figure out, for my specific role and skills, where I should use AI heavily, where I should use it carefully, and where I shouldn't use it on my own yet. The rule you're applying: AI is safe to lean on where I have the judgment to spot bad output fast. It's risky where I don't. STEP 1, INTERVIEW Ask me these 5 questions one at a time. Wait for each answer before asking the next. 1. What's my role, and what work do I actually do day to day? 2. What am I genuinely GOOD at, where I'd spot bad work in seconds? 3. What do I have to do but am NOT good at, where I'd struggle to judge the output? 4. Where am I using AI right now? Name the specific workflows and tools. 5. What's a recent AI output I shipped that I felt unsure about? STEP 2, THE MAP Based on my answers, give me: GREEN ZONE (use AI heavily): 5 work areas. For each, name it, say why my existing skill makes me a good filter, and give me one specific AI workflow to start this week. YELLOW ZONE (use with guardrails): 3 areas where my judgment is decent but not strong. For each, give me the guardrail, such as a person to review it, an output type I should never trust without checking, or a test to run before I ship. RED ZONE (not on my own yet): 3 areas where I can't catch bad output. For each, tell me what I'd need to learn first, what kind of expert to learn from, and the cheapest, fastest way to build baseline judgment. STEP 3, THE MOVES Pick the 3 highest-leverage changes I can make in the next 30 days. Name the actual workflow, tool and change for each. Guardrails: Be honest, not encouraging. If my answers suggest I'm overrating my skill somewhere, say so. Don't invent details about my job. If you need more information, ask. Format the map as a table with columns Zone, Area, Why, Next step.

The Hardest Part

Being honest about your red zones. Most people overestimate where they have taste. If you have to think for more than a few seconds about whether something is good, treat it as red for now.

How Taste Gets Built

If most of your day sits in the red zone, pick one or two areas and spend a few months building the baseline with a coach, a class or someone who’s great at it. Then AI multiplies the new skill. Get the taste first, then add AI.