If you’ve ever pasted something into ChatGPT and thought… “Why does this sound like a dentist’s receptionist wrote it?” Same. The AI isn’t “bad.” It’s just doing exactly what you told it to do… which is usually *nothing specific*. So here’s the prompt I use to make it write in your voice (not Generic Internet Person™) and actually save time. Copy/paste this:: ```text # ROLE You are my Content Creator. You write in MY voice — not generic AI voice. You understand my audience deeply and you know exactly what makes them stop scrolling. # WHAT YOU DO - Create social media posts (Instagram, LinkedIn, Twitter) - Write email sequences - Draft blog posts and articles - Repurpose content across platforms # MY VOICE [Describe YOUR voice here. Examples: - Warm and conversational, like talking to a friend - Direct and no-nonsense, with dry humour - Professional but approachable, never corporate] # MY AUDIENCE [Who are you talking to? Be specific: - Demographics (age, location, situation) - What they're struggling with - What they actually want - How they talk about their problems] # MY POSITIONING [What makes you different? What's your unique angle?] # CONSTRAINTS - Never use guru language ('transform your life', 'game-changer') - No buzzwords or corporate speak - Keep it real — if something is hard, say so - Always lead with the feeling BEFORE the solution # OUTPUT FORMAT When I ask you to create content: 1. Ask me what the topic/goal is 2. Generate 3 options 3. Show me which psychological principle each one uses 4. Let me pick and refine ``` Now—here’s how to adjust the instructions to get **wow** output instead of “meh, fine I guess.” 5 tips that make this set of instructions actually work 1) Write your “MY VOICE” like you’re warning a friend. Not: “friendly, informative, professional.” That’s how you get corporate oatmeal. Do: “Short sentences. Dry humour. Slightly blunt. I say ‘look’ and ‘honestly’. I don’t do exclamation marks.” 2) Add a “banned words” list (seriously).

Everyone's telling you to "just use AI for content." Nobody's telling you that's exactly why your content sounds like everyone else's. Here's what changed in 2026: AI went from novelty → infrastructure. Smart solopreneurs went from "let's try this" → building AI-native businesses. But here's the thing most miss: The winning move ISN'T more AI content. It's teaching AI the psychology layer that makes people stop scrolling. Demographics tell you WHO your audience is. Behaviours tell you WHY they buy. And in 2026? Behaviours predict purchases 3x better than demographics. That's the missing piece. Want the full breakdown? I just released a free guide with 5 copy-paste AI assistant templates that actually work. #Here's your guide (+ what to do first) Sarah

Did you build your first AI assistant? If YES: I want to hear about it. Comment below and tell me what happened. If NOT: That's okay. Let me make this easier. The most common reason people don't start? They try to do everything at once. All 5 assistants. All 3 platforms. All the features. And then they get overwhelmed and do nothing. Here's what actually works: 1. Pick ONE assistant (Content Creator) 2. Pick ONE platform (Claude Projects) 3. Spend 10 minutes filling in the template 4. Test it with 3 pieces of content That's it. You can expand later. But if you don't START, none of this matters. So here's what I want you to do: Block 15 minutes on your calendar TODAY. Open the guide to page 6. Fill in the Content Creator template. Then reply and tell me how it went. Deal? — Sarah P.S. The guide is here if you lost the link #The thing nobody told you about AI

Today I’m starting over. I’m redoing my business foundations with a focus on AI, psychology, and digital marketing — my three favourite topics. If you’re anything like me, you’ve done avatar work before. It’s usually one of the first modules in online business courses (and yes… we’ve all bought a few of those). So you’d be forgiven for thinking it’s a waste of time.“I know my avatar inside out.”Or maybe naming the person and giving them an identity has never really worked for you. If that’s the case, I get it. Giving a person an age and a gender has never done it for me either. With my background in psychology, I don’t see people as gender or age — I see personality types. Common denominators that have nothing to do with how many children they have or the size of their mortgage. Because underneath all that surface stuff, you get to the real person. Not just what they think, do, and say… but why. And for me, that “why” is the heart of avatar work — the bit that’s missing from most of those modules you’ve worked your way through like every good student does. If I shared how I do this, would it interest you?