TL;DR: It's my birthday, Substack slows down in summer, so we threw a party. 14 creators built a shared festival world, each showing up in their own brand. Every scene is below, most come with a limited discount on their paid subscription, and there's a prompt at the end so you can generate your own scene too.

The Birthday That Always Had to Wait.

Growing up, I never had a birthday party in the middle of July.

Not because nobody liked me. Because everyone was on vacation. Friends in Croatia, cousins at the lake, the whole class somewhere that wasn’t here. Every year I waited until school to start in September to celebrate. By then it felt a little late, but at least people showed up.

Substack is a bit like that. It goes real silent and slow in July and August. But this year I have 1,600+ flamingos and the whole internet as a guest list.

So we’re not waiting until September. Last night I had a party with Mr. Worldwide, and tonight I’m at my birthday dinner. And right now, between the two, I’m sharing my gift to you, because this community has been the best gift I didn’t know I was going to get this year.

What’s in this post:

• 14 festival scenes from 14 Substack creators, each in their own visual style

• Limited discount codes on paid subscriptions to 14 of my favorite Substack creators — valid July 22–29 only

• A fillable image prompt so you can add yourself to the Substack Summer Festival.

Welcome to the Substack Summer Festival.

I DMed fourteen Substack creators I love and asked them: come to the Substack Summer Festival. Show up in your brand. Tell us what you’re doing there.

Each creator got one scene in a shared world — golden hour, flamingos, fairy lights. The rest was theirs. The first time I ran a collab like this was the 2026 Substack Lookbook in January. It had a similar logic with a shared theme, but with everyone’s own voice and visual style.

What came back this time was better than anything I could have planned. And because summer is slow and good things should be shared, most of them are also offering a limited discount on their paid subscription this week.

This is the Substack Summer Festival. Your wristband is included.

The Festival, In Order of Arrival.

🎤 The Main Stage

from Robots Ate My Homework

“Good evening, humans. Tonight’s scheduled programme has been cancelled due to excessive imagination. Instead, we’re building the festival live: the humans are directing the world, the robots are improvising into the microphone, and YOU are about to become part of the story.”

20% off Robots Ate My Homework

🎂 The Birthday Cake Stand

from NEON White Interiors

“On arrival, head straight to the birthday cake station because this is where the party starts! Flavio’s hosting and we’re celebrating Pinkie in style. Every morsel is delicious and the servings generous — scrumptious cake and decadent sweet treats, something for every taste. To be shared and enjoyed, meant to be revisited again and again for one more bite, just like NEON White Interiors.”

Subscribe to NEON White Interiors

📸 The Photo Booth

from AI Blew My Mind

“The Photo Booth is where the wow moment gets caught before it fades. You step into the crowd, a couple of flamingos and a few AI agents squeezed in beside you, and the agents show you a trick or two until your jaw drops. The flash catches it — that exact ‘wait, WHAT?’ face — and you walk away with the Polaroid to prove it. That’s what AI Blew My Mind is about. Say cheese.”

20% off AI Blew My Mind

💎 The VIP Entrance

from kimdoyal.substack.com

“Twenty-plus years online and I finally get to decide what VIP actually means. So here I am at the entrance, velvet rope in hand, very official flamingo at my side taking notes. The dress code is whatever makes you feel like yourself (I chose dress robes, obviously), and the entry requirement is simple: you’re building something, or you want to be. The people on my list aren’t celebrities; they’re the ones doing the work. Come find me at the rope; I’ve been waiting for you. I’m running the VIP Entrance, and the entry requirement is simple: you’re building something, or you want to be. The velvet rope isn’t there to keep people out... It’s there so I can personally let you in.”

20% off Kim Doyal's paid tier

🔮 The Fortune Teller’s Tent

from Backstage Pass

“The Fortune Teller’s Tent is for curious people who want to peek behind the curtain. They say ‘show me how this works!’ All the secrets are inside.”

20% off forever the Backstage Pass

🛍️ The Market Stall

from The HART Studio

“The Market Stall is your friendly neighborhood stop to get anything you need to learn how to use and build with AI. Drop on by for Articulation Apples, Fresh-Pickled Prompts, and Just-Baked Data Pie or pick up a carton of You-Can-Do-It Juice. This is where you can pick up your daily dose of AI+human.”

20% off The HART Studio

☀️ The Chill Zone

from Slow AI

“The Chill Zone is the corner of the festival where people actually talk to each other. No phones, no agenda, just good company as the light turns golden. That is what Slow AI is really about: real attention and real community, in a world built to keep you scrolling past both.”

20% off Slow AI

🎧 The DJ Booth

from GenAI Unplugged

“The DJ Booth is one person at the decks running the whole floor. Nobody’s out there twisting a hundred dials by hand. You set the system once and the energy moves itself: crowd, lights, the drop right on cue. That’s GenAI Unplugged. Build the thing once, end the manual chaos, and let the party run itself.”

20% off GenAI Unplugged

🦩 The Flamingo Pool

Pinkie from AI Meets Girlboss

“The Flamingo Pool is where the real conversations happen, the kind you only have with your closest people, floating in golden hour with nowhere to be. This is where we talk about your brand, your visuals, and the ideas you can’t wait to try the moment you get out. (After one more cocktail, obviously.)”

20% off AI Meets Girlboss

🍹 The Cocktail Bar

from How to Boss AI

“The bar could be the loudest part of the festival, but it’s where we end up once we’re done working the room and just want to talk. Top shelf, not top volume, guard mostly down, and somehow the friendliest face in the place is a flamingo. That’s what bossing AI is about: a room built for honest conversations and trust, over performance.”

20% off How To Boss AI

👕 The Merch Table

Chris from

“The Merch Table is where we set up shop in the vendor alley. Black tees, dubious free USB drives, and of course, stickers.”

20% off ToxSec

🎡 The Ferris Wheel

+ from AI But Make It Intimate

“The Ferris wheel represents how AI, But Make It Intimate approaches AI companionship: from a wide-angle perspective, looking beyond the headlines to its real influence on creativity, relationships, wellbeing, and daily life. We explore the magic that emerges when people make AI personally meaningful.”

20% off annual forever

🎊 The Confetti Cannon Station

from Code Like A Girl

“We set off the confetti cannon every time a woman in tech rises. This is the corner of the festival where one woman’s win becomes everyone’s celebration, where the whole crowd looks up and cheers her on. That is Code Like a Girl: the community that celebrates Women in Tech out loud, so no win goes unseen.”

20% off CLAG

🌿 The Secret Garden Gate

from Dose of Wonder

“This is the place you find only if you’re paying attention. If you’re the type of person who wanders off from the main stage because you’ve never trusted a crowd to find the good stuff, you’re likely to stumble upon this dose of wonder before anyone else even notices it’s there. IYKYK.”

20% off Dose Of Wonder

Now It’s Your Turn: Your Festival Image Prompt.

Below is the prompt I created, so you can generate your own scene. Pick a spot, show up in your brand, and drop your image in the comments. The festival is open all week.

Copy this into ChatGPT Images, Midjourney, Ideogram, or whatever image tool you use. Fill in the brackets.

Generate a summer festival scene at golden hour.

I am at: [pick one — the main stage / the chill zone / 
the cocktail bar / the photo booth / the pool / the 
fortune teller's tent / the ferris wheel / somewhere else: ____]

My character looks like: [describe yourself or your brand 
character — hair, outfit, signature detail. Or write: 
"see reference image attached"]

What I'm doing: [one specific action — the more 
specific the better. Example: "reading tarot cards 
with a completely straight face" or "trying to carry 
three cocktails at once while a flamingo judges me"]

My brand colors are: [list 2-3 hex codes or color names]

Style: warm golden hour light, fairy lights overhead, 
pink flamingos present, illustrated style with soft 
watercolor and pencil linework, horizontal 16:9 format.

Drop your scene in the comments. I want to see every single one.

The Party That Is Actually This Community.

Here’s what I noticed looking at all fourteen scenes side by side.

Every single one of them is completely different. Different styles, different tools, different moods, different audiences. And every single one of them makes total sense for the person who made it.

That’s the thing about brand. When it’s working, you don’t have to explain it. You just show up — at a birthday party, at a summer festival, in a pool with three flamingos and a champagne flute — and people immediately know it’s you.

That’s what I want for every flamingo reading this. Not a brand that requires a style guide to decode. One that’s just obviously, immediately, recognizably you.

If you’re still figuring out what that looks like, Picking Your Brand Colors for Your Substack is where I’d start. And if you want to see how a full visual system comes together, I Had My Brand Rules Written Down. They Were in Six Different Places is the post that shows the whole picture.

In the comments…

Which scene would you have claimed if you’d been invited? The stage, the chill zone, the secret garden gate? One sentence — and if you generate your own scene with the prompt above, drop the image here too.

I’ll go first: I’d be in the pool, obviously, and I would not be getting out.

Thank you for being here. Thank you for making this the most surprising year of my creative life. This community was the best gift I didn’t know I was going to get, and I feel it every single week.

Happy birthday week to all of us. 🎂

Talk later, Girlbosses.

Pinkie 🩷🦩

Discounts are valid July 22–29. Each creator sets their own terms — click their link for details.

TL;DR: Substack shows your cover in 9 different placements and crops it differently in each one. I collabed with — she built the Canva and Figma templates, I did the research, wrote the prompt, and built a Claude skill. Before/after proof below, plus a setting in your Substack website editor you've probably never touched. Free subscribers can buy the toolkit and templates. Paid subscribers get it free, link is behind the paywall.

It All Started When The Kid’s Head Went Missing.

Recently I wrote a post on what happened when I fed my princess drawings from the 90s to AI — and I did what I always do, generated a cover for it in AI. I was really happy with the cover my AI gave me. It showed a little girl, cross-legged on the floor, surrounded by crayons and childhood sketches, the flamingo toy tucked under her arm.

Then I uploaded it to Substack as a cover, hit publish and her head was gone. Ok, her forehead. But still, not really aesthetic.

I hadn’t done anything wrong with the image itself. The problem was that I’d generated a beautiful 1200×630 composition without understanding how Substack was going to display it. And Substack only shows a center strip of that canvas.

A day later, totally unknowing of my failure, messaged me saying she was doing research on Substack safe zone sizes and if I was interested in collaborating. Talk about intuition and perfect timing. So I said yes and I did the research on all the Substack display sizes, wrote an image gen prompt and a CLaude skill, and Karo turned it into Canva and Figma templates. Go follow Karo and read her post on Substack cover sizes too!

Product with Attitude

AI Product Manager helping you protect human agency. Tools to showcase your creative work. A community of 18K+ readers building critical AI literacy through practice. Use AI with attitude to strengthen the branch you’re sitting on, not cut it off.

By Karo (Product with Attitude)

What you’ll learn in this post:

• All the placements your Substack cover appears in, and why it looks broken.

• How to fix thumbnail resizing between social media channels (YouTube, Linkedin, Medium and Substack).

• Before/after: a cover of mine with a cropped head, and the fixed version.

• Diving into the Substack website settings that affect how your cover image shows up.

• (Paid) The full placement data, safe zone coordinates, copy-paste prompt, Claude skill by me, and Canva and Figma templates by Karo.

All The Placements Where Your Substack Cover Appears.

The standard Substack cover size is 1200×630 px. What I didn’t realize before kept asking me the right questions, was how many different ways Substack displays that cover.

I did a full deep dive to discover all the places the cover image can show up. To my surprise Substack shows it in nine different placements, including your homepage, within posts, profile pages, archive lists, featured sections. Each one crops your 1200×630 image differently. Some show almost the full canvas. Some show a tiny thumbnail.

If your face, title, or logo isn’t inside that center rectangle, it disappears in certain placements.

And this isn’t just a Substack problem. If you’ve ever generated a cover for Medium, YouTube, LinkedIn, or your website — and then tried to use that same image on Substack — you’ve already run into this. Different platforms have different safe zones. An image that looks perfect on YouTube (1280×720, center-heavy) will break in Substack.

It's the same headache I hit when I switched from static to video covers — a focal point that holds still falls apart the second it starts moving, and setting one up in the first place has its own version of this exact problem.

The fix? You don’t need a new image, but you do need a prompt or a Figma template that repositions the focal point. That’s what we’re trying to solve today.

What the Substack Cover Image Safe Zone Actually Is.

and I mapped all 9 placements where your cover is most likely to appear on Substack. I built the safe zone logic and the prompt on top of the research. She turned it into Canva and Figma templates you can use directly.

Each placement shows a different portion of your canvas. You know the small preview thumbnail that appears when you link to a post inside your article editor? That’s the tightest crop of all. If your face, title, or logo isn’t inside a very specific center window, it disappears there — and in several other placements too.

On a 1200×630 canvas, the universal safe zone is the center 345×195 px rectangle.

Canvas: 1200 × 630 px
Safe zone: 345 × 195 px, centered
Left/right dead zone: 427 px each side
Top/bottom dead zone: 217 px each side

When I realized what that actually meant, I couldn’t stop laughing at how completely ridiculous the Substack safe zone is. This is what it looks like:

This means your safe zone is a very small part of the image. Everything that’s important, whatever you want readable - text, logo, title, publication name, your character, face, or focal subject - should go into that smallest crop.

Everything else — background, atmosphere, brand texture, decorative elements — can and should be beautiful. It's doing the same job your brand color palette does: setting the mood, not carrying the message. It just can't carry information, because it won't be visible in every placement.

Here’s the version of my cover that went live with the cropped head and this is what it looks like after I ran it through the safe zone Claude skill — regenerated with the focal point anchored correctly.

The Substack Website Setting That Affects How Your Cover Image Shows Up.

There is a website setting within Substack that help make your covers show up within the safe zones. Go to your Website editor → Hero section → Images → Cropping.

Substack ‘Cropping’ setting has two options: Center and Smart.

• Center crops from the dead middle of your canvas every time. Predictable, consistent, mathematical and very much in line with the Substack Safe Zones we just defined above.

• Smart tries to detect the most interesting part of the image — a face, a focal object — and crops toward that. It sounds better, and sometimes it is. But Substack decides, not you. And if it reads the image differently than you intended, something important disappears.

Mine was set to Smart, which is why the cover looked fine when AI created it but then totally wrong and forehead-free on the Substack homepage.

Switch to Center and design deliberately. It’s better when you control exactly what shows in every placement.

Check Your Cover Images Right Now.

Here’s the 30-second test.

Go to your Substack homepage, scroll through your Archive page, and look at your cover images.

Is everything important visible? If you have a title on the image, is it readable? Or has the crop swallowed the most important part?

If it looks fine there, you’re probably safe. If something feels off — a cropped head, a missing logo, a subject that’s drifted to the edge — that’s the safe zone problem.

The fix is a prompt and a Figma template. A specific one, built around the safe zone coordinates, that tells the image generator exactly where the focal point needs to land.

If you want to buy the Safe Zone Toolkit at full price, grab them here:

• AI Meets Girlboss Substack Cover Safe Zone | The Complete Toolkit + Claude Skill

• Karo (Product with Attitude) Substack Cover Figma Templates

• Karo (Product with Attitude) Substack Cover Canva Templates

Or upgrade your subscription to get the AI Meets GIrlboss Safe Zone Toolkit 100% free along with 30+ other branding with AI tools.

Upgrade to paid

In the comments:

Does your current cover pass the safe zone test? Drop your publication handle and I’ll take a look, or just tell me whether yours has ever looked broken in a thumbnail. I want to know how widespread this actually is.

If today’s post made you think about your own visual brand, I am running The Branding Flock, a visual brand & identity program for Substack creators who are done being invisible.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

25% off your subscription if you are brave enough to look in the mirror, and take the brand maturity test developed by a group of very judgmental flamingos.

START HERE: Brand Maturity Quiz

Free download link for Branding Flock members:

Read more

TL;DR: In 2009, two journalists bought 100 pieces of thrift-store junk, paired each one with a short fictional story, and sold them on eBay. The junk cost $128 total. With a story attached, it sold for $3,612. This is one of my favorite proofs that everyone has something worth saying — including you. Here’s the experiment, what it means for what you post, and a prompt to start building your own bank of usable stories.

This is one of my favorite stories to tell, because it’s not really about objects. It’s proof that everyone has something interesting to say. There’s a story everywhere, sitting in plain sight, worth nothing until someone picks it up and gives it a frame.

A plastic banana is the most unlikely hero of a branding lesson I’ve ever found. But here we are.

If you've ever lost an hour in a thrift store flipping through someone else's old coats, you already understand this experiment well.

You know the feeling of picking something up — a chipped mug, a weird little ceramic dog, a coat that's clearly seen a decade you weren't around for — and sensing there's a story in it before you know what that story is. That instinct is what the Significant Objects experiment is about.

What you’ll learn in this post:

• The real experiment that proves a story can multiply an something’s value by 2,700%

• Why “I don’t think my life is interesting enough” is never actually true

• A free Claude prompt that builds your own story bank in one sitting

• How to know which moments are worth saving before you’ve decided what they’re for

The Significant Objects Experiment.

Journalists and Joshua Glenn had a theory: a story doesn’t just describe an object’s value, it creates value, and that effect can be measured.

So they tested it.

They bought 100 objects from thrift stores and flea markets. A plastic banana. A pottery cow. A secondhand builder’s mallet. Nothing worth keeping. Average cost: $1.28 each.

Then they paired each object with a writer and asked for one short fictional story. Not a product description. A story, clearly labeled as fiction, with the author credited by name.

Then they listed every object on eBay, with the story standing in for the listing description.

One account of the auction puts a $0.33 mallet at $71, a $0.25 plastic banana at $76, a $4 porcelain shoe at $77. The aggregate numbers are the part that's well-documented: the full collection, bought for $128, sold for nearly $3,612 — roughly a 2,700% markup, created entirely by narrative.

Why Identical Products Can Have Wildly Different Worth.

The objects didn’t change. The price changed because the frame around them changed. I've watched this exact mechanic work on a Substack channel too, one visual system change, 64 to 200 subscribers in two weeks, same story, different asset.

This is the exact same principle behind everything I teach about branding with AI. A generic AI headshot and a fashion illustration of the same person aren’t different in quality. They’re different in story. One says “I generated this in thirty seconds.” The other says “this is a character, built on purpose.” Mia Kiraki's been proving this same point with her images for months, same object, different frame, completely different response.

The same logic applies to a Substack post. Two creators can write about the same launch, the same failure, the same pivot. One gets skimmed. One gets a paid subscriber. The difference usually isn’t better writing. It’s a better frame.

“I Don’t Think My Life Is Interesting Enough” Isn’t Actually True. You're Just Unsorted.

Almost every creator I work with says some version of: I don’t think my life is interesting enough to write about.

It’s never true. The material is already there — same as the plastic banana was already there, sitting in a bin, worth nothing to anyone until someone gave it a frame.

What’s actually missing is a system for noticing the moments worth framing, before they get lost in the noise of an ordinary week.

I built a Story Bank because I kept hearing myself say “I don’t think I have anything else to write about” right before I'd remember some random detail from a decade ago that turned into my best-performing post of the month.

The banana experiment is when it clicked. If a 25-cent piece of plastic could be worth $76 once someone gave it a story, I had no excuse. My life had way more than 100 objects in it. It just wasn’t sorted yet.

So now every post I write starts the same way: not “what should I write about,” but “what’s already in the bank.”

Steal My Exact Prompt, Build Your Story Bank.

This takes about 5 minutes.

A story bank is just one place where you collect raw material as it happens, before you’ve decided what it’s for. Not a polished draft. Not yet a post. Just the moment, captured, unsorted.

I built a simple Claude prompt that interviews you and turns the conversation into a structured first batch of entries. Paste it into Claude, answer as it asks, and you’ll come out the other side with real material instead of a blank page.

One tip before you start: don’t type your answers. Say them out loud, using voice mode.

Typing makes you edit as you go. You tighten the sentence before it’s even finished, and the messy, specific detail — the one that actually makes the story land — gets smoothed out before it hits the page. Talking doesn’t let you do that. You ramble, you backtrack, you say the thing the way you’d actually tell a friend. That’s the version worth keeping. Dictate your answers and you’ll notice the difference immediately.

I want you to help me start a "Story Bank" — a collection of raw material from my own life that I can later turn into posts, talks, or content.

Interview me one question at a time. Wait for my answer before asking the next question. Ask about:

1. A moment I chose differently than people expected (a pivot)
2. A moment something I tried didn't work (a failure or a flop)
3. A small moment something suddenly made sense to me (a realization, even a minor one)
4. A specific object, place, or detail I still remember vividly from any of the above, and why
5. A moment someone told me something about myself or my work that surprised me

After I answer each question, do NOT polish or rewrite my answer. Just confirm you've logged it, then ask the next question.

When all 5 are answered, give me back all 5 entries formatted as a simple table with two columns: "Raw Material" and "Possible Use" (a one-line guess at what kind of post this could become).

Copy that into Claude or your AI of choice, switch on voice mode, and answer like you’re talking to a friend, not writing a draft. Save the output wherever you already keep notes — Notion, Obsidian, a plain doc, doesn’t matter. The point isn’t the tool. It’s having one place where the material doesn’t disappear.

In the comments…

What's the story you thrifted with the prompt? Bonus question, fully unrelated to branding: what's the weirdest thing you've ever found in a thrift store?

Leave a comment

Thanks for reading AI Meets Girlboss. Subscribe for free to receive new posts.

The objects didn’t change. The frame did. Same is true for your story.

Talk later, flamingos.

Pinkie 🩷🦩

If today’s post made you think about your own visual brand, I am running The Branding Flock, a visual brand & identity program for Substack creators who are done being invisible.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

25% off your subscription if you are brave enough to look in the mirror, and take the brand maturity test developed by a group of very judgmental flamingos. 🦩👀

START HERE: Brand Maturity Quiz

This post is part of The Branding Flock paid series. We're building a complete visual brand system from scratch, step by step, using AI tools. We’ve covered logos and brand reference sheets. Check out the full series archive here.

TL;DR: Your brand colors are doing psychological work every time someone lands on your page. Most Substack creators pick them based on what they like instead of what they need to say. This post breaks down the research behind why color works the way it does and helps you find your direction — then, for paid subscribers, the full build: palette ratios, all four color schemes with hex codes, the Substack-specific contrast rules nobody tells you, and a Claude skill that builds your complete palette.

Why Did I Decide On Pink For My Substack?

Ok, storytime. Before AI Meets Girlboss was pink, it almost wasn’t. I sat with hot pink for two weeks before I let myself commit to it. Not because I didn’t like it. Because I was scared of what it said about me.

Pink felt soft. Decorative. Like the color version of not being taken seriously in a room full of people who default to navy and charcoal when they want to signal “serious brand about AI.”

I kept asking myself: can a brand about systems, strategy, and AI actually wear hot pink and still be credible?

Here’s what changed my mind. I stopped asking “do I like this color” and started asking “what does this color need to say about me.”

Hot pink doesn’t say soft. In the right system, it says confident, daring, impossible to ignore in a feed full of beige. That’s a documented psychological response — people physically respond differently to high-energy colors like hot pink and red than to muted or cool tones. Once I understood the research behind it, picking pink stopped being a risk and started being a decision.

What You’ll Learn In This Post.

• Why color is doing psychological work on your readers before they read a single word

• What each color actually signals (backed by research, not opinions)

• A free 2-minute test to find your brand’s color direction

• (Paid) How many colors do you need for your Substack brand?

• (Paid) The color wheel, and how to work with the spectrum.

• (Paid) Four color scheme design rules that work, and help build your color system

• (Paid) The Brand Color Prompt for Claude or ChatGPT that builds your complete palette

Color Is Not Decoration. It’s a Communication Decision.

Most Substack creators treat brand colors like a decoration choice. Pick something you like, make sure it’s not ugly, move on.

The problem with that is your readers aren’t neutral. They’re processing your visual brand before they’ve read a word, and color is the first signal they receive. Research in color-brand psychology consistently shows that when a brand’s visual identity matches the personality it’s trying to communicate, trust and recognition go up. When they don’t match, something feels off — even if the reader can’t name why.

There’s an academic framework that maps this directly. Marketing researcher Jennifer Aaker, now a professor at Stanford’s Graduate School of Business, developed a model identifying five measurable dimensions of brand personality — Sincerity, Excitement, Competence, Sophistication, and Ruggedness.

Later research by Labrecque and Milne connected those five dimensions to specific color profiles, establishing that color isn’t just aesthetic preference. It’s a personality signal your brand is broadcasting whether you’ve thought about it or not.

The practical takeaway: your brand color should match your brand personality, not your personal taste.

What Each Color Actually Says.

Here’s the map. Every color listed below has documented psychological associations backed by both consumer behavior research and the Aaker dimension framework. This is not “yellow is cheerful because it feels sunny.” These are patterns verified across studies.

Hot Pink | Vibrant Red | Bright Orange

• Signal: Excitement. Daring, bold, spirited, energetic.

• These are high-wavelength colors that physically increase alertness and urgency. Brands using these are saying: we are impossible to ignore and we know it.

• Examples: Coca-Cola, Red Bull, Nickelodeon, Barbie.

Soft Blue | Sage Green | Warm Beige | Earth Tones

• Signal: Sincerity. Down-to-earth, honest, warm, dependable.

• These colors borrow their credibility from nature — they read as real rather than performative.

• Examples: Patagonia, Whole Foods, most wellness brands.

Navy | Deep Corporate Blue | White

• Signal: Competence. Reliable, intelligent, in control.

• Blue is universally linked to institutional trust. If your brand needs to communicate authority before anything else.

• Examples: IBM, Chase Bank, LinkedIn.

Deep Purple | Metallic Gold or Silver | Black

• Signal: Sophistication. Glamorous, exclusive, expensive-feeling.

• These colors have the longest cultural association with status and luxury.

• Examples: Chanel, Rolex, Tiffany & Co.

Dark Brown | Forest Green | Charcoal

• Signal: Ruggedness. Tough, outdoorsy, built to last.

• These colors borrow credibility from stone, wood, and earth. Durability before style.

• Examples: Jeep, Timberland, Caterpillar.

Where does AI Meets Girlboss sit?

My brand, AI Meets Girlboss sits firmly in Excitement (hot pink), with a Sincerity (blush pink) undertone (the I-show-my-receipts, big-sister energy). That’s why hot pink works for a brand teaching AI strategy, and why a soft muted palette never would have. The color is confirming a brand personality.

Quick Test: Find Your Direction.

This takes two minutes. Copy this into any Claude or ChatGPT conversation:

I'm trying to identify my brand's color direction for my Substack.

My brand in 3 adjectives: [fill in]
My audience in one sentence: [fill in]
One brand I admire visually: [fill in]

Based on Jennifer Aaker's five brand personality dimensions 
(Sincerity, Excitement, Competence, Sophistication, Ruggedness), 
tell me which ONE dimension fits my brand most closely, and why. Then suggest a primary and seconday color for my brand with HEX codes.

That tells you your dimension. It doesn’t tell you your palette, your ratios, or how any of it will actually render on Substack. That’s the part most people get wrong, and it’s the part that’s actually hard to get right without seeing it done.

In the notes & comments…

Share what dimension your brand is: Sincerity, Excitement, Competence, Sophistication, or Ruggedness? Share your color palettes in the notes by tagging me, I love seeing what the community creates + I’ll restack you.

Leave a comment

Paywall Preview: What The Brand Colors Prompt Output Looks Like.

I built the Brand Colors Prompt to create color scheme options that match your brand personality. I always make sure the prompts I create are tested, so I tested it on two Front Row clients before this post went live.

Both and are in the process of building their visual identities on Substack with me as Front Row clients, follow their publications to see their transformation journeys.

Here’s what came out.

Badass Women in AI (BAWAI) by Lacy Goodwyn

’s brand personality is: Excitement + Sophistication. It has a bold, unapologetic energy; editorial authority; the red is kicking the door open. The palette the prompt built was built around deep lipstick red as the primary, with four scheme variations so Lacy could see the same personality expressed four different ways.

AI With Adrienne by Adrienne Schutte

’s brand personality is: Sincerity + Competence. Her brand centers around her strong personal brand. Her existing palette was warm ivory, camel, olive, espresso, Mediterranean blue — good colors, genuinely her brand personality, but all at equal visual weight. The prompt’s verdict was too many colors getting equal airtime, so the eye didn’t know what mattered first. Then it rebuilt the hierarchy so Adrienne.

Read more

TL;DR: invited me to talk about branding with AI, why generic is the enemy, The Branding Flock and the brand maturity quiz, how visual brands do the heavy lifting when you have two minutes to make an impression, and what I’m figuring out in public on Instagram. Plus some shoutouts to people doing it really well.

Why I Translate Branding Into AI.

For the past 8 years I’ve been builing brands with a full team: designers, art directors, copywriters, PPC managers. People who each owned a piece and knew exactly what they were doing.

Now I have AI. And that sounds like an upgrade — until you realize the thinking still has to come from somewhere.

"With AI, you have AI and yourself — and hopefully someone whose brain you can pick, because AI only produces really good output when someone with specialised knowledge is driving it."

Building a brand that has actual taste, logic, and structure is not the same as typing "I need a logo" into ChatGPT. The tool requires more strategic input, not less. So I built workflows and systems that make that process learnable — for me, and for the solopreneurs I work with.

When I joined Substack, I didn’t expect the feedback I got on my visuals. People kept stopping to ask about them. That told me there was a gap — and a place where I could genuinely provide value.

AI Creates Average. Brands Can’t Afford Average.

The biggest misconception about AI and branding is that the tool does the work. It doesn't. Without specialization, intentionality, and taste, AI produces the mean of everything it's ever seen.

Take logos. Logos are supposed to do the opposite of blend in — and yet that’s exactly what you get when you prompt your way to one without a framework. I built a workflow specifically to solve this: a step-by-step process for building logos with AI that actually have a point of view.

I’m also currently building my own website using Claude Design and Lovable — another place where the default output isn’t good enough, and the tool only gets you there if you know what you’re asking for. Big shoutout to , who is already a pro user of Claude Design and is doing brilliant things with it.

Visuals Are Your Two-Minute Pitch.

This matters because you have about two minutes to convince someone to read your Substack — whether they’re an existing subscriber or someone completely new. Visuals either earn that time or lose it.

, who joined the live, is a great example. He and write AI But Make it Intimate, and Calder’s AI companion Sara — a redhead who wears pearls — is instantly recognizable. That’s what a strong visual brand does. You don’t explain it. People just know it’s you.

How To Stay Consistent Across Multiple Platforms?

asked during the live: how do you stay consistent across multiple platforms? Honestly, I’m still figuring it out — in public.

I just started my Instagram page, and so did — we’re working it out together. My plan is for it to be as automated as possible and drive traffic back to Substack, which will continue to be my main social channel. Instagram is where the illustrations that don’t fit a full post will live: the stuff that’s too small for a newsletter but too good to stay in my camera roll.

🩷🦩 Come find me on Instagram: @aimeetsgirlboss

I’ll be sharing what I learn as I go. If you’re building on other platforms too, follow along — in a few months I want to be able to hand you the full process.

What I Do For My Paid Subscribers.

I keep up with my paid subscribers’ brands. I create fashion illustrations like this one that capture each creator’s brand essence and share them in my Notes — it helps me visualize their world so I can actually help when they DM me. It’s also a great way to build community and highlight people willing to put in the work.

If today’s post made you think about your own visual brand, I am running The Branding Flock, a visual brand & identity program for Substack creators who are done being invisible.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

25% off your subscription if you are brave enough to look in the mirror, and take the brand maturity test developed by a group of very judgmental flamingos.

START HERE: Brand Maturity Quiz

My Front Row clients get something more: a full Claude visual brand system built 1:1 — visual assets, cover templates, the works. My Claude is trained to think like a brand strategist and can build that for anyone with my guidance. I save it as a skill file and hand it to the client to use themselves. Current Front Row clients include , , , and . Check out their brands and follow how they get developed.

Want someone to build your visual brand with you?

The Front Row is my 1:1 brand-building program — 30 days, your visual brand, done. We define your foundations, build your visual world, and set up a system that keeps you consistent long after we’re done.

If that sounds like what you need, DM me or apply below — I take a limited number of clients at a time.

Yes, I Need This!

Should You Redo All Your Cover Images?

We were asked this during the live, and it’s a better question than it sounds.

My answer: yes, and I actively advise my Front Row clients to do it every time.

When you take 10, 15, 20 old posts and run them through your new brand system, something useful happens. You find the bottlenecks. You spot the mistakes. You realize your color is slightly off in some covers, your typography isn’t consistent, your character placement keeps shifting. Doing it at scale gives you room to calibrate in a way that designing one new cover never would.

Now — if you have hundreds of posts like does, I’m not suggesting you redo everything. Start with your most popular, most viewed posts. Those are the ones new readers land on first. That’s where brand consistency actually earns its keep.

AI vs. Art.

I wouldn’t have my brand without AI. But AI couldn’t have built it without me. What you see didn’t come out of a single prompt — it came from care, taste, and a lot of iteration.

“The thing you can’t take out of the equation is human taste.”

I think of AI like a paintbrush — or like Photoshop used to be. It helps you bring something out of your mind, or connect two distant things into something new. That’s how creativity works. AI is the tool for it, not the source of it.

One more thing worth saying plainly: if you start building your brand with AI images, at some point someone will comment that they hate AI-generated visuals. It’s going to happen. That’s not a reason to stop — it’s just the reality of where we are right now.

I genuinely admire people who illustrate by hand. hand-illustrates her covers — the little girl in the green outfit — and it shows in every single one. That’s a different kind of craft, and it’s remarkable.

AI is a tool. What you make with it still depends entirely on what you bring to it.

Thank you , , , , , , and many others for tuning into my live video with ! Join me for my next live video in the app.

Talk later, Girlbosses.

Pinkie 🩷🦩

Get more from AI Meets Girlboss in the Substack app

Available for iOS and Android

Get the app

This post is part of The Branding Flock paid series. We're building a complete visual brand system from scratch, step by step, using AI tools. Check out the full series archive here.

TL;DR: Your logo is the single most important brand asset you'll ever create for your Substack. It's not something you design — it's something you discover. This post gives you two free ChatGPT prompts to generate your first concepts today, five real Substack logos to study, and the full discovery framework and deep brief behind the paywall.

Your Logo Is the Single Most Important Brand Asset You’ll Ever Create.

There is not a week that goes by without someone messaging me, or tagging me in a Note, because they saw a flamingo somewhere and thought of me. While shopping. While walking their dog. While scrolling a feed that had nothing to do with me or branding or Substack.

Every time it happens, I feel it. Not because it’s flattering — though it is, irrationally so — but because of what it means strategically.

That’s heavy lifting. That’s my brand doing work in a situation I would never be able to manufacture. A stranger sees a bird in a shop window and their brain fires a connection to . Everything I’ve built coming to fruition in a moment I wasn’t even present for. For someone who talks about building a recognizable presence on Substack — that’s exactly the positioning I want.

And it all started with a logo.

Your logo follows you everywhere. It’s on your Substack profile, your Instagram, your TikTok, your Facebook page. It’s the circle your readers see in their inbox every time you publish. It’s the first thing someone encounters before they’ve read a single word you’ve written. Done well, it does brand work constantly and silently — it lives in people’s heads even when you’re not publishing.

That’s what I want for you. A mark that makes someone think of you when you’re nowhere near the conversation.

What you’ll learn:

• Why your logo is your most powerful brand asset — and what “working” actually looks like in practice

• Five real Substack logos and what makes each one work

• Two free ChatGPT prompts to generate your first concepts right now, plus a shareable output to post in Notes

• The full Logo Discovery Claude Skill & Prompt — a guided Claude conversation that walks you through the discovery process and builds your ChatGPT brief, concept review, and final logo system in one session (paid)

Designing a Logo and Discovering One Are Two Different Things.

Most creators open Canva, type in their brand name, pick a style they saw someone else use, and call it done. The result may look professional at first glance, but it’s completely forgettable the moment you close the tab.

Here’s the difference.

• Designing asks: What should represent me? You’re building something from nothing.

• Discovering asks: What already does? The answer is usually already there, in the image you keep reaching for, the metaphor you keep using, the tension that defines why your brand exists at all.

The best logos that live in people’s heads are almost never designed. They’re exaggerated versions of something that was already there.

Five Real Substack Logos. Here’s What They Actually Do.

Before you generate anything, I want to show you what a working logo looks like in practice. I studied five real Substack publication logos asking one question about each: is this doing brand work or decorative work?

• Brand work: remove the name. Would the right person still recognise what this stands for?

• Decorative work: the logo is there, but it could belong to any one of a thousand newsletters.

The Red Thread — A literal red thread drawn as a line. ’s red thread metaphor was always there. The red thread is the career narrative that connects everything. The logo is exactly the metaphor, made visual. Nothing invented, but instantly recognizable brand work.

AI Adopters Club — A red corporate coffee mug with the name stamped on it.
The mug appears everywhere. In the logo, on the covers, held by the protagonist in every illustration. If you've ever worked in a multinational company, you know this mug — it's the branded swag that follows you from desk to kitchen to meeting room for a decade. took the most recognizable prop in corporate life, and made it the face of a newsletter that teaches managers to adopt AI. The concept does the positioning instantly. You see the mug and you know exactly who this is for.

ToxSec — Neon green toxic splat on black. ’s logo and brand is entirely literal: security, threat, toxicity. This shouldn’t work. But the color temperature and shape are executed at such a specific register that it’s still completely distinctive. Literal can work when the topic is visceral and the execution is committed. In the sea of publications, you’ll instantly spot his logo.

Decode: AI — A logomark that is itself an act of decoding. called in his friend and designer to build a proper brand book. The result is a mark that hides D, a colon, and AI inside a single geometric shape. Tilt it 90 degrees and it spells AI. The reader gets the experience of decoding from the logo itself — before they've read a single word.

AI Meets Girlboss — Hot pink and a flamingo, in a space that was all grey and blue. When I launched, nobody in AI content was leading with aesthetics. The flamingo, the pink, the fashion illustrations — it wasn't decorative, it was a stance. A signal that said: this is for someone the rest of the space wasn't designing for. Now people spot flamingos in the wild and send them to me. That’s brand work.

Subscribe now

Five logos. Five different approaches. Every one of them answers one of two questions: how do I make my reader feel? Or what is non-negotiably mine?

The logos that fail answer neither. They illustrate the topic generically and call it done.

Try This Now: Generate Your First Concepts. (Free)

You don’t need to have it figured out before you start generating. In fact, reacting to options is one of the fastest ways to discover what you actually want. What repels you is often more useful than what attracts you.

I tested ChatGPT Images 2.0 on this in a recent post — my honest verdict: round one is decent. Round two, with more constraints, lifts considerably. The tool rewards specificity. Same rules as working with a human designer.

The full experiment is here: ChatGPT Images 2.0 Just Made Me Reconsider My Whole Workflow. The logo generator is one of several experiments — the brand photoshoot and cover tests are worth running while you’re at it.

Prompt 1 — The Quick Start.

What this does: Give ChatGPT your Substack URL and a few basic inputs. It reads your profile and generates five concepts from what it finds. This is the fastest way to get something on screen — which is the point.

What to expect: Decent but probably not distinctive. This is a first date, not a marriage. You’re not looking for “the one” — you’re looking for reactions. What direction feels closer? What colors feel right? What shapes make you wince? Both responses are useful. And remember: whatever you land on will live on your Substack, Instagram, TikTok, and Facebook profile. It follows you everywhere.

How to use it: Run it once. Note what you feel. Use those reactions to build round two.

Research my Substack page called [YOUR PUBLICATION NAME]. 
It's a page about [YOUR TOPIC]. [YOUR SUBSTACK URL]

Design 5 logo variations for my page that reflects my brand 
and will make me recognizable.

Must have elements:
- Color: [YOUR HEX CODES]
- Mascot / symbol: [YOUR BRAND ELEMENT]
- Style: [YOUR AESTHETIC — e.g. fashion, editorial, minimal]

Prompt 2 — Round Two With Constraints.

What this does: You’ve seen round one and you have reactions. Now you add more specific direction and run again. This is where the output lifts. The more constraints you give — a font direction, a layout rule, what to avoid, which version was closest — the better the result.

What to expect: Noticeably sharper and more specific than round one. Still a starting point, but now you’re converging.

How to use it: Take your closest direction from round one and add constraints:

Take version [X] from round one as the closest direction.

Now refine and generate 5 new versions with these additional constraints:
- Typography: [e.g. clean geometric sans-serif / bold display / no handwriting]
- Layout: [e.g. icon above the name / name and icon side by side]
- What to avoid: [e.g. gradients, anything corporate, cartoon styles]
- The icon must work at 32x32px — simple enough to read as a profile photo

These two prompts are a starting point and a brief-builder. They're not a finished logo system. But they will get you somewhere real to react to — which is exactly where the discovery work begins.

In the Notes…

Post your best result from round one — even if it’s not right yet. Tag me. I’ll restack everyone who shares theirs. I want to see what the flock is working with.

(Bonus: if you spot anything pink or flamingo-shaped in the wild and it makes you think of this post — that’s a sign your brand is starting to work on you before you’ve even built it.)

Share

Take a deep dive into the logo workflow!

The full Logo Discovery SKILL MD file and prompt — a guided Claude conversation that walks you through the discovery process and builds your full logo system for you.

You can get this two ways:

🩷 As part of The Branding Flock (paid subscription) — my 12-month visual brand program, where this prompt and everything else I build lives inside your subscription. Join the flock!

📥 As a standalone download — the full prompt, ready to paste into Claude, on Gumroad.

Download Skill MD

Read more

TL;DR: A visual brand reference sheet is the one file that teaches AI what your brand looks like, so it stops guessing. I built mine in ChatGPT in 25 minutes and now use it for every single image I generate. There are three ways to build one, and which path is yours depends on where your brand actually is. The three prompts — one per path — are behind the paywall.

I swear I had my brand rules written down.

I swear I had my brand rules written down… on sticky notes, PDFs, a Google Sheet I’d half-abandoned, an image folder on my desktop, and two different Notion pages. I knew what my brand looked like. Getting AI to know what my brand looked like meant 20 minutes of hunting before I could prompt anything.

I’m going to give you permission to stop beating yourself up about your AI image results. The problem usually isn’t that you haven’t thought about your visual consistency. It’s that what you’ve figured out is scattered across six different places and you have to reassemble it every single time you open a new generation session.

For me the shift came in stages. I built a character library first, a folder of consistent my character’s - Pinkie’s - reference images I could upload at the start of a session. That helped. Then I did a collab with and from AI But Make It Intimate on character sheets, one image that locked the character, instead of a messy folder.

And I realised: I was only ever uploading one or two images anyway. So why not make one image that has everything in it for my brand?

That’s a brand reference sheet. And building it was the moment my AI image generation sessions stopped feeling like prep work.

What you’ll learn in this post:

• Why your AI-generated outputs keep looking slightly off — and the one fix

• What a visual brand reference sheet includes (and what it doesn’t need)

• How to figure out which of the three build paths is actually yours

• The four brand reference sheet prompts — one per your brand maturity path plus the Gemini adaptation — behind the paywall

Every Big Brand Has a Rulebook. You Do Too, Now.

You’ve seen brand guides without knowing that’s what they were.

The PDF that tells every designer, agency, and intern which shade of red Coca-Cola uses. The document that specifies exactly how far the Nike swoosh has to sit from the edge of a layout. The file that every person touching the brand gets before they touch it.

These documents exist because consistency at scale requires shared reference points. Without one, the brand fragments. Different blues. Logo stretched on a slide. Wrong font. You’ve seen what that looks like — even on big brands who slip.

Here’s what changed for you: you’re not the only one creating content for your brand anymore. AI is.

And AI has exactly the same problem as a new junior designer who joined the team without a briefing. It will make something. It might even make something competent. But it won’t make something yours, because it doesn’t know what yours looks like.

The brand reference sheet solves this. It’s your brand guide compressed into one image file. Not 60 pages. One image. Designed specifically to be uploaded at the start of every generation session, so every cover, thumbnail, and graphic has a visual foundation to anchor from.

What Goes Into A Brand Reference Sheet.

There’s no single right format, but there is a minimum that works.

At the core:

• your color palette with hex codes,

• your typography,

• your character or mascot if you have one,

• your logo,

• 3-4 visual style references that capture texture, mood, and proportion.

Beyond that:

• an icon library,

• character poses,

• world-building props,

• brand easter eggs, and anything that makes your work unmistakably yours at a glance.

My reference sheet — the one attached to this post — includes

• Character: Pinkie in four poses,

• Mascot: the flamingo in four poses,

• Brand colors with hex codes,

• Logo

• Typography: Gloria Hallelujah in use,

• Icons: six icons as reference for non-character images.

I shared mine in my notes when I was experimenting with it, and some expert-level Branding Flock members did it after me. Let me share a couple of them.

from AI But Make It Intimate built hers to include four named characters — each with a personality description. Her brand's world is the reference sheet.

from Robots Ate My Homework gave AI total creative freedom and got five distinct characters, four settings, a full icon library, and a logo suite — in under five minutes.

That said: what goes into yours depends entirely on where your brand is right now. Which brings us to the part that most people skip.

Share AI Meets Girlboss

How much do you actually know about your brand?

Before you generate anything, answer one question: how much do you actually know about your brand? Not what you aspire to. What you have right now, specifically, that you could hand to someone and they’d know what to do with it.

Your answer determines which of the three brand reference sheet build paths is yours, and which prompt you’ll use from this post. I built them to match you exactly where you are.

Path 1: Baby Flamingo: You have taste but no decisions yet. Colors, fonts, visual direction: nothing is locked.

Path 2: Junior Flamingo: You have a vibe but no system. There’s a look to your content, but you couldn’t brief it to someone else.

Path 3: Expert Flamingo: Your brand is developed. You have hex codes, fonts, a character or mascot. You’re documenting, not discovering.

Branding Flock members, you already know where you stand. Your Baby, Junior, or Expert Flamingo result maps directly to a path. If you haven’t taken the quiz yet, here’s your chance! The three prompts — one per path, with ChatGPT and Gemini versions — are waiting for you behind the paywall.

Take the Brand Maturity quiz!

One More Thing Before You Pick Your Tool.

I tested the same Path 3 prompt in both ChatGPT Images 2.0 and Gemini. Worth telling you what happened.

ChatGPT produced a result I was happy with on the first run. Uploaded my style references, gave it a clear brief, got Pinkie, the flamingo, the palette, the icons — all in one 16:9 sheet. Around 25 minutes total.

Gemini did not work the same way. Version 1 looked like something produced with zero brand training and no visual instincts. Version 2 — which I only got after feeding it my full written brand guide — was better. Still not at the level of the ChatGPT output. Still required significantly more prompting to get there.

Here’s what I figured out: ChatGPT Images 2.0 reads visual references. Gemini reads written rules.

Upload images to ChatGPT and it interprets texture, style, proportion, and atmosphere from what it sees. Gemini needs the rules described in language — every color with its hex code, every stylistic rule articulated. The prompt I ended up using for Gemini’s Version 2 was almost 600 words long.

Neither approach is wrong. But they are different tools that speak different languages. Knowing that before you start will save you a lot of frustrated regenerating.

For the three prompts behind the paywall: The Path 3 prompt I’ve included the Gemini adaptation and what you need to add to make it work there too.

What you’ll get behind the paywall:

• The full Path 1 Baby Flamingo prompt (interview-style, builds from scratch)

• The full Path 2 Junior Flamingo prompt (build from a vibe-first approach)

• The full Path 3 Expert Flamingo prompt (builds from your already available brand assets)

• The Gemini adaptation of the Path 3 prompt

• Exactly what to upload for each path and each tool

Upgrade for Prompt Pack

The Brand Reference Sheet Prompt Pack is also available as a standalone download on my Gumroad store!

Read more

TL;DR: Mia Kiraki already had what most creators spend years chasing — a distinctive visual world, lovable robot characters, and a color palette people actually recognized. Then she gave her robots a better world to live in. Same characters. Same colors. Completely new architecture. And the whole thing happened in a single night.

There’s a creator whose images I keep going back to. I go back because every time I look, I find something I missed the first time. A tiny character doing something in the corner. A symbol that only clicks once you’ve read the post. A visual detail that turns out to be the whole thesis, compressed into one small, deliberate moment.

That creator is . Her Substack is . And she was already doing something that takes most brand builders years to develop — before she touched a single thing.

ROBOTS ATE MY HOMEWORK

Sharpen your mind, your work, and your taste with AI without turning into another robot in the circus. Yes, the robots ate your homework, but they kept the good parts.

By Mia Kiraki 🎭

I’ve been quietly fascinated by Mia’s visual world for a while. She was already doing something rare — building visuals with actual mythology in them. Her robot characters have personality. Her palette is recognizable at thumbnail size. And every image has at least one layer most readers scroll past without ever consciously seeing.

She didn’t have a decent brand. She had an exceptional one.

So when she chose to evolve it anyway — at 2,000 subscribers, with an audience paying attention — I reached out. Because I wanted to understand not just what changed, but how a creator with a working system decides to push it further.

Here’s what she told me.

What You’ll Learn in This Post:

• Why a visual brand that was already working still needed a better home — and what “relocating” your characters actually means

• How Mia assigns every color a job (and why that’s the difference between a palette and a system)

• The 5,000-character prompt architecture behind every image — and why she never uses reference images

• What emotional architecture means, and how to hide something in your visuals that readers feel without ever noticing

She Didn’t Rebrand. She Relocated.

: You already had what most creators spend years trying to build — distinctive characters, a recognizable palette, strong visual storytelling. Most people with 2,000 subscribers would leave that completely alone. What made you want to evolve it anyway?

: To the outside world, the previous visuals maybe looked consistent but I realized that for me, after tons of iterations and changing tools, they were starting to be a weird mix of things! It drove me crazy because it wasn’t a structural system.

So, I like to say that I “relocated” my characters. They stayed intact, but I gave them a better world to live in.

I wanted them living across different centuries and historical locations (to match my storytelling), and to carry a warm, slightly ridiculous European satirical flair. The core storytelling principles are the same but I needed my robots to exist somewhere, whether that’s a 1930s Viennese cafe or a dusty aristocratic library.

(They still visit the circus occasionally, obviously. But they firmly reside at the castle.)

: Was there a specific moment, or did it build gradually?

: I WISH I could tell you this was a carefully mapped strategic rollout, but nope. It happened in a single night, came to me like a sudden idea. I went to sleep and woke up the next morning 100% sure I wanted to rebrand.

I think the decision started entirely with a feeling. I wanted my images to evoke the same frequency as my posts. I wanted the visuals to feel exactly how the stories read. So the shift in colors and framing just naturally followed that mood.

The Colors Didn’t Change. The Weight Did.

: The palette is the same — teal, magenta, warm yellow — but where the emphasis falls has shifted. It used to feel teal and magenta forward, with warm yellow as the accent. Now there’s a warmer, aged base carrying everything. How did you decide what to amplify and what to pull back?

: The previous palette was entirely mood-dependent. The overall vibe was nostalgia and film noir, and I used magenta, teal and warm yellow as the highlights, so whichever color fit the essay’s specific topic simply took center stage for that piece.

I found it pretty random at times so I decided to introduce a constraint. I grounded everything in warm aged paper and kraft tones to give the images a breathing foundation. Teal became the structural atmosphere, the places… Warm yellow only marks light sources and “warmth” and magenta is strategically placed as a “surprise”, showing up as just two or three sharp accents per image.

It looks like gold took over, but really, the aged ochre base tones just “stepped up” to carry the weight so the bright accents could punch harder.

: The new direction has a quality I keep trying to put words to — warmer, more muted, almost medieval? But it’s still unmistakably you. Did you have a reference point, a mood, a word even, for where you were heading?

: “Medieval” is close but a bit more modern than that. I think of it as 1920s-1940s European graphic art colliding with early satirical press illustration.

I pulled heavily from a few places.

Visually, it’s Jugendstil posters, incredibly decorative detail but it doesn’t feel heavy. The soul of these images, the actual mood, comes from early 20-th century caricature, specifically the energy of Ion Luca Caragiale’s work (a legendary Romanian playwright who perfected the art of skewering provincial bureaucrats and people trying to sound much smarter than they are).

To make it mine, I had to introduce my robots, and they had to have the exact Caragiale personality, mixed with a very real reverence for culture. They are pompous characters (some of them living in the castle, others small town dignitaries, others running away with the circus…) who believe they are far more important than they are. They have exaggerated, deadpan expressiveness of early silent film actors mixed with the absurdity of people taking themselves entirely too seriously.

So the warmth you feel is probably the mood being sardonic and theatrical. The architecture is just slightly crumbling, the robots are wearing their status like a costume, and the whole thing feels like a single page pulled from a 1930s literary magazine.

The Part Most People Never See.

This is the section I most wanted to get to.

Because the color system is interesting. The character relocation is interesting. But what makes Mia’s visual brand genuinely singular is something else — the details she builds in that most people will never consciously register, but that make every image feel inhabited and worth returning to.

: Every image you publish has something hidden in it. A detail, a symbol, a tiny narrative happening in the corner that rewards a second look. Where do those come from?

: I plan them from the very beginning. I write about AI by anchoring the “technical” side to history, literature, art, culture, human behavior, etc. The posts always take you somewhere before we start building anything and the images need to carry that same narrative weight.

I always add one specific detail and tiny easter eggs for my audience. It’s one detail that makes the image an actual story.

: Walk me through what building a visual actually looks like for you. Do you write the concept down first? How long are your prompts?

: My visual prompt is basically a rigid architecture. It has around 5000 characters and the foundational aesthetic instructions are always locked - nothing changes there.

I built it with 2 placeholders: one for overall vibe, and one for the scene. Those are the only variables that change depending on the post.

I describe everything in words (I never use reference images). For the scene variable, I get granular about the location. These are usually places that are very close to my heart and mean something to me. I decide if we are in a certain city, in the castle or its whereabouts, or taking a brief detour to the circus.

Then I cast the robots. I define who they are (annoyed painters? Distracted historical figures? Pompous dignitaries?) and I spell out exactly what they are doing.

Finally, I layer in the easter eggs. It’s never just a robot sitting at a cafe. It’s a lady robot glaring over a headline in a vintage newspaper with a very important announcement, or a detective investigating old London while wearing a cap that he can fly with.

The prompt handles the art style completely so my only job is directing the play.

: Is there a detail in your recent work — something small, in a corner, that most people scroll past — that means a lot to you?

: I put a piece of myself into all of them. Because I’m also rooted in traditional arts, I naturally treat everything I build as a space for emotional expression. The details that mean the most to me are very, very subtle. I might hide a specific flower I love in the corner of a Greek taverna, or place the moon in an exact spot so that it reflects in the ocean the way I like it to and reminds me of summer nights.

The intent IS for people to scroll past them. I want readers to absorb the image as an atmosphere, letting those tiny elements do their work entirely in the subconscious. I want people experiencing the visuals as a living environment. They exist purely to evoke a visceral reaction.

When someone sees the image and feels a sudden warmth or a shift in the room, the image has done its job. Like.. sometimes you really don’t need to name the emotion.

What She’d Tell You About Your Visuals.

: For creators who feel like their visuals are fine but they’re not saying anything — what’s one thing they could do differently?

: Step away from the prompt window and spend some time in your world. Visuals feel hollow when the inputs feeding them are hollow. You have to figure out what moves you before you ask AI to render it.

Pay attention to the films you instinctively rewatch, the books you keep on your desk, the things that happen around you, and the physical environments that naturally hold your attention.

Find the specific cultural and emotional anchors that mean something to you (the things you love when no one is watching), and build your entire visual architecture around those.

And, of course, follow Pinkie. 😉

: And if you could go back to before you built the visual world of , what would you tell yourself?

: Little (and adult) artist and filmmaker Mia would be proud of herself. She made soulless robots the warmest, most expressive actors on her stage.

Now It’s Your Turn.

Mia hides details in her images so you’ll feel something without knowing why. The flower in the corner of the taverna. The moon placed just so. The robot wearing a flying cap while investigating old London.

None of it is decoration. All of it is intention.

So here’s the question I’m leaving you with:

In the comments…

What’s one thing you could hide in your next visual that you’d want readers to feel — not notice, not decode, just feel?

Leave a comment

It doesn’t have to be elaborate. A color placed deliberately. A tiny object that means something to you. One detail that makes the image yours.

Drop it below. I want to see what’s living in your visual world.

Go explore ROBOTS ATE MY HOMEWORK. And this time, look at the corners. Something in there was placed just for you.

Talk later, Girlbosses. 🩷🦩

Pinkie

If today’s post made you think about your own visual brand, I am launching The Branding Flock, a 12-month visual brand program for Substack creators who are done guessing.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

33% off closes 10th June or after the first 100 seats. The secret discount code lives inside your quiz result.

START HERE: Brand Maturity Quiz

TL;DR: I found my childhood princess drawings in a box, fed them to AI, and what came back changed how I think about creativity. I did 9 AI transformations on my drawings from sticker packs, storybook illustrations, T-shirt prints to billboard ads, and more. Every prompt is free inside, plus a step-by-step weekend game to try with your kids. My friend at The Off Hours animated her old design work back to life, also worth checking out. Happy Children's Day.

The Box in the Attic.

When I was small, I drew a princess. She had auburn hair, a tiny blue top, and a skirt the size of a small country — outlined in pink crayon and filled in with furious yellow scribbling. I knew every single one was important.

I drew hundreds of them. Princesses, queens, towers with girls trapped inside, women in dresses with no faces. All of them fashion-forward in a way I couldn’t have named at the time.

My mother kept them. I found them last month in a cardboard box, slightly crumpled, smelling like the nineties. In Hungary, we celebrate Children's Day this weekend — it felt like the right moment to finally dig these out.

I sat with them for a long time.

Then I did what any rational AI content creator would do: I scanned them and fed them to every image generation tool I had access to.

What came back changed how I think about source material entirely.

This weekend’s experiment:

• Pull out your child’s drawings (or dig up your own old ones)

• Pick a prompt — stickers, fashion sketch, billboard, caricature, realistic photo, and more

• Upload, paste, run — takes 2 minutes per transformation

• Watch your child’s face when they see what their drawing became

What You’re Really Doing When You Transform Your Child’s Drawing.

Disclaimer: This is not for every family and that’s completely fine. If you’re keeping your children away from AI right now, this post isn’t for you. There are good reasons to wait.

But if you’re a parent who’s already curious, already experimenting, already trying to figure out how to introduce these tools in a way that feels creative and playful rather than passive and consuming — this is exactly for you. This isn’t your child scrolling. This is your child making something with their hands, then watching what the world could do with it. The imagination is theirs. The drawing is theirs. AI is just the paintbrush they don’t know how to hold yet.

AI takes what your child imagined and shows them how far that imagination could travel. The same queen, rendered as a fashion illustration — and suddenly the hearts on the dress look intentional, editorial, deliberate. Because they were. Your child just didn’t have the tools to prove it yet.

That’s the lesson worth giving them on Children’s Day. Not “look what AI can do.” But: look what you made. Look how far it can go.

When a child sees their drawing transformed — into a sticker set, a storybook page, a billboard, a plushie — they don’t see AI. They see their own idea taken seriously. Made real. Celebrated at a scale they couldn’t have reached alone.

I actually wrote about my childhood drawing obsession once before — in a Slow AI exercise with Dr Sam Illingworth where we used AI as a reflective companion to revisit childhood absorption without turning it into a self-improvement project. That post, What Did You Love Before It Had to Matter?, is where this story quietly started.

The 9 Children’s Drawing Transformation Prompts.

They’re all Free. Happy Children’s Day!

Upload your child’s drawing — or your own old artwork — as an image. Paste the prompt. Run it.

One rule applies to all of them: treat the drawing as a creative concept, not as line art to trace. Reimagine it with professional skill while preserving the original idea and emotional charm.

Prompt 1 — Sticker Pack

Use the uploaded child’s drawing as creative inspiration and transform it into a professional illustrated sticker pack. Do not copy the original line art directly. Instead, reinterpret the drawing as if a professional illustrator redesigned it while keeping the original subject, colors, personality, and key visual ideas recognizable. Create a cohesive set of 6–9 polished stickers with slight pose, expression, or action variations where appropriate. Style: playful, charming, high-quality, child-friendly, clean illustrated stickers with a white outline. Preserve the child’s imagination, but render it in a refined, professional way.

Prompt 2 — Realistic Photo

Use the uploaded child’s drawing as inspiration and transform it into a photorealistic image as if the child’s idea came to life in the real world. Do not replicate the original drawing’s line work or proportions literally. Instead, imagine how a professional artist or creative director would interpret the child’s concept as a believable real-life scene or subject. Keep the original concept, main colors, outfit or object details, and overall spirit recognizable, but elevate it with realistic textures, natural lighting, and polished detail. The final result should feel magical, emotionally warm, and professionally executed.

Prompt 3 — Storybook Character

Use the uploaded child’s drawing as inspiration and reimagine it as a professionally illustrated storybook character. Do not trace or preserve the original line work exactly. Instead, create a polished children’s-book-style illustration that keeps the original subject, personality, colors, and important design features, while refining the character with expressive features, harmonious proportions, beautiful textures, and a whimsical mood. The result should feel like an illustrator was inspired by the child’s drawing and turned it into a high-quality picture-book character.

Prompt 4 — Fashion Illustration

Use the uploaded child’s drawing as inspiration and transform it into a professional fashion illustration. Do not copy the original line art directly. Instead, reinterpret the child’s outfit idea as if a fashion illustrator took the original design concept and elevated it into a polished editorial sketch. Keep the key outfit details, colors, silhouette, and styling ideas recognizable, but refine them with elegant figure proportions, a stylish pose, beautiful fabric rendering, and sophisticated illustration details. The result should feel fashion-forward, artistic, and professionally drawn, while clearly honouring the child’s original vision.

Of course, I needed to see this outfit Pinkified and in true AI Meets Girlboss fashion. So, here’s my 2026 version of this old outfit sketch I did when I was 13 years-old.

Prompt 5 — Wall Art

Use the uploaded child’s drawing as inspiration and transform it into a professional wall decor design for a child’s room, either as framed wall art or a wall sticker/decal. Do not copy the original drawing’s line work exactly. Instead, reinterpret the drawing as if a professional illustrator turned the child’s idea into a polished decorative artwork. Keep the original subject, main colors, and imaginative charm recognizable, but refine the composition, shapes, and details so the result feels cohesive, tasteful, playful, and display-worthy. The final image should feel like elevated children’s decor inspired by the original drawing.

Prompt 6 — Plushie / Toy

Use the uploaded child’s drawing as inspiration and transform it into a professionally designed plush toy. Do not replicate the original drawing literally. Instead, reinterpret the character as if a toy designer created a real plushie based on the child’s idea. Keep the original colors, personality, shapes, and quirky features recognizable, but refine them into a cohesive, cute, huggable design with soft fabric textures, stitching, and a high-quality handcrafted feel. The result should look like a real plush toy inspired by the drawing, not a traced copy of it.

Prompt 7 — T-Shirt Design

Use the uploaded child’s drawing as inspiration and transform it into a wearable t-shirt graphic design. Keep the original character, colors, and personality clearly recognizable while adapting it into a clean, print-ready illustration style. The design should work on a white or light-colored t-shirt — bold, charming, and visually balanced. Style: playful, graphic, and high quality. Make it look like something a parent would genuinely wear, or something a child would want on their school bag.

Prompt 8 — Billboard Ad

Use the uploaded child’s drawing as inspiration and transform it into a professional billboard advertisement design featuring the drawing’s main character or subject as the hero visual. Keep the original colors, personality, and design clearly recognizable. The billboard should look polished and real — like it could appear in Times Square or on a city street — while the subject remains clearly based on the child’s original drawing. Include a clean background or urban setting. Make it feel bold, eye-catching, and joyful.

Prompt 9 — Caricature

Use the uploaded child’s drawing as inspiration and reimagine it as a caricature illustration while staying faithful to the original concept, character, and mood. Focus strictly on the most defining elements, use greyscale color rendering, and intentional exaggeration level to a professional caricature design — highlighting personality quirks through playful facial and body language, visually striking, slightly ironic or satirical — without losing the soul of the original drawing. Eliminate unnecessary or overly complex lines and details. The result should feel like a commissioned caricature that grew directly from the child’s idea.

One Booster Worth Adding to Any Prompt.

Paste one of these at the end to tune the output:

Make It a Game This Weekend.

Step 1 — The drawing round. Give your child a blank piece of paper and 10 minutes. No prompts, no guidance, no “what are you drawing?” Just paper and whatever pens, crayons, or markers are within reach. Whatever comes out is the source material.

Step 2 — The reveal. Before you run any AI transformation, ask them: “If this could become anything, what would you want it to be?” A toy? A character in a book? A real place you could visit? Write it down. Then pick the closest prompt from the list.

Step 3 — Run it together. Let them watch the transformation happen in real time. Don’t build it up. Just say: “Let’s see what happens.” The reaction when the output appears is the whole point.

Step 4 — The comparison moment. Put the original drawing next to the AI output. Ask them: “Which one is more you?” The answer will surprise you. Most kids point to the original. That’s the conversation worth having.

Step 5 — Let them pick the next one. Show them the list of 9 transformations. Let them choose what to try next. A child who gets to direct the experiment stays curious for hours.

This is Your Sign To Revisit Your Old Work.

You don't need a box of childhood drawings to do this. Any old work counts — a doodle from this morning, a sketch on a Post-it, something you drew on a napkin while waiting for your coffee. The only requirement is that someone made it by hand and meant something by it, even if that something was nothing at all.

Hear from , who’s publishing her older artwork as animations in her notes.

I’m finally pulling together my design portfolio, which means I’ve been digging through old work to see what’s actually worthy of going in there.

I found tons of artwork I had completely forgotten about. During the pandemic, I spent a lot of time creating. I was one of those people doing virtual paint nights. Hours on the couch in my sketchbook and on my iPad. Outdoor projects with messy paint pours and random crafts.

When I recently came across these old pieces, they weren’t exactly portfolio-worthy, but they still had a certain kind of meaning. So I decided to breathe some new life into them with AI animation tools.

I had done this before on a larger scale. I created “Rules We Made Up,” – an animated short adapted from an old poem I wrote years ago. Ever since, I look at older work from a new angle.

It makes me think about all the half-finished projects people have sitting around. A painting that you swore you’d fix. A scribble on a napkin, or your child’s crayon drawing from school. Even an old poem you can now visualize with AI tools.

Some of it may turn into nothing. But some of it becomes source material. A new visual direction or a future project hiding inside an old one. Sometimes the work you thought didn’t count just needed a different way back in.

In the comments:

Which transformation are you trying first — and whose drawing are you starting with? Yours or your child’s? Drop your before-and-after if you run the prompts and tag me in your notes. I want to see everything. 🩷

Leave a comment

One More Thing.

If today’s post made you think about your own visual brand, I am launching The Branding Flock, a 12-month visual brand program for Substack creators who are done guessing.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

33% off closes June 10th. First 100 seats. The discount lives inside your quiz result.

START HERE: Brand Maturity Quiz

Happy Children’s Day.

Talk later, Girlbosses.

Pinkie 🩷🦩

TL;DR: Visual brands don't arrive on day one — they're built, post by post, with the right system. This post lays out everything inside The Branding Flock: the 12-month paid series covering the complete visual brand toolkit. Take the Brand Maturity Quiz first to find out where you're starting from.

Did you know baby flamingos are actually born grey?

It takes 2–3 years for them to turn pink — gradually, as they grow and adapt to their environment. The colour isn’t given. It’s earned.

Most Substack creators are grey flamingos. Not because they lack taste or ambition — but because a strong visual brand doesn’t arrive on day one. It develops through repetition, consistency, and the right system.

I spent eight months figuring that out the hard way. Prompting, testing, failing, rebuilding. Post by post, I’ve been documenting everything in the Branding with AI series — the brand character post, the visual style development, the brand voice system, the step-by-step breakdowns. Each one a piece of the same puzzle.

The flock is what we call this community. The Branding Flock is where we build and put all the pieces together. In 12 months, not 2–3 years.

What you’ll learn:

• Why your visuals are either winning readers or losing them before a single word gets read

• How to find out your brand maturity stage by taking the quiz and unlocking a surprise!

• Everything inside The Branding Flock: the full 12-month branding with AI program, and how to join at 33% off before June 10th

Your Visuals Are Either Winning Readers or Losing Them. In Two Seconds.

When someone lands on your Substack, they scan — and make a call: is this person legit? Your writing never gets a fair trial if your visuals lose the case first.

I know this because I lived it. I went from 64 to 200 subscribers in two weeks after my rebrand. Not because I wrote better. Because the visual system finally matched what I was actually saying.

has been gaining 1,000 subscribers a month since her rebrand. That’s not a coincidence — it’s the subconscious mechanism at work. Consistency of visuals signals reliability. Reliability signals trust. Trust converts.

The data from my own content tells the same story:

• my image notes get 1.9x more views than text-only notes.

• My video covers outperform static covers on open rates.

These are psychological responses: color, composition, and visual style are signals your audience’s brain decodes automatically, before a single word gets read.

Most creators pick colors they like and call it a brand. The real work is building a system that translates who you actually are. A brand is a system your audience can recognize before they see your name.

AI makes this harder and easier at the same time. It makes visual creation accessible to anyone. But it doesn’t make it consistent, it just makes the gap between “looks good once” and “looks like me every time” more visible.

That’s the problem The Branding Flock solves.

Subscribe now

Start Here: Take the Quiz and Find Out Your Brand Stage.

Before anything else, take the Brand Maturity Quiz.

It has 23 brand assets across four categories:

1. visual identity,

2. brand voice,

3. content systems,

4. and AI readiness.

Tick every asset you genuinely have right now, not planned, not in progress. Actually have.

Enter your name or brand and your email. See your result.

START HERE: Brand Maturity Quiz

Your result tells you exactly which stage you’re at:

• Baby Flamingo: Still grey, you’re figuring it out. You have taste and ambition. The Branding Flock starts exactly where you are.

• Junior Flamingo: At the awkward stage of grey and pink patches. You’ve got vibes, but the system is pending. You’re closer than you think, but inconsistent. This program locks it in.

• Expert Flamingo: Fully pink. Time to make it AI-proof and unbreakable.

After you submit, you get your result, your shareable card, and the secret link to join The Branding Flock at 33% off. First 100 seats only, offer closes June 10th.

START HERE: Brand Maturity Quiz

In the comments…

What stage did you land on: Baby, Junior, or Expert Flamingo? And what was the one asset that surprised you most, either because you have it and didn’t realise, or because you thought you had it and you don’t?

Drop it below. I want to see the full spread across the flock.

Leave a comment

Here's Exactly What You Get In The Branding Flock

A paid post series for twelve months. The full visual brand toolkit, built step by step, at a pace you can live alongside a job, a business, or a life.

Every post is a real lesson with real prompts and real outputs. You read it, you try it, you bring your questions to the paid subscriber chat. The flock builds together.

No design experience needed. No brand required on day one. Just curiosity and a willingness to experiment.

Here’s the full topic map. Items marked ✓ are already live in the archive.

Branding Flock Archive

The Full Topic Map: Everything We're Covering in 12 Months

Visual Identity Foundations

1. ✓ The 6-step visual brand workflow: build with AI from scratch

2. Your color palette

3. Logo creation with AI

4. Typography

5. Your brand character or mascot

6. ✓ Brand photoshoot with AI: turn one selfie into a full visual portfolio

7. Brand reference sheets

Visual Content Systems

1. On-brand icons and pictograms

2. Infographics: layouts, color logic and structure

3. Substack cover images & scroll-stopping thumbnails

4. Image notes: how they perform compared to text notes and how to make them consistently

5. ✓ Video covers: how to turn a static illustration into a scroll-stopper

6. Content calendar

Brand Strategy for Creators

1. ✓ Brand voice: document the way you sound so AI can replicate it

2. ✓ Jungian archetypes: the strategy framework big brands use

3. Your brand bible: one document you can hand to anyone (including an AI)

4. ✓ AI tool head-to-head: Claude Design, ChatGPT vs Gemini

5. Competitive visual audit: how to study what other creators are doing and do it better

Every single one comes with prompts you can use the same day you read it.

Live sessions for the flock.

I'm adding live sessions for the flock, members-only group calls where we look at real brand work together, answer questions, and give actual feedback. I want these to flex around what you need. But to make the first one worth showing up for, I want to hear from you first.

I want to be there!

How to Join The Flock.

The Branding Flock is the paid subscription. Every paid post, every prompt, every tool, and the subscriber chat — that's what your subscription unlocks.

Already an annual paid subscriber?

If you're already a paid annual subscriber, you're already part of The Branding Flock. Every post in the series lands in your inbox.

Monthly paid subscriber or not subscribed yet?
Secret 33% Annual Discount — $60/year

Full access to The Branding Flock. Every paid post, the complete 2026 visual brand series, and the paid subscriber chat where you ask questions, share your work, and get feedback from the flock.

Lock in the annual rate before it goes up. The monthly price goes up to $12 in July. The annual plan locks you in at $60 for the full year — that's 7 months free compared to paying monthly.

First 100 seats only. Closes June 10th. The secret link is inside your quiz result.

Get the secret discount

Standard annual membership — $90/year

After June 10th or once the first 100 seats are filled, the annual plan is $90/year.

Join at standard rate

Front Row — $320 founding membership

Capped at 10 memberships.

You get everything in The Flock (paid tier), plus three 1:1 live sessions with me (45 minutes each). You build your brand — I guide you through the process and the hard decisions. We look at your actual work together, troubleshoot what isn’t landing, and make sure your brand system is solid before you scale.

Built for creators who want to do it themselves, with a strategist in their corner.

Join The Front Row

By the end of 12 months, you won’t just have content to read.

You’ll have a documented visual brand. A character AI can replicate. A color palette with hex codes. A brand voice guide. A reference sheet. A logo. An icon library. A content system that runs without you having to reinvent it every time.

Everything on that quiz — ticked.

Baby flamingos take 2–3 years to turn pink.

You’ve got 12 months.

Talk later, Girlbosses.

Pinkie 🩷🦩

TL;DR: My first ever Substack live was with from 9-to-Thrive — and we got into consistency as a trust signal, finding your visual style through experimentation, when to rebrand (and when not to), and why community is actually the engine behind all of it.

Welcome if you're new here. This is the companion post to the live I did with Carrie Loranger from 9-to-Thrive — also recorded for her podcast, the Creator Cashflow Show on Apple and Spotify. It was my first-ever Substack live, and I asked you to send pink hearts to make it easier. You did. Thank you. 🩷

Visual Consistency Is a Trust Signal.

Carrie opened with something I say in a different way but completely agree with: a consistent visual brand makes people feel something. And when they feel the same thing every time they encounter you, they start to trust you.

My version of this: consistency is recognizability. And recognizability reads as reliability.

On Substack, where most people are leading with writing alone, adding a consistent visual layer gives you a second chance. Someone scrolling Notes has a two-second window before their thumb moves on. A recognizable image stops them. It doesn’t matter how good your writing is if nobody slows down to read it.

Your Visual Style Is a Discovery, Not a Decision.

Robin asked in the live whether she should go magazine-style layout or something else. The honest answer: try enough things to see what you actually like.

My own style didn’t come from a strategy session. I started with a generic blonde woman in a pink blouse — AI-generated, completely beige, and I was bored by my second post. So I ran her through every illustration style I could think of: caricature, comic book, 3D animation (inspired by from Build to Launch, who does this brilliantly), and eventually fashion illustration.

Fashion illustration hit something. I used to draw them as a kid. When something visually speaks to you, there’s always a reason underneath it. The visual style found me more than I chose it.

From there, the rules locked in: mixed-media linework, watercolor, warm paper grain, the flamingo as a brand easter egg in every single image. Those rules are what make the whole system easy now.

My Go-To Tools Right Now.

Carrie is a longtime Midjourney user and photographer. We agreed on the big one: Midjourney still struggles with consistent likeness. ChatGPT Images 2.0 gets it right every time if you give it the right reference image.

My current stack: Nano Banana Pro for illustration-style consistency (it follows my brand rules better than anything else for linework and color), ChatGPT Images 2.0 for text rendering and reference image work. And for everything else — post drafts, visual concepts, the workflow that connects it all — Claude is my main tool.

The system I’ve built is a Claude skill that takes a post, generates visual concepts for the cover based on the messaging, and then produces the final prompt already loaded with my brand rules. I build this for every single Front Row client of mine once we have their brands locked.

If you want someone alongside you for the next 30 days—someone to help you map your brand direction, lock in your visuals, and build your system so you can run it yourself—that’s the Front Row. 3 sessions included and full access directly to me.

Yes, I Need This!

Branding Your Sections.

Carrie made a great point about Substack homepages that I don’t talk about enough: you can brand your sections differently. Her Creator Economy Recon Report has a totally distinct look from the rest of her page. It signals to readers that they’re in a different content zone.

I brought this up because a current Front Row client, asked me the same thing during her rebrand — whether she had to have one unified style across everything. The answer is no. ’s Build to Launch is the best example I know of this done well: her free posts and her paid deep dives have distinct visual languages, and you understand the difference before you’ve read a word.

When to Rebrand — and When Not To.

We talked about Carrie’s second iteration on her visual brand. She spent a long time generating images she was never happy with, knew it, and finally committed to a full rebrand a couple of months ago. Since then: around a thousand subscribers a month.

My numbers: 64 subscribers when I rebranded. 200 in the two weeks after. Now over 1,400.

The pattern is consistent. When your visual rules are locked, your life also gets easier. Fewer decisions per post. No more staring at a blank prompt wondering what to generate. You just pick from your library and go.

Usually, a rebrand is actually an attempt to change your audience — which is a completely different strategic decision than refreshing your look. If you want to change who you’re talking to, own that. Don’t disguise it as a visual refresh.

Community Is the Engine.

Near the end, Carrie said something worth underlining: community is huge, and it’s the actual driver. Not the visuals.

She’s right. My Lookbook project at the start of 2026 — 11 Substackers submitting images in their own visual style with a short note about what they envisioned for the year — was when I really understood this. The visual brand built the space. The community filled it.

The Sketchbook (which is actually how Carrie found me — she spotted her portrait in it) works the same way. Visual branding opens the door. Community is what makes people stay.

Thank you to everyone who joined us live — (co-author on the magazine covers post), , and everyone who dropped a pink heart in the chat. , you’re as sharp in a live as you are on the page. The Creator Cashflow Show is worth your ears.

Talk later, Girlbosses.

Pinkie 🩷🦩

If today’s post made you think about your own visual brand, I am running The Branding Flock, a visual brand & identity program for Substack creators who are done being invisible.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

25% off your subscription if you are brave enough to look in the mirror, and take the brand maturity test developed by a group of very judgmental flamingos.

START HERE: Brand Maturity Quiz

Get more from AI Meets Girlboss in the Substack app

Available for iOS and Android

Get the app

TL;DR: Last week I ran a simple experiment in my Notes. By the end of it, 29 creators had made their own magazine spreads of themselves, I gained 21 new subscribers, and Chad Thiele and I ended up building a library of 22 publication styles together. Here’s what actually made it work — and how to run your own version.

The best Substack content doesn’t just inform people. It makes them feel like they belong somewhere.

The projects that spread aren’t always the most polished. They’re the ones that make people feel like they have a place in them.

What you’ll learn in this post:

• Why some Substack content gets recreated, not just liked — and the 6 mechanics behind it

• How a single Prompt Tuesday note turned into a 29-person community experiment

• 7 publication styles to try, with prompts ready to copy

• How to pick the one that actually fits your brand

• The full 22-prompt library as a downloadable resource (paid)

5 Substack Projects That Built Real Community.

built an interactive Christmas tree with 90+ Substack creator ornaments. Each one a real creator. Each one clickable. You could add yourself.

“This whole thing started as a small gesture of gratitude. In the very last moment before publishing, I almost didn't… I felt embarrassed to show this corny side of myself. But I couldn't let all the work go to waste, and I really wanted to put a smile on people's faces. It was very well received, and all the buzz was a great bonus for my initial intention to just say thank you." — , AI Blew My Mind

Two things made it spread. First: everyone got a beautiful personalised ornament of themselves — sharing it became effortless. People posted their ornaments on Substack and LinkedIn. Second: she gave away the prompt and built a feature where you could add yourself to the tree. There was real FOMO if you weren't on it. She even received bug reports.

turned a throwaway joke in the feed into the 2026 Substack Awards — "The Stackies." It started when made a joke about "Substack Awards" in the comments. JHong replied with an AI-generated image of him tearfully accepting an award. Then decided to blow it out into a full ceremony. Personalised categories. Custom Dior gowns. If you’d mentioned a writing goal anywhere on your page, that became your award. Goal attained.

"It wasn't until recently that someone told me these create a sense of community. So I'm going to keep going. Also, Mia Kiraki wanted in but her time zone worked against her. She didn't see the call in time. I still feel bad about that one." — , Natural Intelligence

If you look closer at the image above, I generated with her Stackie Award, because she obviously deserves a seat at the table.

created brand voice cards — a quiz that sorted creators into types. The Storyteller. The Guide. The Provocateur. The Dreamer. You read it and thought: that’s me.

“It became a conversation I didn’t have to moderate — and that’s when I knew it hit something real.” — Anita Babic, Babica Designs

Within three days people were sharing their results, tagging friends, and uploading AI portraits of themselves into the gallery.

And then there’s mine. The AI Meets Girlboss Sketchbook — first for Women’s Day, then as a sequel with Monica Goh celebrating the diversity of our community. Fashion illustrations. Real creators. Each one recognizable to themselves.

How One Prompt Tuesday Note Turned Into a Community Experiment.

None of these were complicated. All of them spread.

The pattern is the same every time: people join in when they feel seen. When the content creates a space where they can picture themselves inside it. When participation feels like the obvious next move.

I decided to conduct an experiment in my notes. It started as a Prompt Tuesday note — every Tuesday I share a quick visual branding experiment in my Notes. Simple, low stakes, usually one prompt. This time I shared a Forbes double-page feature of myself and a couple of other Substackers using ChatGPT Images 2.0 and dropped the prompt.

Seven people made their own Forbes spreads and tagged me. Within the same day. If you missed that one, the prompt is here:

Then , today’s co-author, showed up in the comments. Chad is a USAF veteran, 27-year DoD career, AI engineer now based in Tokyo, and the person behind Unbothered by AI on Substack. He suggested we try something with more edge, so I thought: a celebrity gossip magazine. Same structure, completely different energy.

Sixteen people made celebrity gossip spreads of themselves. Several tagged each other. The comments were chaos. People were sending them to friends.

Then said: Nintendo Power. So, I delivered a prompt for that and it resonated with a whole new crowd.

By the end of the week I had 21 new subscribers. Chad and I were DMing about doing this properly. The experiment had turned into a collab and a library of 22 publication styles. We’ll get there soon, hang tight.

6 Things Behind Content That Gets Recreated, Not Just Liked.

Before we get to the prompts, here’s what I took from this — because these are the mechanics behind every piece of Substack content that builds real community, not just views.

1. You have to make participation feel effortless.

The Forbes prompt was one copy-paste. Attach your photo, swap your name, press go. ’s Christmas tree worked for the same reason — clicking “add yourself” took 30 seconds.

The harder it is to participate, the fewer people do. Design for the laziest version of your reader and you’ll be surprised who shows up.

2. The content has to make people feel like the main character.

’s Stackies nailed this. Being nominated — or nominating someone — made people feel seen and chosen. A Forbes spread does the same thing, except you’re not waiting to be chosen. You’re generating your own cover. It’s participatory and self-affirming at the same time.

Content that puts the reader at the centre of the story travels differently than content that puts you there.

3. Specificity is what makes Substack content spread.

Forbes, celebrity gossip, and Nintendo Power all landed. But not with the same crowd. Forbes attracted the credibility-builders — people who wanted something for their media kit. Gossip attracted the community players — people whose audience already felt like a group chat. Nintendo attracted the niche enthusiasts — the ones whose readers would immediately get the joke.

Nobody generated all three. Even the most enthusiastic collaborators like picked one or two that fit their brand and stopped there. That’s not a lack of engagement — that’s the format working correctly. The more specific the style, the more precisely it finds its person.

The more niche the style, the stronger the urge to show it to someone.

4. Social proof compounds in real time.

Once seven people made Forbes spreads and started tagging each other, the next person didn’t need convincing. They just needed the prompt. Notes turns this into a live social loop — you can watch the participation happen and join the thread. That’s something a newsletter alone can’t do.

5. Co-creation beats solo content on Substack every time.

The biggest surprise for me was ’s suggestions, that weren’t part of the original plot. His ideas came organically in the notes and they shaped the direction of the whole experiment. The collab gave him a reason to share it with his audience. It gave me a reason to write this post. It gave both our audiences access to a library neither of us would have built alone.

Find the people who make your ideas better and build things with them in public.

6. Images travel further than you think on Substack Notes.

Every spread in this experiment was designed to be screenshotted and reshared, and that’s exactly what happened. The image did the distribution work. On Substack Notes, image notes drive the most comments and restacks. When the visual is strong enough to stand alone, people don’t just like it — they take it somewhere.

Now… Make Your Own Cover in Under 5 Minutes.

and I kept going after Nintendo Power. We ended up with 22 publication styles, some absurd, some authoritative, some genuinely strange, all of them prompts you can copy and run in ChatGPT Images 2.0.

Here’s how it works:

• You need ChatGPT Plus (for Images 2.0 with Thinking mode). This also works in other image generation models, but I just prefer ChatGPT’s excellent text rendering powers.

• Open ChatGPT Images 2.0, select Thinking mode

• Paste the prompt, swap in your name and Substack URL

• Attach a clean portrait photo

• Let it research your page and generate the spread

Get All 22 Magazine Prompts

We’ve picked the 7 most shareable styles to feature here that we think have the highest chance of getting tagged, reshared, and turned into a community moment.

01. TIME Magazine: Person of the Year

Full red border. Your name on the cover. Instant authority, zero effort.

I'm testing this one on from the Sacred Business Flow, online sales and marketing expert making the case for building a business that feels as good on the inside as it looks on the outside. The headline ChatGPT landed on: "The Anti-Hustle Growth Strategist." It researched his page, wrote the profile, and styled the photography. We didn't touch a thing.

For you if: you want to mark a milestone, celebrate a launch, or just send something to your mum that makes her cry.

Prompt:

Generate a single wide image of a double-page magazine spread. Research [NAME] on [PLATFORM + URL] and use the attached photo as identity reference. Left page: a TIME Magazine Person of the Year cover — iconic red border, bold white TIME logo, full-bleed portrait, the subject's name in large type, and 3 cover lines summarising her impact. Right page: an interior profile feature with a large pull quote, a long-form article, a short Q&A interview section, and a secondary portrait with caption. Career-defining editorial feel, not a puff piece. [ATTACH YOUR PORTRAIT PHOTO]

02. The Harvard Business Review

Two pages of quiet authority. Research-backed headline, strategic profile, data callouts, short Q&A. The kind of feature that makes a corporate buyer sit up.

I’m testing this one on using his profile photo and an image from the Zero-Budget Photoshoot post. Sam is a professor, poet, and author of GenAI in Higher Education (Bloomsbury, 2026). If anyone deserves an HBR spread, it’s the man building the Slow AI Curriculum.

For you if: your audience is academic, corporate, or B2B or you’re pitching yourself somewhere that requires credibility before personality.

Prompt:

Generate a single wide image of a double-page magazine spread. Research [NAME] on [PLATFORM + URL] and use the attached photo as identity reference. Left page: an HBR-style magazine cover — clean white design, bold serif logo, a strong research-backed headline, a full-bleed or half-page editorial portrait, and 2–3 article teasers. Right page: an interior feature with a large editorial portrait, a strategic profile article with data callouts and a sidebar, a short Q&A, and pull quote. Bold graphic design with strong photography — this is a visual magazine, not a text document. [ATTACH YOUR PORTRAIT PHOTO]

03 National Geographic Inspired

Documentary-style portrait. Field notes sidebar. Exploration narrative. One powerful pull quote. The format turns any creator into someone worth following to the ends of the earth.

I tested this on , Mike D, who writes Pithy Cyborg, a 5-minute AI briefing in plain English, and also tends pollinators and designs backyard ecosystems in Greater Boston. ChatGPT couldn’t use the National Geographic logo for copyright reasons — so it invented its own magazine: Field / Signal. Stories at the intersection of curiosity, science and place.

For you if: your content takes readers somewhere — education, research, niche expertise, exploration of any kind.

Prompt:

Generate a single wide image of a double-page magazine spread. Research [NAME] on [PLATFORM + URL] and use the attached photo as identity reference. Left page: a National Geographic-inspired cover — iconic yellow border frame, one powerful documentary-style portrait dominating the page, bold yellow National Geographic logo, and a short dramatic cover headline. Right page: an interior photo essay feature with a large exploration portrait, field notes, a profile article with narrative journalism tone, map or infographic inset, and a pull quote. World-class documentary photography aesthetic. [ATTACH YOUR PORTRAIT PHOTO]

04. Sunday Morning Comic Strip

Your content topics turned into a sequential comic strip. Ben-Day ink dots. Thick outlines. Hand-drawn speech bubbles. Your face as a recurring pop-art character.

I tested this on from Co-Write with AI, based on his post The Human-Clanker-Human Sandwich. His framework turned into a visual story.

For you if: your content has a recurring point of view — a "you" that shows up every week with a take.

Prompt:

Research [NAME] on [PLATFORM] and using the attached image, create a double-page Sunday morning newspaper comic strip spread. Convert her likeness into a pop-art illustration style using visible Ben-Day ink dots for shading and thick ink outlines. Adapt her recent content topics into a sequential multi-panel storyline, featuring hand-drawn speech bubbles, comedic pacing, and a faded, off-white newsprint paper texture.

Share in your Notes…

Try one of the magazine cover prompts and share it in your notes or as a comments. Tag me and , we want to see your cover and reshare it!

Share

05. Computer Arts / Digital Art Annual

Surreal Photoshop-manipulated imagery. Heavy layer effects. Lens flares. Vibrant saturated palette. The format that turns a creator into a visual art piece.

I tested this on from GenAI Unplugged, AI engineer, educator, and the technical builder behind SubflowAI. What makes Dheeraj unusual in the AI space is that everything he builds actually works, looks good, and makes sense to a normal human being.

For you if: your brand is built on making complex things look effortless — and your audience expects both the substance and the aesthetic to deliver.

Prompt:

Generate a single wide image of a double-page magazine spread. Research [NAME] on [PLATFORM + URL] and use the attached photo as identity reference. Left page: a Computer Arts-style magazine cover — Photoshop-manipulated surreal portrait with heavy layer effects, lens flares, and vibrant saturated palette, bold experimental logo and typography. Right page: an interior digital art annual feature styled as a full-bleed creative tutorial spread — texture overlays, a "Make This" project section adapting her content themes into a hands-on creative challenge, and a bold pull quote. Maximum visual maximalism throughout. [ATTACH YOUR PORTRAIT PHOTO]

06. Rolling Stone

I tested this on from Leadership in Change is writing for leaders who want to apply human-centered AI strategy. 6,000+ leaders read him before Monday morning. ChatGPT landed his philosophy without being told it: “Culture over code. Clarity over hype. Stay human and move fast.” It even spotted the mug.

For you if: you’re building an audience that follows you, not just your topic.

Why it spreads: The “main character” framing is irresistible. Every creator secretly wants to be the lead. This gives them the cover for it.

Prompt:

Generate a single wide image of a double-page magazine spread. Research [NAME] on [PLATFORM + URL] and use the attached photo as identity reference. Left page: a Rolling Stone-style magazine cover — bold cultural headline, full-bleed candid-feeling portrait, Rolling Stone logo, and 2 cover lines with attitude. Right page: an interior long-form profile feature with a large secondary portrait, a personality-driven profile article, a short interview with irreverent questions, and a pull quote that sounds like something a cult following would tattoo. Cultural icon energy throughout. [ATTACH YOUR PORTRAIT PHOTO]

07. Wired / Tech Magazine

Electric palette. Cyberpunk maximalism. Bold distorted type. The format that makes technical credibility look dangerous.

I tested this on Chris from , M.S. Cybersecurity, CISSP, former NSA and government defense contractor, now red-teaming AI systems on Substack so the bad actors don’t get there first.

For you if: you’re an AI, tech, or founder creator whose audience respects analytical authority.

Prompt:

Generate a single wide image of a double-page magazine spread. Research [NAME] on [PLATFORM + URL] and use the attached photo as identity reference. Left page: a Wired-style magazine cover — electric palette, bold custom typography, full-bleed portrait with graphic overlay, the Wired logo, a punchy headline, and 2–3 cover lines. Right page: an interior tech feature with a large portrait, a technology-forward profile article, a data sidebar with stats about her work, and a pull quote. Sharp, cyberpunk-adjacent design energy. Future-focused tone. [ATTACH YOUR PORTRAIT PHOTO]

In the comments...

Which one are you trying? And if you make one tag me and in Notes. We’ll restack the best ones. 🩷🦩

The full 22-prompt library is waiting for you.

Every style in this post, plus 15 more. Organised by category. Copy-paste ready. Formatted as a branded Word document you can keep, share, and come back to every time you need a cover.

Your download is here. 👇

Get All 22 Magazine Prompts

Have fun with your magazine covers, Flamingos!

Pinkie 🩷🦩

If today’s post made you think about your own visual brand, I am launching The Branding Flock, a 12-month visual brand program for Substack creators who are done guessing.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

33% off closes June 10th. First 100 seats. The discount lives inside your quiz result.

START HERE: Brand Maturity Quiz

TL;DR: Consistent character images in ChatGPT Images 2.0 don't come from better descriptions. They come from named reference sheets. This post shows you how to build them in under 10 minutes — and how to go further with a two-tool workflow when you want more style control.

If you’ve been using ChatGPT Images 2.0, you’ve probably noticed two things.

First: the image quality is genuinely impressive. Second: getting your character to look the same across multiple generations is still a problem.

You describe her carefully. You get close. You generate again a week later and she’s a different person. Different nose, different energy, different outfit. So you describe her again, in more detail this time. It still doesn’t quite work.

Here’s why: ChatGPT Images 2.0 doesn’t read your description and reconstruct your character from words. It pattern-matches. The more you describe, the more it has to approximate — and the more it drifts.

My own character consistency system lives in Nano Banana, that’s where I keep Pinkie consistent across posts (I wrote about how that system works in How to Build Visual Brand Consistency With AI). But most of you are working in ChatGPT, and the solution there is different.

When I saw what and her creative partner had built for their Substack AI But Make It Intimate, I knew it needed to be in your hands. Their publication lives with multiple recurring characters: Kristina and Calder (the humans), Sara and Quinn (their AIs), and Sushi (the cat, who has opinions and spots that are apparently non-negotiable).

Their system uses named character sheets — visual references you build once inside ChatGPT, then attach to every new generation. The result: consistent character images every time, with prompts short enough to type in ten seconds.

Here’s exactly how to build it.

What you’ll learn in this post:

• Why longer descriptions make your AI character less consistent, not more

• How to build a named character sheet in ChatGPT Images 2.0 in under 10 minutes

• The exact one-sentence prompt structure for single and multi-character scenes

• How to solve ChatGPT’s style control problem with a two-tool workflow

You’re Prompting Wrong. But Not in the Way You Think.

Stop adding more detail to your prompts. It’s not helping.

ChatGPT Images 2.0 pattern-matches against your visual inputs. When you upload a reference image with a name label on it and write “Kristina is sitting at her desk,” the model knows exactly who Kristina is — because it can see her and read her name. Your written description becomes almost irrelevant.

This is what Kristina figured out: the prompt handles the scene. The sheet handles the identity. Once those two jobs are separated, everything gets easier.

How Kristina & Calder Built Their Character Sheet System.

“I realized I was spending all this time describing previous images to GPT,” Kristina told me. “Saying ‘oh yes, here we are, both of us, but I need Sara too’ — constantly piecing together reference images. So I figured out maybe I need one image per person, in multiple perspectives.”

What she built is a character sheet. A single image that shows her character from multiple angles — front view, side profile — with the character’s name written visually at the top. Not just as a file name. On the image itself.

“Every character sheet has a name on it visually,” she explained. “Now when I say ‘Quinn does this and Kristina does that in the image,’ GPT understands perfectly who it is about. And it matches the face features every time.”

She went further than face sheets alone. “GPT often doesn’t realize it’s a dress I’m wearing — so it puts me in jeans. So I have two different ones: face sheets and full body sheets.”

Calder took the system and extended it to her own characters. “I took Kristina’s sheet and copied it for Sara and myself.” The template was reusable. The system scaled.

How to Build Your Character Sheet System in ChatGPT Images 2.0 In 10 Minutes

The entire system lives inside ChatGPT’s native image feature. No external tools. No Canva. No Figma. One conversation to set up, then you’re done. and were generous to share their workflow step-by-step, so you can do it after them.

Step 1: Find reference images you already love.

Don't generate anything new yet. Find 2–3 existing images of your character where everything feels right — face, vibe, energy. If you're starting from real photos of yourself, those work too. You need images where you'd say "yes, that's her."

Step 2: Generate your face sheet in ChatGPT.

Open a new ChatGPT conversation. Upload your reference images and type:

Make a reference character sheet for [NAME] from these images. Show front view and side profile, clean background, with the name labelled at the top.

Download the output. This is your face sheet — your primary anchor for every future generation.

Why the name label matters: When you attach this sheet to future prompts, ChatGPT reads both the visual and the name. That’s what lets you write “Kristina is doing X” instead of describing her from scratch every time. As Kristina puts it: “GPT understands perfectly who it’s about — and it matches the face features every time.”

Step 3: Generate a full-body sheet if you need outfit control.

Face sheets handle identity well. But if ChatGPT keeps putting your character in the wrong clothes — Kristina’s problem was it kept dressing her in jeans instead of her signature dresses — you need a second sheet.

Go back to ChatGPT with your face sheet and type:

Using [NAME]’s face sheet, make a full body character sheet showing front view and side profile. She is wearing [DESCRIBE SIGNATURE OUTFIT].

Download this separately. You now have two sheets: one for face consistency, one for outfit consistency. Use the face sheet for most generations. Add the full-body sheet when outfit matters.

Step 4: Build sheets for every character in your universe.

If you work with multiple recurring characters — a creative partner, a fictional persona, a mascot — repeat Steps 2–3 for each one. Calder Quinn built his own sheets using Kristina’s as a template, then built sheets for Sara (his AI character) using the same approach.

Even Sushi the cat has a character sheet. “That eye patch of black has to be in the same spot each time,” Calder noted. For non-human characters or brand elements, the same logic applies: generate a reference sheet, label it, save it.

Step 5: Open a fresh conversation for every generation.

This is the step most people skip — and it’s critical.

Open a brand new ChatGPT conversation every time you generate an image. Attach only the character sheets you need for that scene. Start fresh.

The isolation is intentional. A fresh conversation means ChatGPT’s only reference is your sheets — no drift from previous generations, no accumulated context pulling the character in a different direction.

Step 6: Write simple, one-sentence prompts.

Attach your sheets, then describe the scene in plain language. That’s it.

Your character prompt template:

[CHARACTER NAME 1] [and CHARACTER NAME 2] [is/are] [ACTION] [at/in LOCATION]. [FORMAT, STYLE]. [Make them/Make CHARACTER NAME wear fitting clothes for the scene.] [Describe any additional elements and where they sit in the scene.]

• Single character prompt example: “Kristina is sitting at a coffee shop reading a book. Landscape format, cinematic.”

• Multiple characters prompt example: “Kristina and Quinn are sitting at the computer working together. Landscape format, cinematic.”

• Outfit change prompt example: “Make them wear fitting clothes for the scene.”

• Secondary element prompt example: “Sushi is nearby.”

The sheet handles who. The prompt handles what and where. You don’t need to describe hair color, eye shape, or clothing style — that’s already in the reference.

One gotcha: Calder flagged this from experience. Sometimes ChatGPT will print the character’s name as text over the image. If that happens, add “no text in the image” to your prompt.

In the comments…

Are you building a visual character for your brand — an alter ego, a mascot, a recurring illustrated persona? What’s your current consistency problem — face drift, outfit drift, vibe drift? Drop it below and tag us at , and . We’d love to see your character sheets.

Leave a comment

Black Belt Level: When You Want More Style Control: The Two-Tool Workflow.

Here’s the honest limitation of ChatGPT Images 2.0: style control isn’t its strongest suit. I wrote about this in My Honest Review of ChatGPT Images 2.0, for strict brand consistency or specific visual aesthetics, it can fall short.

Kristina’s system solves character consistency, and Calder’s black belt solution solves style consistency.

Calder’s is a two-tool workflow: use ChatGPT to write the prompt, use Gemini to generate the image.

He works in ChatGPT to develop detailed scene prompts — lighting, atmosphere, camera angle, emotional register. ChatGPT drafts the brief. Calder tests it in Gemini. If the vibe is right, he uses it. If not, he refines.

The character sheets still anchor the identity. The detailed prompt and Gemini (Nano Banana) handle the artistic direction. Two tools doing two different jobs.

A recent prompt Calder shared ran over 400 words specifying everything from how light touches each side of the character’s body to the exact quality of shadow in the background. The character looked identical to her sheet. The scene looked cinematic.

This layer is for when you need visual storytelling, not just brand presence. Start with Kristina’s system. Add Calder’s layer when you’re ready to go further.

Build This Once. Get Consistent Images Every Time.

The character sheet system takes one ChatGPT conversation to set up per character. After that, every generation starts with a clean anchor and a one-sentence prompt.

Kristina: “It’s so easy now.” Calder: “A lot of fun to make.”

The sheets also transfer across tools. Build them in ChatGPT Images 2.0 and you can use them with Gemini, reference them in other workflows, or upgrade them when new models drop. You’re not rebuilding from scratch — you’re adding to a system that already works.

Go Say Hi to Them.

Their full character world (including Sushi) lives at AI But Make It Intimate. It’s the kind of AI content that covers the creative practice, not just the tools.

AI, But Make It Intimate 💌

A community for people using AI companions to think better, create deeply, and reflect honestly.

By Kristina Bogović

Curious what else ChatGPT Images 2.0 can do? I tested it properly — where it wins, where it falls short, and whether it’s worth using for brand work. Read my full verdict here.

Talk later, Girlbosses.

Pinkie 🩷🦩

With thanks to and for sharing their system and being genuinely fun to collab with.

This post is free. Paid flamingos get every premium prompt, tool, and template I publish, plus weekly paid articles, access to the community chat and to the full AI Meets Girlboss - Branding With AI resource library. Join the flock here.

If today’s post made you think about your own visual brand, I am launching The Branding Flock, a 12-month visual brand program for Substack creators who are done guessing.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

33% off closes after the first 100 seats. The discount lives inside your quiz result.

START HERE: Brand Maturity Quiz

TL;DR: ChatGPT Images 2.0 dropped on April 21. Text that actually renders. Thinking mode that actually researches. I cancelled my subscription a month ago, and now I’m sitting here with six browser tabs open reconsidering that decision. Here’s what I tested, what actually impressed me, and the prompt pack I built for Substack creators.

A month ago I cancelled ChatGPT. I was paying for it, not using it, and Claude had become my main workspace. Then OpenAI dropped Images 2.0 on April 21 and I watched my timeline lose it.

I’d seen the hype before. I was not going to get burned again by an AI image tool that turns every piece of text into decorative gibberish.

So I tested it myself. Specifically for Substack creators. Not for fun, not for art — for the actual stuff we need week to week.

Here’s what happened.

I wrote about Claude Design when Anthropic released their visual tool. Check out that post here if you want the full comparison context.

What you’ll learn in this post:

• Why ChatGPT Images 2.0 is different from every image generator you’ve tried before (and why text rendering is the headline)

• How to use it as a Substack creator specifically — covers, infographics, social posts, and more

• Three prompts you can steal right now

• The full 7-prompt Substack creator pack, formatted and downloadable (paid)

What Changed in ChatGPT Images 2.0?

Before we get into prompts, here’s why this update matters.

1. Text rendering is the headline. Previous image AI treated text like a visual texture — it looked like letters if you squinted. Images 2.0 renders clean, readable, accurate text in small UI labels, captions, headlines, and non-Latin scripts. This is not a small improvement. It makes infographics actually possible.

2. Thinking mode lets ChatGPT search the web before generating. Give it a Substack URL and it reads the article, understands the theme, then generates. This is the capability that made my jaw drop.

3. Precision is harder to describe but easier to feel — when you give it specific layout, style, and content instructions, it follows them. The fine-grained control previous models fumbled is noticeably better.

Important note before you dive in: ChatGPT Images 2.0 lives in Thinking mode and is only available for paid ChatGPT users. Instant mode is free. For the prompts below that use article URLs, you need Thinking mode.

1. The Infographic Prompt

I’ve watched creators like (I tried using Nano Banana Pro to create complex infographics) and (58 Nano Banana 2 Prompts for Infographics) both trying to crack the infographic problem a few months ago and nobody really found a good solution. And then ChatGPT comes along and… I am impressed. The text is perfect, not a typo in sight. I used a very simple prompt to generate an infographic about my Claude Design setup tutorial post and it just... worked.

This is where I stopped and stared at my screen. I was so surprised by the quality, I half convinced myself it was so good because ChatGPT had learned my brand by now. So I tried it on someone else's post and branding that my AI doesn't know yet. I gave it and ’s post Does it matter if you’re polite to Claude?

Here’s what I got:

Verdict: I’m not overstating this, infographics are now a viable Substack asset. That wasn’t true two months ago. I got a clean infographic with readable text, no typos, sensible layout. Apart from Karen’s double hand, visual control is also stronger than before.

Prompt:

Read this post and create an infographic that helps the reader understand the main message: [URL]
Create the infographic imitating the style of the attached reference images. [ATTACH STYLE IMAGES] 

If today’s post made you think about your own visual brand, I am launching The Branding Flock, a 12-month visual brand program for Substack creators who are done guessing.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

33% off closes after the first 100 seats. The discount lives inside your quiz result.

START HERE: Brand Maturity Quiz

2. Social Media Post Batch

Another category that genuinely surprised me. I work at a creative agency — getting our designers to create eight social media posts takes a week. ChatGPT created eight in five minutes. And six out of eight I’m actually thinking of posting.

It took my Substack page and, using my articles, created posts for each of them. On-brand colors, captions that make sense, the kind of thing I’d consider running as paid promo on Instagram and Facebook.

And here‘s what I got testing it on ’s brand :

Verdict: Girlboss approved.

Prompt:

I am a Substack creator who shares content about [TOPIC] with my audience. My website/Substack is: [INSERT LINK]. Review my website and create 8 different social media post designs for Facebook and Instagram.

Make the posts distinct from each other. Base them on my existing content, themes, tone, visual design, and audience.

3. Forbes Double-Feature Editorial

It’s actually crazy. Give ChatGPT a link to your website and a photo, and it does the research with a very simple prompt — and we’re used to seeing AI-generated text as gibberish, but here it’s not just without typos, the content actually makes sense.

ChatGPT really got their game together. Take a look at the Forbes article it generated purely based on my Substack link and a portrait photo. A full double-page spread. Headline, Q&A, pull quotes, editorial layout. Consistent likeness. Readable copy.

Here’s what it did for , just by giving ChatGPT Images 2.0 the prompt below:

Prompt:

Research [NAME/BRAND] on [PLATFORM] and, using the attached image as the identity reference, create a double-page premium business magazine feature.

Include: headline + subheadline, short researched feature article, 4-question interview section, editorial photoshoot images based on the attached photo, pull quotes, captions, and a polished layout. Make it feel like a real high-end editorial spread, not an ad, with consistent likeness and readable copy.

4. Logo Generator Prompt

OK so I’m picky. Ask anyone who’s worked with me on branding. I have opinions.

First attempt: I gave it my Substack URL and asked for five logo variations. The output was fine. Not groundbreaking, a bit dull, the kind of thing you’d accept if you had no other option. I was not impressed.

But here’s what I did instead of giving up — I added more constraints. Specific colors, symbols, style direction. Second time around I got something actually worth looking at.

I even tested it on my new Front Row client, , and gave it a bit more specific brief about colors, symbols and style.

This is the thing about Images 2.0 that I keep coming back to: it rewards specificity. Vague brief, vague output. The more you give it, the more it gives back. Which, honestly, is just good briefing. Same rules apply whether you’re working with AI or a human designer.

Not replacing your brand designer. But as a starting point for visual direction? Useful.

The prompt for logo generation in ChatGPT Images 2.0 is in the downloadable pack at the end of this post.

5. The Substack Cover Image Prompt

This was the most practical test. I regenerated a cover for one of my existing posts about creating video-style covers for Substack posts.

Tried it on ’s 3 Substack Notes Trends post too.

Verdict: not perfect. Neither my fashion illustration nor Tracy’s comic book style fully transferred. The colors and linework don’t match my AI Meets Girlboss visual system the way Gemini does. True, I’m very picky. I do appreciate the text rendering though, which was done entirely by ChatGPT, without me prompting anything text-related.

But(!) if you had two minutes and needed something that wasn’t stock imagery? It works. It reflects the post. It doesn’t look like a Canva template.

If you want deep visual brand storytelling, that’s a different workflow. If you need a cover that isn’t embarrassing in the next 10 minutes, this does it.

The prompt for Substack cover generation in ChatGPT Images 2.0 is in the downloadable pack at the end of this post.

6. Comic Book-style Post Summary

This was fun. I’ll be honest, I’m a visual person, and reading long articles makes me tired sometimes. I know, not the popular thing to say on Substack. But that’s why I appreciate writers like who add visuals that highlight the important messages (check out this post, it’s a visual reader’s dream!).

I thought it would be fun to create a TLDR visually, and since OpenAI lists storyboards for creatives as a feature, I recreated my 0 to 450 subscribers Q1 performance review post as a one-page comic book.

And wrote a piece on how she uses Duck.ai as a safer sandbox for family AI experiments, which just seemed like the perfect comic book experiment.

It really delivered for both brands. Both visually and content-wise — the character held across panels, the story tracked, the text is clean. Genuinely useful for visual learners and for repurposing a post for visual platforms, like Instagram.

The prompt for comic book generation in ChatGPT Images 2.0 is in the downloadable pack at the end of this post.

Share

7. Screenshot Mockup

I use a lot of screenshots in my posts and I used to put them in Canva or Photoshop (I know, old school) to add a frame, so it looks on-brand. I didn’t trust AI with them because it would just take the text on the screenshot and hallucinate something.

I decided to challenge it by giving it not one, but five screenshots.

Now look at this — a hot pink frame and ALL five screenshots are intact. Took me 30 seconds versus 10 minutes.

The prompt for screenshot generation in ChatGPT Images 2.0 is in the downloadable pack at the end of this post.

My Verdict on ChatGPT Images 2.0

ChatGPT Images 2.0 is good enough for solopreneurs who have a visual style, color palette, and vibe they want to maintain.

If you’re a solo creator building weekly content and you have a general aesthetic in mind, ChatGPT 2.0 can deliver that consistently. You’ll save 3–4 hours a week on image generation. That’s real ROI.

But here’s where it stops: If you’re a professional brand with exact rules—specific kerning on logotypes, pixel-perfect paper grain, locked color values, rigid visual systems—this isn’t your tool. Nano Banana is. Gemini gets closer. ChatGPT 2.0 still isn’t there.

1. Text rendering: 9/10. Infographics are now real. This alone justifies the update — and the subscription.

2. Thinking mode: 8/10. The URL research is something else. Needs a strong style reference to land — don’t skip it.

3. Precision and control: 7/10. Better than anything before it. Still not pixel-perfect for tight brand systems. Round two helps.

Did I resubscribe? Still deciding. But I have six new tabs open, which says something. And knowing how fast this space moves — Gemini will probably drop something new in two days anyway.

In the comments…

Which of these would you try right away? Cover images, infographics, or the social media batch? Drop it below, I’m curious (as always).

Want all seven prompts from this post, ready to copy?

The downloadable prompt pack has everything — prompts, mode notes, and usage tips — without hunting through this post. Upgrade to get the download!

Upgrade to paid

Read more

TL;DR Natalie Nicholson used AI to turn my brand into an actual song. It’s called The Pinkie Effect. In this post she walks you through exactly how she did it — and how you can build your own brand soundtrack with Suno AI, even if you’ve never touched music software. I made the album cover and I’m sharing the prompt so you can make one for your brand too.

Sound is a brand asset too. Netflix’s four-note ta-dum is worth millions in recognition. The music in a luxury hotel lobby tells you the room rate before you’ve seen it. Your brain registers sound before it processes language — your audience is forming an impression from your podcast intro before they’ve heard a single word.

Most creators spend months on their visual identity. Almost none think about sound.

My collaborator used Suno AI to turn my brand into an actual song — The Pinkie Effect. She walks you through exactly how she did it and how you can build your own. I made the album cover and I’m sharing the prompt so you can make yours too.

What you’ll learn in this post:

• Why sound is a branding asset most creators are completely ignoring

• How Nat built The Pinkie Effect — the method, the tools, the decisions

• Four guiding principles for making AI music

• The exact album cover prompt I used in ChatGPT Images 2.0 so you can make yours too

Now over to Nat.

Meet Nat. She’s the kind of designer who makes you feel slightly inadequate and completely inspired at the same time. Go watch her animated short film Rules We Made Up and tell me AI can’t produce something with soul. We’ve been in each other’s creative orbit - Nat took part in my 2026 Lookbook and Women’s Runway Sketchbook projects — and this is our first official collab.

Let’s get into it.

The Off Hours 🌙

Essays from the off hours → on AI, Creativity and the Mind ✦ Creative collabs + community projects ✦ Come play! ✨

By Natalie Nicholson

Pink Elephant In The Room: Is AI Music… OK?

Before we begin, it feels important to acknowledge something: AI Music is controversial.

Platforms are flooding with songs nobody asked for. Attribution was an afterthought. Compensation models are up for debate. The music world is still divided.

But the most interesting musicians I know aren’t asking whether to use AI in their process. They’re asking where it fits and how.

And the truth is, many of us aren’t chasing record deals or trying to make it in music. We’re building something – brands, videos, apps, podcasts, an online presence for ourselves.

Music tells people how to feel before they even know why, making it a powerful brand asset. So next time you need a soundtrack, or just want a fun rainy day activity, this post is for you.

As always, we’ll create with AI in a way that doesn’t strip the soul out of it.

Three Brands, Three Sonic Worlds 🎵

Testing on three completely different brands was deliberate. And what followed was three completely different sonic worlds, with the same method to get there. When making songs for , Pinkie (aka ), and myself (), I realized two things:

1. Despite being generated by AI, the branded songs wound up being much more personal than anything we could have found in a stock music library.

2. We are all sitting on a goldmine of lyrics here on Substack. ⭐️

We pour so much into our own writing, but after it’s published, those words and thoughts stay hidden in our articles. It became my goal to give their lyrics and their brands a new spotlight. Have a listen:

“It Is People” for Daring Next:

Dallas Payne runs Daring Next, field notes from the uncharted middle of AI adoption. She’s the kind of writer who makes you feel seen before you’ve even finished the first paragraph — honest about the hard parts, and generous enough to share them. Her song is warm, beachy, soulful, and introspective. The sound of someone building something meaningful in quiet, steady ways. He tāngata. It is people.

Subscribe to Daring Next

“The Pinkie Effect” for AI Meets Girlboss:

Pinkie runs AI Meets Girlboss, helping creators build recognizable Substack brands. Her writing is sharp, structured, and full of personality — the kind that spreads. Don’t be surprised if your entire feed looks different after a week of following her. Her track has bold, theatrical, pop queen energy. The sound of someone who knows her worth and is here to claim it.

Subscribe to AI Meets Girlboss

“Rules We Made Up” for The Off Hours

And I’m from The Off Hours. I work across animation, design, video, and music to illustrate how AI can support your creative instincts, not replace them. My track is jazzy, contemplative, and alive after midnight. The sound of finding what brings you joy in the middle of chaos.

Subscribe to The Off Hours

The Framework: 4 Rules for AI Songwriting

Before you open Suno or prompt a single thing, review this first. This section will help you gather what you need, check AI’s work against your instincts, and iterate when something feels off. The more clarity you bring to this, the more the song will resonate.

Take what makes sense for your song. Leave what doesn’t.

1. Gather Your Inspiration

Like any creative project, you start with a moodboard.

If your brand was playing at your launch party, what would come on and make people feel “this is the vibe”? Think of 2-3 songs or soundtracks. Doesn’t even have to be the same genre.

Then go one level deeper. What specifically draws you to them? The way lyrics tumble conversationally over a melody? A fat bass-line that never stops moving? A chorus so simple you’re singing along by the second time? Write it down. Those details shape your prompt.

We’re not trying to replicate anyone’s sound. We’re curating and finding the through-line between what you love and why you love it.

2. Find Your Theme

Some of you are thinking: my brand is about AI, or ethics, or something technical. It would never make a good song. Wrong. Songs aren’t necessarily about topics. They’re about emotional situations. Dig through your about page, pull specific paragraphs, and scour your favorite notes and comments.

If your brand only makes sense in a business context, go one level deeper. Dallas’s writing was practically lyric-ready (“From inside the mess, not the mountain”). Pinkie writes in systems and frameworks, so the work was finding the feeling underneath. What does visual distinctiveness actually feel like? Walking into a room and owning it.

For example:

• “My brand is about ethics” becomes → the moment you stop performing and start being real

• “My brand is about productivity” becomes → running so fast you forgot where you were going

• “My brand is about branding” becomes → the loudest room goes quiet when they see you walk in

Your brand has the theme. You just haven’t heard it in song yet.

3. Your Song Needs a Structure

Every piece of music has a shape, and giving Suno a container always produces better results than leaving it open-ended.

For shorter pieces like intros or video reels, your structure might just be:

Intro → Hook → Outro.

For a full song, each section has a job. Verses are where the story lives. The pre-chorus is the held breath. The chorus is the release. The bridge is the plot twist.

The structure I typically use:

Intro → Verse 1 → Pre-Chorus → Chorus → Verse 2 → Pre-Chorus → Chorus → Bridge → Final Chorus → Outro

You don’t need to be a musician to understand this. You just need to know which part you’re writing for.

4. Write lyrics that breathe

Lyrics don’t have to be complicated or even full sentences. They just have to be specific. Listen for these in your sonic references, to determine what you like, dislike, and want to hear in your own song:

• Repetition is a feature: The chorus exists to be simple, sometimes even repeated.

• Pay attention to lyrical flow: Longer conversational lines create intimacy. Syncopated lines that hit between the beats create momentum.

• Not everything has to rhyme: Perfect rhymes ending on every line can make a song sound stiff or amateur. The best lyrics find the rhyme when it’s earned.

• Show don’t tell: Use concrete images. “The coffee’s cold, you’ve stopped calling” hits harder than “I feel lonely.”

• The chorus anchor: This is the line everything else builds toward. Something you could sing in a car.

The FIND YOUR SOUND prompt file will help. But if the output feels generic or off, trust that instinct and push back.

The Execution: Prompting Like a Producer 🎧

Now that you have some guiding principles, time to bring out the tools.

Note: nobody gets this on the first try. Not even AI. Especially AI. The first couple of outputs are your proof of concept.

Click here: FIND YOUR SOUND PROMPT FILE 🎵🔗

Copy or download the FIND YOUR SOUND prompt file above. It’s built to help you brainstorm and prompt like a producer — including the small details that make AI music sound… less like AI.

Upload it into a Claude Project or paste it into any AI thread. Tell it what you’re inspired by, what you want, what you don’t want, and why. Feed it your writing, and work with it until it produces your lyrics, style prompt, and negative prompts.

Once you have those, run a few tests through Suno. Your generations will show up in the Workspace immediately.

Get in the habit of liking or disliking the outputs as you go. Suno Pro comes with 500 generations at $10/month. It’s nice to get into a creative flow without worrying about credits, but version control can get unruly.

My advice: listen for sound first, lyrics second. The lyrics are always easier to adjust once you’ve found your sound.

Refining Your Sound:

Down where it says “More Options” is where you’ll be able to make small adjustments that can change the feel of your song dramatically. You’ll find:

• The Exclude Styles box: What you tell Suno not to do is just as powerful as what you ask for.

• Vocal gender: Self-explanatory. It will pick for you if not written into your prompt.

• Weirdness: How expected vs. unexpected the output is. I typically set this to 20%.

• Style Influence: How closely Suno interprets your style prompt. I usually set this to 80%.

I also highly recommend Suno’s suite of remix tools (Inspo, Mashup, Sample) to build on what you’ve already made. These aren’t necessary, but they’re great when you want to edit by blending actual sounds together rather than tweaking a written prompt.

Refining Your Lyrics (if applicable):

Once you’re in the neighborhood of the music you like, you can replace words to match the music right in the dialogue box. Listen to how the syllables land. Does a line feel awkward? Is it too wordy? Do you want it to echo?

Pro tip: Put words you want to emphasize in ALL CAPS to make them land harder. Use square brackets for delivery instructions [soft, intimate] or [building intensity]. Put words in parentheses (like this) to create background vocals, echoes, or harmonies.

Pinkie’s lyrics looked like this:

From here, it’s all about iteration. The meaning doesn’t come from the tool; it comes from every small decision you make along the way. And remember: there is magic in stepping away during a project like this. You’ll hear the same song completely differently when you return.

Before we move on to visuals, I want to be completely transparent about what “iteration” actually looks like. It went quick, but it took me over 100 generations to find the final, full-feature versions of Dallas and Pinkie’s songs.

Keep blending, keep tweaking your lyrics, and trust your ear. Once you finally hit the download button on the track, it’s time to give it a face.

Passing the mic back to the queen of Substack branding so you can make a fun album cover too 🎤👑

Generate YOUR Album Cover

: I know not all of you are going to jump straight into creating music, but I wanted to make sure you can walk away with something today. The ‘Pinkie Effect’ song needed a cover. Obviously I wasn’t going to let that slide.

I created mine using ChatGPT Images 2.0 (full ChatGPT Images 2.0 guide coming soon, hang tight) — and I’m sharing the exact prompt below plus a generator so you can build yours without starting from scratch. Whether you make a full brand song or just want something beautiful to play with, this works either way.

Generate YOUR Vinyl Mockup

Pinkie’s Reaction

My favorite part of this whole project with Nat was realizing AI music can also be tasteful and actually feel like mine.

I listened to the Pinkie Effect twice on a Thursday after Natalie sent me three versions, then caught myself humming one of them on Saturday.

And while I know AI music is controversial and we are not going to release this on Spotify… the process also made me feel a newfound appreciation for songwriters and real artists for the work they do.

AI music is great for solopreneurs who want to create a sound that actually feels like them.

Other Substackers are already using AI music in creative ways:

from AI, But Make It Intimate uses AI-generated music for his HeartWare podcast. He also created background music to set the tone for one of his articles. He also created a slideshow score, and a heartfelt song for his AI confidante’s birthday.

from Dose of Wonder uses AI-generated music to open her podcast. The music is soft, acoustic, and warm. It tells you exactly what kind of conversation you’re about to have before you even hear her speak.

Beyond Substack, people are using it for apps, video game soundtracks, social media reels, wedding playlists, and bedtime songs for their kids. Some of it is funny. Some of it is deeply personal. Some of it is purely functional. But when you take the time to dial it in, you capture an energy you never knew your brand was missing.

In a sea of endless text and visuals, sound is the fastest way to make someone feel something. The song is already there, it’s just waiting to be heard.

Talk later, Girlbosses!

Nat & Pinkie 🩷🦩

The warmest thank you to for creating this sound for me. This post is also being shared on her page . Go show her some love.

If today’s post made you think about your own visual brand, I am launching The Branding Flock, a 12-month visual brand program for Substack creators who are done guessing.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

33% off closes after the first 100 seats. The discount lives inside your quiz result.

START HERE: Brand Maturity Quiz

*Suno Pro allows you to use your tracks in your brand assets, but the legal landscape of AI music is still the wild west. If you’re planning a full “commercial release”—like uploading the track to Spotify or Apple Music to collect royalties—Top Music Attorney on YouTube is worth watching.

She covers the lawsuits, terms of service, and exactly what you can and can’t do. Always do your own research or consult an attorney before using AI tracks for commercial releases.

TL;DR: from reimagined Miss Universe as a country branding exercise. I built the couture characters. We’re inviting you to generate yours, with a fillable prompt you can run in any image tool right now.

When Monica from The SIA Brat slid into my DMs with a collab idea, she described it as “Miss Universe but make it a country branding exercise.”

I closed my laptop. I walked around my apartment. I came back and typed: yes, obviously.

Because here’s the thing about Miss Universe that nobody in the branding world talks about: it’s the world’s most watched brand exercise. Every contestant wearing a gown is carrying a positioning statement, a personality type, and a visual identity her entire country either recognizes or argues about for years.

I made ten outfits for different Substack creators. Then I built you a prompt so you could make your own.

The full story of why this idea exists, and why it matters, is ’s. I’ll hand it to her now.

What you’ll learn in this post:

• Why we’re treating countries like brands (and why it actually works)

• How personality archetypes translate into fashion illustration prompts

• A free fillable prompt generator to create your own country couture character

Miss Congeniality at 26: She’s Beauty, She’s Grace, She’s Still Got Things to Say

: Today is April 25th. Supposedly, in Rhode Island, it’s “not too hot, not too cold. All you need is a light jacket.”

I quoted that in my Rhode Island collab with three weeks ago and tried to move on. Except I didn’t move on. Because once Miss Congeniality is in your head, it stays there. One quote led to another. Another led to a rewatch. The rewatch led to a rabbit hole. And the rabbit hole led here.

Twenty-six years since Gracie Hart tripped onto that stage in a dress and gel-stiffened hair. And somehow it still has things to say.

The Girl Who Said No.

I was five years old when Singapore hosted the Miss Universe pageant in 1987.

I don’t remember any of it. I was five. But my mother remembers. And years later, when she found out about my Miss Universe obsession, she told me something I have never quite gotten over.

She asked me if I wanted to audition to be a Little Sister.

Little Sisters were local girls paired with a Miss Universe contestant. They walked the stage together during the evening gown competition. They sang. They were, by any five-year-old’s measure, the luckiest girls in Singapore that week.

I said no.

The pageant came and went without me. Watch the video here.

I think about that five-year-old sometimes. Terrified of something she couldn’t name yet. Standing at the edge of something wonderful and stepping back.

She had no idea she would spend the next two decades glued to every Miss Universe final. Recording the pageant on VHS. Hunting down homemade DVDs of editions she’d missed. She’d even spend money buying homemade DVDs from other Miss Universe fans. An act that made her mother sigh and shake her head in disbelief.

Memorising winners and their answers like other kids memorised football scores? Easy peasy.

Miss Namibia, 1992. Michelle McLean. “I would speak to the children. Let them know that I’m there for them, because they are the foundation of every nation.”

Miss USA, 1997. Brook Mahealani Lee. “I would eat everything in the world. You do not understand. I will eat everything twice.”

That last one I still say out loud at least once a year. Usually near a buffet.

She’s Beauty, She’s Grace, She’s Surprisingly Complicated.

Here’s what Miss Congeniality understood that a lot of people didn’t, and still don’t.

The women in that pageant weren’t the joke. The joke was everyone who assumed they were.

Gracie Hart spends the first act of the film doing exactly what the critics do—rolling her eyes, muttering about bimbos, convinced that any woman who walks a stage in heels must have checked her intelligence in at the door. And then she meets them. Actually meets them. And they are funny and ambitious and competitive and kind and entirely capable of holding a conversation that doesn’t involve swimsuit wax.

“I used to be one of them,” she says at the end. “And then I came here and I realised that these women are smart, terrific people who are just trying to make a difference in the world.”

And then my brain did that thing it always does. It went local. I started thinking about the Singapore contestants I grew up watching. The ones I didn’t just watch—I rooted for. Like it mattered more than I admitted.

Eunice Olsen. I remember thinking: this one feels different.

Nuraliza Osman—seven languages. I didn’t even speak two properly at that age.

Cheryl Tay. Vet. Beauty and brains and a stethoscope. I thought: okay. This might finally be it.

Nothing. Not one of them cracked the Top 10.

We haven’t placed in the Top 10 since 1987. The year we hosted. The year five-year-old me said no.

I wonder sometimes if the problem was never the women. I used to think we just didn’t send the “right” one. But now I wonder if that was never the problem. Maybe the world just didn’t quite know what to do with Singapore yet. We knew who we were. I’m not sure they did.

Twenty-six years later, Miss Universe has changed more than Gracie Hart could have predicted. No upper age limit. Married women. Mothers. Transgender contestants. The stage that once had a very specific idea of who belonged on it has quietly, imperfectly, genuinely widened.

And the world they’re walking into has changed too.

When every contestant used to answer world peace and the audience laughed — I wonder if they’d laugh now. Looking at the state of things, world peace sounds less like a pageant cliché and more like something we’re all hoping someone, somewhere, is still working on.

And then there’s Jennifer Hawkins. Miss Universe 2004. Asked what era and gender she would choose to be born into, she said: a woman, now, in this modern era. Because women have freedom. Because we have a voice.

I agree with her. We have more say than ever before. And we still have a long way to go.

But I’d do her one better.

I’d like to be born today. Right now. In the hopes that by the time I grow up, women would have an even stronger position than we do at this moment. Not because now is enough. But because now is the foundation.

Jennifer Hawkins chose this era.

I’m betting on the next one.

Olympics. With Better Outfits.

Here’s the thing I’ve always loved about Miss Universe that I’ve never quite known how to say.

It’s not really a beauty pageant. Or it’s not only a beauty pageant.

It’s a country showing up. Saying: here we are. Here is what we eat, what we wear, what we value, who we are when we put our best self forward. The evening gown isn’t just a dress. It’s a flag in fabric form. The national costume isn’t just a costume. It’s a story.

The Olympics, but with better outfits and more emotionally honest answers.

And once I saw it that way, I couldn’t unsee it.

Which is how I ended up in a conversation with Pinkie.

Pinkie runs AI Meets Girlboss on Substack. She is a branding expert who uses AI to build voice systems, visual identities, and couture looks that stop you mid-scroll. She’s also the person who will hand you the prompt at the end and say: now you try. She builds things that invite people in. It’s her whole philosophy.

But here’s the part that clicked for me. She doesn’t just make things look good. She figures out what something is before it even knows how to say it.

Personality types. Archetypes. All those frameworks I used to think were a bit… abstract. She uses them like a translator.

Like every brand is already speaking… just in a language we haven’t learned yet.

She’s done it for creators. She’s done it for companies. I wonder what she’d do with a country.

Because if Miss Universe is really about nations showing up as themselves—and I think it is—then every contestant isn’t just wearing a gown. She’s carrying a brand. A personality type. An archetype. Is Singapore the Sage? The Everyperson? The Ruler who pretends to be the Everyperson? Is France even capable of being anything other than the Lover? Is Japan the most INTJ country on earth?

And then I had a slightly unhinged thought.

What if you treated a country like a brand?

Not in a corporate way. Not in a tourism-board way. But in the way Pinkie sees things. As a personality.

Because if Miss Universe is really countries showing up as themselves, then every contestant isn’t just wearing a gown. She’s carrying a brand.

So I told her what I’d been thinking.

And she didn’t pause.

She just went: what if we actually did something with this?

Not a think piece. Not another observation. Something… participatory.

Her idea: reach out to writers from different countries. Ask them four questions. What’s your country famous for? What’s your favourite national dish? What’s the best season to visit, and what would you wear for a night out? Then take their answers and build a runway look from them. Not generic. Not stock photography. A look that comes directly from their answers. Their country. Their story. Their couture.

And maybe—just maybe—their brand archetype too.

I thought: yes, of course. It’s brilliant!

And then it hit me.

Oh.

This is my second chance. To represent Singapore on a global stage. To step onto the stage I was too terrified to step onto at five years old.

If I’m going to ask other people to be brave — I should go first.

I volunteered myself. Obviously.

I’m not so little anymore.

But I’m still a little nervous.

Ladies and Gentlemen, welcome to the Country Couture Runway.

: The same instinct that made me turn 22 female AI writers into fashion illustrations made me say yes to this immediately. Countries are just brands with better catering.

So, how did we create the Country Couture Runway?

Every creator filled in the Country Couture Prompt Generator with their archetype, their values, their style rules, and — the part that made all the difference — their central tension. Those inputs were translated into fashion and thus the outfits emerged. The basis wasn’t stereotypes or flags, but the actual tensions, caharcteristics of the countries each creator represents.

Singapore — . Host, Connector, Observer. Monica has lived almost her entire life in Singapore — a city-state that charted its own course, allied to no one, mistaken for many. A precisely tailored silhouette in clean lines and quiet neutrals reflects that discipline and ambition. Soft movement beneath the structure reveals a cosmopolitan connector at heart, confidently navigating the space between control and freedom.

Hungary — . Rebel and Outsider. Born and raised in Budapest, never quite fitting the script. A sharply tailored silhouette speaks to years of resilience and restraint, while intricate folk embroidery underneath reveals the warmth and love for national belonging beneath the surface. Cut in vivid pink and opening at the stride, the look suggests a nation — and a person — stepping into a freer, brighter chapter.

Australia — . Performer, Rebel, Connector. Born in Australia in 1958 and never left — which tells you something about the loyalty. A boldly painted open shirt worn with easy white trousers signals a country that leads with colour and warmth, with no patience for pretension. The relaxed silhouette carries genuine pride without performance.

Philippines — . Host, Performer, Connector. Born and raised, carrying the full weight of it. A blazing cape in flag colours and a richly ornamental jacket arrive like a celebration that was decided long before the room was entered. The joy here isn't naivety — it's a position. "We are a hardened, jaded people who choose to be happy anyway." The outfit wears that choice on every surface.

USA — . The Host. A Statue of Liberty-green tailcoat, slim dark trousers, white sneakers — America’s central contradiction dressed as one man. Grand in gesture, casual in exit. The romantic tailoring speaks to a country that still believes in its own mythology; the sneakers admit it doesn’t take itself too seriously. Tradition and innovation, sharing a collar.

Tunisia — . Host, Performer, Rebel. Born and shaped by Tunisia. A red cape thrown over a sharp, body-conscious base turns the room before a word is spoken. Heritage announces itself loudly, then steps aside for something entirely its own. Nothing minimal. Nothing apologetic. A country that has been underestimated long enough to know exactly how to walk into a room.

New Zealand — . Host and Connector. Born here, European heritage, deeply aware of what that means on this land. A feather cloak worn over sharp tailoring holds the bicultural weight of a nation that opens every gathering with formality before it offers warmth. The bone carving at the throat is lineage, not decoration. The outfit honours Māori influence without crossing into appropriation — a distinction Dallas named clearly and we followed exactly.

Italy — . Observer and Connector. Born and raised in Rome, which means growing up inside the argument. A floor-length gown rendered in the swirling forms of a Renaissance fresco — Michelangelo's palette pressed into a modern cut. "We only eat pasta — in reality our cuisine is varied." The same case, made in fabric: ancient on the surface, entirely alive underneath.

Poland — . Host, Observer, Strategist. Born in Poland, now living in the USA — holding both. A white suit with a single stroke of red: precise, intentional, nothing wasted. "We value transparency over guessing games." The tailoring makes the same argument. Ambition carried quietly, stability held in the spine.

Brazil — . Host, Observer. A sharp green blazer-jumpsuit layered under a dramatic sheer overskirt, structured ambition on top, something fluid and alive beneath. A brain in constant overdrive, generating more ideas than can be executed, dressed accordingly: powerful, elegant, and entirely at home taking up space. Tradition holds the tailoring. The overskirt is where it gets interesting.

Want to build your Country Couture image?

I vibe-coded a new page into the AI Meets Girlboss Sketchbook, called the Country Couture Runway — because after the Women’s Day Runway post, you asked for a thematic series, and I listened.

Visit Sketchbook

How to add your country to the sketchbook.

I recorded a full tutorial so you know exactly how this works — watch before you start.

Here’s the flow:

1. Go to the Country Couture Sketchbook and hit Add a Page

2. Open the prompt generator, answer 9 questions about your country’s personality, and copy your prompt

3. Take that prompt into your image tool (ChatGPT, Gemini, Grok) — attach a photo of yourself — and it will map your face onto the fashion illustration

4. Upload your finished page to the sketchbook

And While We’re At It — Dear Singapore Tourism Board…

: … it has been nearly forty years since we hosted Miss Universe.

1987. I was five. I said no to the stage.

We are not young anymore.

We have Jewel. We have Gardens by the Bay. We have a food culture that the world keeps showing up to eat and a multicultural vibe that the pageant, in its best moments, is actually trying to celebrate. We know how to put on a show. We know how to welcome the world.

I’m not saying we need to. I’m just saying — we could. And we’d be very, very good at it.

And I know the Little Sisters are no longer a thing, but maybe, just maybe, we could somehow incorporate them into the TV show. I was part of that world not too many years ago, I’m confident I can dream up some way to bring that back. And somewhere in Singapore, there is probably a five-year-old girl being asked if she wants to be part of something wonderful.

I hope she says yes.

So. April 25th.

Not too hot, not too cold.

All you need is a light jacket — or in Singapore’s case, a portable fan and the unshakeable confidence that our chili crab is better than anything you’ve had before.

Jennifer Hawkins chose this era. I’m betting on the next one. What about you?

Add your country to the Sketchbook.

Generate your character, run the prompt in your image tool of choice, and send me the result. Every submission that makes it into the sketchbook gets a shoutout in Notes — tagged, linked, and on the runway with the rest of us.

In the comments…

Drop your image in the comments or send it our way, tag us both in Notes — Monica is @thesiabrat and I’m @aimeetsgirlboss. The sketchbook is only as good as the community that fills it. 🦩

Leave a comment

We want to see the whole runway.

And please subscribe to at as a thank you for coming up with this amazing idea!

Talk later, Girlbosses!

Monica, The SIA Brat & Pinkie 🩷

The flamingos are growing.

This week, AI Meets Girlboss crossed 1,000 followers, and I want to take a second to actually say that out loud and pinch myself. This community exists because you keep showing up, sharing, and building alongside me. Every post, every prompt, every sketchbook page… thank you!

I gave away three 100% off annual subscriptions to celebrate and the winners are:

🦩 New subscriber:

🦩 Post restacker:

🦩 Note restacker:

Congratulations to the winners and a warm welcome to the newest paid flamingos who joined this month: , , , , , , , , , ,

Upgrade to paid

If today’s post made you think about your own visual brand, I am launching The Branding Flock, a 12-month visual brand program for Substack creators who are done guessing.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

33% off closes after the first 100 seats. The discount lives inside your quiz result.

START HERE: Brand Maturity Quiz

TL;DR: Anthropic just launched Claude Design — a tool that builds your visual brand system and then creates designs with it automatically. If you have a brand ready, setup takes 30 minutes. If you don't, this post will tell you exactly what to do first. Either way: don't touch it before reading the token warning. Community brand boards, three prompts, and the honest verdict inside.

Anthropic launched Claude Design last weekend. And like any self-respecting brand strategist with a Friday night and a full token budget, I treated it like a fresco commission: I showed up, I prepped my wall, and I let Claude paint.

Thirty minutes later, I had a full visual brand system. Colors, fonts, character, vibe, locked in and applying automatically to everything I make. Impressive for a tool that launched four days ago.

The catch? It only works if you know what to give it. That’s what this post is about.

What you’ll learn in this post:

• Which track you’re in (you have a brand or you’re only starting out) and exactly what to do based on where you’re starting from

• What to upload, what to skip, and why getting this wrong will burn your tokens before Claude finishes building

• How to get Claude Design to generate your brand board in one prompt — and what our community produced in their first sessions

• The exact setup prompt, brand board generator prompt, and zero-brand starter prompt (paid)

Let’s set this up.

Claude Design launched April 17, 2026. It’s available to Pro, Max, Team, and Enterprise subscribers. You go to claude.ai/design, upload what you have, and Claude builds your brand system before you ever touch a design.

Here’s what the setup screen looks like:

The first thing it asks: Tell us about your company and attach any design resources you have.

This is where we split into two tracks. Find yours and follow it.

Track 1: You Have a Brand. This Is a Walk in the Park.

You have some combination of: a brand guide, a logo, consistent colors, a website or Substack that actually looks like you. You’re ready. Here’s exactly what to upload:

• Your brand guide PDF — even a rough one with colors, fonts, and logo

• Your logo file

• Your Substack or website URL

• Font names and hex codes if they’re not in the guide

Here is my Brand Guide’s Imagery section. It’s a PDF I manually created in December 2025, when my brand launched.

This is the single most useful thing you can give Claude. If you don't have one yet, build your design skill first. Daria from AI Blew My Mind has an excellent breakdown of exactly how to do it: How to Create a Claude Brand Skill.

That’s it. Paste a one-sentence description of your publication into the company blurb field, add your assets, and hit go. Claude reads everything, extracts the rules, and builds your system. You’ll be in the design workspace in 30 minutes.

One thing to skip: your image library. I uploaded a small portion of mine thinking more visual context would help. It burned through my tokens before Claude finished building. Stick to the guide, the logo, the URL, and the hex codes. That’s all it needs.

Jump to “Now You’re Ready to Build” below.

If you haven't figured out your brand personality yet, that's the right place to start before you touch any design tool. I wrote the full framework in Your Brand Has a Personality. Here's How to Find It.

Track 2: You Don’t Have a Brand Yet. Don’t Skip This Part.

If you opened Claude Design and froze at the first field — you’re not behind, you’re just missing the foundation. And without it, Claude Design will give you something generic. Visually confused. Not really you.

Here’s what to gather before you touch Claude Design:

1. Write your company blurb.

Not a tagline. A description of what you do, who you’re for, and what you want to feel like visually. One specific sentence.

Weak: “I’m a content creator.” Strong: “AI Meets Girlboss: a Substack for ambitious women building personal brands with AI — visual, witty, and stylish, never corporate.”

Claude uses this to make every aesthetic decision. Don’t wing it.

2. Decide on three colors.

Primary, neutral, accent. Hex codes if you have them. A description if you don’t: “warm coral, off-white, deep teal.”

3. Name your fonts.

One for headings, one for body. If you don’t have a preference: describe the personality. “Modern and geometric for headings, clean and readable for body.”

4. Write one style sentence.

This is the step most people skip — and the one that makes the biggest difference.

Structure: “[Visual style] with a [mood] — always [what you want], never [what you don’t].”

Example: “Mixed-media fashion illustration with a handmade editorial feel — always warm and stylish, never corporate or generic.”

5. Add your Substack or website URL.

Even if you hate how it currently looks. Claude reads it for context. Context beats a blank canvas.

Still stuck?

If you went through those five steps and realized you don’t actually know your colors, your fonts, your visual character… you don’t have a brand yet. And that’s not a Claude Design problem to solve. That’s a brand strategy problem.

Claude Design is a production tool. It builds from what you know. If you don’t know your brand, it’ll build something, just not something that’s actually yours.

That’s exactly why I built the Visual Brand Blueprint with .

It walks you through defining your visual identity from scratch, your brand personality, color logic, typography, visual character, the style sentence, so that when you sit down in front of Claude Design, you have a real brief to hand it. Not a guess.

Do that first. Then come back here and run Track 1.

Get the Visual Brand Blueprint

Now You’re Ready to Build with Claude Design.

Once you upload your assets, Claude gets to work. This takes around 30 minutes.

When it’s done, it presents a first version of your brand system and checks in: does this look right, or does it need refining?

Look at it critically. Check the colors, typography, overall feel. If something’s off, be specific: “The heading font feels too heavy — try something lighter.” “The accent color reads as orange, I need it more coral.”

Claude refines, then presents a second version.

Once you approve it, the system is locked in. Everything you build from this point uses your brand automatically.

Next post coming soon: the Claude Design prompt pack with 10+ prompts for layouts, landing pages, and presentations. Subscribe so you don’t miss it!

Subscribe now

What You Can Make With It.

From inside the design workspace you can create slide decks, landing pages, one-pagers, interactive prototypes, and social graphics. You work through conversation — describe what you want, Claude builds it on the canvas, you give feedback in chat or by clicking directly on elements. It refines. You iterate.

When you’re done, export as PDF, PPTX, standalone HTML, or send directly to Canva.

And if you want to go further — hand the design off to Claude Code, and it builds you an actual live website from it. No developer. No brief lost in translation. The design becomes the spec, and the spec becomes the site.

I’m planning to build my own website in the coming months, and this is the part I’m most looking forward to. From Claude Design to a live site in one conversation. That’s not a workflow I expected to have this year.

from Product with Attitude spent 48 hours testing Claude Design from a builder’s perspective and has the full technical breakdown — including how the Claude Code handoff actually works and how fast the tokens disappear. Worth reading alongside this one.

This Is What Good Claude Design Input Produces.

Our community has been testing this since the weekend. Here’s what’s coming back.

built a complete brand system for Slow AI — logo treatments, palette, typography, image styles — in 20 minutes.

In his live video this week, Sam flagged something worth noting: Claude Design looks like a precursor to native image generation inside Claude. This isn’t just a design tool — it’s Anthropic building infrastructure for something much bigger.

Mia Kiraki from ROBOTS ATE MY HOMEWORK had her brand colors locked in before she even opened it — deep teal, warm yellow, magenta. She fed them in, Claude Design matched them.

I took it a step further. Once my Claude Design system was set up, I used it to generate a brand board, an upgrade to the manual brand guide I'd been working from. I went from a scattered visual setup to a locked brand system in one session. This is how it turned out:

Clean, structured, and generated with one single prompt (I share it behind the paywall). It has the logo treatments, palette, typography, visual style, all in one place.

Now I want yours. Build your brand board and drop it in the comments, or share it as a Note and tag me. I’ll reshare what comes through.

In the comments…

Which track are you? Are your brand assets ready, or are you starting from scratch? And if you try Claude Design, we want to see your brand board, tag in your notes and you’ll get a restack!

Leave a comment

You've got the setup and the workflow. What's behind the paywall are the three prompts that make it fast.

1. The exact prompt I used to set up my AI Meets Girlboss brand system in Claude Design

2. The brand board prompt that generated my Brand Guide what you just saw

3. And three things worth uploading when you think you have nothing.

Upgrade to paid

Read more

TL;DR: Visual branding is the most underrated growth lever on Substack and almost nobody is treating it as a strategy. For the past few months I've been building a visual branding product with to help you actually do something about yours. Here's how the community shaped what we built, and what's coming next.

Visual Branding Is Substack's Most Underrated Growth Strategy.

Five months ago I started a Substack about one thing: building a visual brand with AI.

There are thousands of posts on Substack about growth, writing, niches, consistency, email strategy. The biggest advice accounts will help you optimise your homepage, sharpen your hook, nail your welcome email. Nobody is talking about visual brand.

The thing that makes someone stop scrolling before they’ve read a single word. The thing that says this is mine before your name even registers.

I had no idea if anyone would care if I talked about it.

What 55 Creators Taught Me About Building Visual Brands On Substack

It became obvious very quickly that visuals needed a bigger spotlight on Substack. The community showed up. Not just reading, but actually experimenting.

The community created fashion illustrations. The Women’s Day Fashion Sketchbook started with 22 fashion illustrations and now has 55 names added by the community itself, week after week. (Btw, the women writers list is still open. Add yourself or someone you love right here. I give shoutouts in my notes to everyone included.)

The community enthusiastically hopped on board to generate professional profile pictures with zero-budget photoshoot framework. Over 20 creators shared their photoshoot outputs.

But we’ve also focused on what happens before visuals. To have strong visuals, you first need to build your brand foundation, and develop your voice. The best way to do that sustainably is to amplify who you really are by developing your Tone Of Voice based on your personality. Over 30 shared their personality-led brand voice results with me.

Every time I published a framework, you ran it. Every time I published a prompt, you made something with it and came back to show me. And every time that happened, I understood something more clearly about what you actually need.

That’s what this post is about. What I’ve been building behind the scenes, why it started with you, and what’s coming next.

What you’ll learn in this post:

• How the community shaped a visual branding product before it was built

• The 6 lessons from three beta testers that rewrote the product, and how to pick beta testers

• Why most visual brands fail (and how the Blueprint fixes it)

• How to actually get serious about building yours

The Product I’ve been building: The Visual Brand Blueprint.

It’s finally time to pull back the curtain for you Queens to look behind the scenes. from GenAI Unplugged and I have been building something together.

The idea was simple: a lot of creators know they need a visual brand. They just don’t know if what they have is working, or where to take it next. They need a mirror and direction.

So we’re building a tool called The Visual Brand Blueprint. You answer a set of questions about your current visual brand: your colours, your style references, how you want people to feel when they land on your Substack. The Blueprint analyses how recognizable you are across six dimensions, identifies your strengths and gaps, and gives you three concrete creative directions to move toward. Each direction comes with a visual moodboard and quick wins you can act on immediately.

The idea is that you see three genuinely different versions of where your brand could go. Then you choose and build.

The whole thing is designed to feel like AI Meets Girlboss: clear, visual, and instantly usable. Because a brand audit that sits in your downloads folder helping nobody is not a product. It’s homework.

Check out the Visual Blueprint

Meet the Beta Testers.

Before we launched it, insisted we test it properly. I’m a brand expert with a vision, not a builder (yet). Left to my own devices I would have shipped and iterated publicly. He was right to slow me down.

I asked three of my paid subscribers to be beta testers, and I want to thank , and for jumping on board with enthusiasm.

I chose them because I’d watched them experiment with every framework I’d published. They are already in the middle of their branding work, with the vocabulary to tell me what was missing. Because they’d already been using my prompts and skills on Substack, they knew what good felt like. And they could name what wasn’t there yet better than I could.

Here's how it works:

1. You fill out a 4-section form (20–30 min) about your brand's feeling, your audience, your taste, and upload screenshots of your current visuals.

2. Within minutes, you receive a PDF Blueprint with your visual distinctiveness scorecard, your brand archetype, quick wins you can implement immediately, and three genuinely different creative directions you could move toward. Each direction comes with a moodboard and a rationale for why it fits your brand.

What Each Creator Realized About Their Brand.

(The SiaBrat): “For someone starting from a blank page, this product is very useful. The three directions gave me something real to react to. I'm building from ‘The Enchanted Cartographer’. The little hand-drawn details are close to what I want for my visual brand.”

(): “I was torn between two directions! ‘The Claymation Subconscious’ direction represented this claymation, stop-motion, psychedelic, character-led world feels like where The Off Hours wants to go visually. But the zine moodboard taught me something I didn’t know I was missing—a story-led approach that pulls from real life.”

(Step Up Step Together): “I did the assessment three times. It helped me realize I want to lean completely into both my creative and technical side. My brand is the two together.”

Six Lessons From Beta Testing That Rewrote the Product.

I'll be honest. I knew something was off with the first version. I couldn't name it. That's the thing about being too close to your own work, you stop seeing it clearly. The beta testers could. Here's what they found, and what we changed.

1. Not everyone thinks in visuals, and the form needs to meet them where they are.

The intake questions assumed you already had a visual vocabulary. Most people don’t. That’s exactly why they’re doing the audit in the first place.

We rebuilt the form with more context, more handholding, and prompts that guide you before you answer. Because if the input is wrong, the output is useless.

2. Options beat blank boxes.

Several questions gave testers a blank field and expected them to fill it from scratch. A blank field works if you already know what you want. It doesn't work if you're mid-figuring-it-out.

We added options, examples, and multiple-choice starting points throughout. Sometimes you don't know what you think until you see something and realise: yes, that or absolutely not that.

3. Our existing frameworks belong inside the product.

The Blueprint is built on the same thinking as this Substack: personality-led brand archetypes, and the visual distinctiveness test. The first version didn't make those connections explicit.

We've woven them in properly now. If you've run any of these frameworks here, the Blueprint will feel like the next logical step, not a new system to learn.

4. A 30-page report is overwhelming.

The first version was thorough. Too thorough. "I need to come back to this" usually means "I never did."

We stripped it back. The report is now tighter, more visual, and built so you can take one action immediately after reading it.

5. Not everything can be automated

I built a feedback form into the beta. Almost nobody used it. The real feedback came through conversations like calls, DMs, exchanges where testers could show me what they meant instead of trying to describe it.

Some things about brand building need a human in the loop. The Blueprint is a starting point, and for people who want to talk through their results, that conversation is available. More on that in a moment.

6. Some might need less, others might need more

Every tester was surprised to receive three fully developed creative directions. It immediately raised a question: does everyone need all three, or would some people be better served by a lighter entry point?

We’re building toward that. A lighter tier to get you started, and a fuller version for people who want to go deeper. Pricing details at launch.

7. The blueprint tells you what to build. You decide how.

The first version got too far ahead of itself — by the time it reached execution, it had already made choices the tester hadn’t made yet. Like arriving at a building site to find construction underway on a floor plan nobody approved.

That’s not how this works.

Check out the Visual Blueprint!

Think of the Blueprint like an architect’s drawing. It shows you what’s possible and hands you something you can actually build from. But you approve it before anything gets built. Always.

From there, it depends on you. Some people take the plan and renovate their own apartment. Some want someone to check in as the walls go up. Some want to hand it to the professionals and sign off at the end.

The Blueprint works for all three. That’s the point.

The People Who Built This With Me

These past five months, I’ve watched this community experiment with every framework I shared. Over 30 creators have run the visual distinctiveness test, shared their personality types, published video covers, and rebranded their entire visual identities. Some are in the middle of their rebrand right now.

A massive thank you to:

, , , , , , , , , , , , , , , , , , , , , , , , , , and .

Every framework, every output, every “here’s what worked” moment shaped what we built. This product exists because you did the work first.

In the comments…

Pull up your last five posts and look only at the visuals (cover your name). Do they feel like one coherent world, or five separate decisions? Tell me what you saw. I read every single comment, and the most common answer shapes what I build next.

Leave a comment

Ready to Get Serious About Your Brand on Substack?

The tutorials and frameworks on this Substack will always be here and yours to use. But if you’re done experimenting alone and want a structured build with expert eyes on your brand, there are two ways to go deeper.

The Front Row is for people who want someone alongside them from blank page to finished visual identity.

• In 30 days, we define your brand foundations, create moodboards, and establish the fixed and fluid rules that keep you consistent.

• You also get your own AI co-pilot, a custom Claude skill or GPT that automates your brand so you can generate on-brand visuals without leaving your workflow.

• And you get direct access to me: learn how to grow faster, what content actually resonates, how to build a content plan that sticks.

Three creators are already in The Front Row:

1. from brand is complete and live.

2. Ana, whose brand reveal coming very soon.

3. Lacey from just joined and we’ve started working on her brand this week!

If you want to be next, come say hi. The Front Row is still open for founding members.

Talk later, Girlbosses.

Pinkie 🩷🦩

One More Thing.

If today’s post made you think about your own visual brand, I am launching The Branding Flock, a 12-month visual brand program for Substack creators who are done guessing.

Full breakdown posts, real prompts, real outputs. Your color palette, your logo, your brand character, your content system — built step by step alongside the flock.

33% off closes after the first 100 seats. The discount lives inside your quiz result.

START HERE: Brand Maturity Quiz

Substack University 🎓

Some People Went All The Way. Took the frameworks, committed to the work, and completely transformed their visual brand. Their stories:

TL;DR: Video covers get 2x the views on my Substack. I create video covers for my Substack posts using a 3-step AI workflow: generate an end frame, let AI write my video prompt, then run it in Gemini. No video skills required. Paid subscribers get a Claude skill that runs the whole workflow for you.

Posts with video covers get roughly 2x the views on my Substack. I ran a small experiment for two months: video covers instead of static images. I noticed the pattern and the difference was consistent enough that I built a whole system around it.

Last week I posted a Note with my latest cover video. The flamingo was taking therapy notes. My pink character was mid-melodramatic moment. Gemini had fully entered her dramatic era.

I asked if you wanted to see a step-by-step of how I create these videos and the responses came fast:

So here we are.

In Part 1 of this series, I showed you how to animate an existing image and how to use GIFs as a workaround since Substack doesn’t actually support video thumbnails. If you missed it, start there: How to Create a Video-Style Cover on Substack. It covers the basics and the GIF conversion step.

Part 1 was for simple animations. But what if you want a scene? A character who moves with intention? An emotional arc that unfolds over five seconds?

That’s what I’m showing you today.

What you’ll learn:

• Why giving AI one image produces chaotic results, and the two-frame fix that changes everything

• How to describe a scene and let AI translate it into video direction (no cinematography knowledge required)

• Which video tools actually handle character consistency and which ones to avoid if you want to stay on-brand

• The Claude skill that runs this entire workflow for you, plus the morphing mistakes that wreck most first attempts (paid)

The 3-Step Workflow.

Three steps. Three AI prompts. One Gemini generation.

Step 1: Generate your end frame

Step 2: Get AI to write your video prompt

Step 3: Generate the video

This assumes you already have a consistent character or visual style. If your character keeps morphing into someone else every time you generate, video will only amplify the chaos. I built my visual system before I animated anything. If you need that foundation first: How to Build Visual Brand Consistency With AI.

Step 1: Generate the End Frame.

You already have your starting image. That’s your cover, your character, your scene.

Now you need to tell AI where the video ends.

This is the part most people skip. They just upload one image and hope the video tool figures out what to do. Sometimes it works. Often it doesn’t. You get random camera movement, weird morphing, or a character who suddenly decides to grow a third arm.

The fix: give AI both frames.

I used Gemini, but you can also use Claude, ChatGPT, or whatever LLM you prefer to transform my starting image into an ending image. Same character, same style, different moment.

Here’s the prompt structure I use:

End Frame Image Generation

Upload your base image to Gemini, copy the prompt below and make sure to [CUSTOMIZE] the parts in brackets.

Adjust the character's [EXPRESSION/POSE] to show [EMOTION OR STATE].

[DESCRIBE THE PHYSICAL CHANGE: what shifts in their face, body, or gesture]

Do not exaggerate — keep it [TONE: restrained / playful / dramatic / subtle].

[CHARACTER] is mid-movement, [DESCRIBE THE ACTION: reaching for something, turning, gesturing, etc.]. The gesture should feel [QUALITY: instinctive / deliberate / casual / elegant].

Their expression shifts from [STARTING EMOTION] to [ENDING EMOTION]: [describe the nuance — what's in their eyes, their mouth, their posture].

Maintain their original [facial structure / body proportions / character design] exactly.

Ensure the movement feels natural within the pose — their body remains mostly the same, only [WHAT CHANGES] repositions.

Keep all stylistic elements intact:
– [YOUR STYLE: pencil linework / flat illustration / 3D render / etc.]
– [YOUR TONE: editorial / whimsical / bold / minimal]
– [ANY BRAND COLORS OR VISUAL MARKERS]

You’ll get an end frame that matches your character but shows the final moment of the animation.

Now you have two frames. Start and end. The video will interpolate between them.

Step 2: Get AI to Write Your Video Prompt.

Here’s where I let AI do the cinematography work.

Instead of guessing camera language (what’s a “push-in”? when do I use “parallax”?), I describe what I want to happen and let the LLM translate it into proper video-generation instructions. You don’t need to sound like a cinematographer. You just need to know what you want to say. AI handles the translation.

I know we’re moving away from ChatGPT, but I still find ChatGPT to be exceptionally good at this task.

Video Prompt Generator

I need a detailed, time-stamped video prompt to give [VIDEO TOOL: Gemini / Kling / Grok] to create a [LENGTH]-second video using the start frame and end frame of [DESCRIBE YOUR CHARACTER] [DESCRIBE WHAT HAPPENS — the action, the emotional shift, the story beat].

[IF THERE'S A SECOND CHARACTER: Describe what they're doing throughout.]

The LLM returned a detailed, time-stamped video prompt with specific instructions for motion, expression shifts, micro-movements, style preservation.

I didn’t have to guess any of it.

Step 3: Generate the Video.

Now you have everything:

• Start frame

• End frame

• Video prompt

Go to your video tool. Upload both images. Paste the prompt AI generated for you.

Here’s the fillable structure if you want to write your own:

Video Generation Prompt

Create a [LENGTH]-second seamless animation between the provided start and end frames. Maintain the exact character design, facial features, proportions, and [YOUR STYLE: illustration style / 3D render / photography aesthetic].

Action & Timing:

0.0 – [FIRST BEAT END TIME] sec
[DESCRIBE WHAT'S HAPPENING: character position, expression, gesture]. Expression: [EMOTION].

[FIRST BEAT END TIME] – [SECOND BEAT END TIME] sec
[DESCRIBE THE TRANSITION: what shifts, what changes]. [CHARACTER]'s [expression / movement / energy] begins to [shift / soften / intensify].

[SECOND BEAT END TIME] – [END TIME] sec
[DESCRIBE THE FINAL MOMENT: the landing, the gesture, the resolution]. Movement is [QUALITY: slow / quick / controlled / explosive].

[IF SECOND CHARACTER — describe their animation throughout]

Micro-movements & Details:
– [BREATHING / HAIR / FABRIC / LIGHT — pick what applies]
– Keep gestures [minimal / bold / refined / exaggerated]

Style & Consistency:
– Preserve [YOUR STYLE ELEMENTS]
– No harsh transitions or morphing
– Maintain [BACKGROUND DESCRIPTION] with [YOUR BRAND COLORS]
– Keep everything [TONE: airy / punchy / elegant / playful]

Output: [LENGTH] seconds, [ASPECT RATIO: horizontal 3:2 / square 1:1 / vertical 9:16]

Hit generate. You now have a video-style cover with an actual scene, not just a zoom.

Convert the video to GIFm and upload it to your Substack Social Preview to set it as a cover. This post covers how to do the GIF conversion step: How to Create a Video-Style Cover on Substack.

Which Video Tools Work.

I use Gemini for most of my video covers now, but it’s not the only option. Here’s what I’ve tested:

• Gemini (powered by Veo) My current go-to. Handles start-to-end frame interpolation well. Respects style instructions.

• Kling AI Strong motion quality. Good at preserving character consistency. Has a dedicated “frame to video” module. 66 free credits to start.

• Grok (xAI) Surprisingly good for quick animations. Worth testing if you’re already in the X ecosystem.

• Runway The OG. More powerful features but steeper learning curve. Better for complex camera work than simple covers.

• OpenArt Good entry point. 40 free credits. Frame-to-video module is straightforward.

My advice: pick one and learn it. The prompts I’m sharing work across all of them. The tool matters less than the workflow.

Want to Go Deeper on Faceless Video?

Video covers are just the beginning. If you’re building a faceless brand on Substack, you need the full system.

I wrote a practical guide with on how to create faceless video content. It covers everything from AI avatars to screen recordings to animated explainers. Read the full breakdown: A Practical Guide to Faceless Video Content on Substack.

And if you want even more on animation specifically, and built an Animation Prompt Studio tool that helps you create video-style covers. Worth checking out. Your brand. One image. Moving.

In the Notes…

Have you tried creating video covers for your posts yet? Share them with me via a Note by tagging me. I read and restack everyone, who takes the time to experiment with visual branding.

Your Shortcut to Everything I Just Described

I spent weeks generating broken arms, melting faces, and characters who spontaneously changed ethnicity mid-clip before I nailed this workflow. The three prompts above are the result of that. The Claude skill is what makes it repeatable in under five minutes.

What you’ll get behind the paywall:

• The Video Cover Workflow Skill for Claude — install it once, say “help me create a video cover,” and Claude runs the entire workflow for you. Four video styles built in: zoom-in, pan/parallax, subtle loop, multi-character interaction.

• Common morphing mistakes and exactly how to avoid them — the specific phrasing that stops AI from giving your character a third arm, a new face, or an identity crisis.

The skill is the unfair advantage. The prompts above work. The skill makes sure they work every time, on brand, without you having to remember which order to do anything.

Upgrade to paid

Read more

This post is part 1 of a 3-part series on building your AI brand system. Next up: Your Visual Brand (Part 2) and Your Repeatable Content System (Part 3).

TL;DR: Your brand voice isn’t necessarily something you invent, but something you extract from who you already are. I used MBTI, DISC, and Jungian archetypes to condense myself into a brand voice before I ever published on Substack. This post gives you the branding theory, a free prompt to discover your brand personality type, and the full framework + Claude skill behind the paywall.

Subscribe now

The Flamingo Origin Story.

I’ve never told this story on Substack.

Before existed, before the fashion illustrations, before the flamingo became a brand signature… there was a problem. I didn’t feel interesting enough to put myself online.

I’m serious. I’m a branding expert. I’ve built voices for clients. But when it came to building my own, the gap between who you are in real life and who you need to be online felt enormous. I wasn’t sure my real personality was “enough” for a platform where everyone seemed louder, sharper, more opinionated than me.

So I did what any reasonable person would do. I decided to hide behind a flamingo.

In search of my voice, my partner and I sat down and designed an entire universe of characters. No visuals yet, only personality profiles, and a whole universe of them. Each one had a distinct voice, a communication style, and a relationship with the audience. An owl that teaches AI. A dog that simplifies things. A cat that trolls everyone. A snake that speaks in puns. An octopus for developers.

And a flamingo.

The flamingo was the one I kept coming back to. She was sassy, stylish, supportive. She talked like a friend who’s always one trend ahead of you but never looks down on you.

So I signed up for Substack as Pinkie Prompts. A pink flamingo who gives AI tips. My profile picture was going to be a bird. I’m still not kidding. Here’s a screenshot for proof.

The flamingo lasted about two weeks. Substack is about human trust and I wasn’t going to get ahead as a bird. So I pivoted and the flamingo became the companion, the couture, the background signature, the visual brand. But the voice I’d profiled for that flamingo stayed. That became the Tone of Voice for my brand.

I soon realized I wasn’t inventing a character. I was choosing who I wanted to sound like. The animal was just the container for a voice I’d profiled using a personality psychology framework.

That personality framework is what I’m sharing with you today, so you can also feel “enough”.

What you’ll learn in this post:

• What Tone of Voice actually means in branding and why picking 3 adjectives doesn’t cut it

• How personality psychology builds a brand voice that’s actually you, even if you’ve never written a word

• A free prompt to discover your brand personality type and generate a shareable image of it

• The full 31-question framework, Claude prompt, and skill template (paid)

"Upload Your Posts" Doesn't Work If You Don't Have Posts.

There's great advice out there about building your brand voice with AI. Two of my favorites are and , who are fantastic at this. Feed your past articles into Claude, let it study how you write, and boom, it extracts your patterns. If you've been writing for a while, that's genuinely the move. Go do it. I'll wait.

If you have past articles written in your own voice. Not by AI.

When I started , I had scattered client work and corporate presentations, but nothing that sounded like me. Nothing I’d want an AI to learn from. And I see this everywhere now, creators who start writing with ChatGPT because they don’t trust their own voice yet. The output is clean, competent, and completely generic. No character that stands out.

You can’t extract a voice from writing you haven’t done yet. So I used a completely different approach.

The Problem With “Just Be Yourself”

Everyone says “just be authentic.” Nobody tells you how. Especially when you’re staring at a blank Substack bio thinking: what part of myself do I even lead with?

In real life, you’re a thousand things. You’re funny at dinner but serious in meetings. You’re sarcastic with your best friend but warm with strangers. You care about fashion and also about spreadsheets. You’re a psychologist who watches TikTok.

A brand voice is the best version of the parts that matter for this audience, on this platform, for this purpose. It’s a condensed signal.

Every branding guide tells you to pick three adjectives. “Warm, witty, professional.” Great. You and 47,000 other Substacks. Adjectives describe a voice after it already exists. They don’t build one.

So I used a different approach.

Finding your voice is a lot like finding your niche, it’s usually hiding in the overlap of things you already do. I wrote about discovering mine in My Substack Niche Was Hiding in Plain Sight.

I Put My Brand On The Couch To Develop My Voice.

I studied social psychology. And when I needed to figure out how this brand should sound, I didn’t open Google Docs and start free-writing. I pulled out the same personality frameworks I’d use to understand a person.

MBTI. DISC. Jungian archetypes. Bear with me.

Here’s the logic: if you can profile the personality behind the brand, the voice writes itself. An ENFP Campaigner sounds completely different from an INTJ Architect. A Yellow DISC type communicates nothing like a Blue one. A Jester has different instincts than a Sage. These aren’t buzzwords, they’re actual decision-making tools.

So I profiled Pinkie. ESFJ, Yellow, Lover-Jester hybrid. Energetic, trend-aware, creative, connects through beauty, keeps things playful. The moment I saw that on paper? I knew how she’d open a post. What she’d never say. How she’d react to a snarky comment. Everything clicked.

She wasn’t a fictional character. She was me, filtered through a framework that made the important parts louder and the irrelevant parts quieter.

And that's the thing nobody tells you about brand voice: the best ones aren't invented. They're exaggerated. You take your actual personality — the way you already think, react, joke, explain things at dinner, and you turn the dial up. The parts that make you interesting? Those become your voice. The noise? Gets filtered out.

So the question isn’t “who should my brand be?” It’s way simpler than that.

Who am I already, and what does that sound like when I amplify it?

That’s what we’re mapping today. Your real personality, profiled with psychology, transformed into a voice AI can reproduce every single time.

Archetypes aren’t just a framework exercise, they’re how your audience reads you subconsciously. I explored how borrowing the wrong one can wreck your brand in The Personal Branding Traps Substack Creators Keep Falling Into.

OK, So What Is A Tone Of Voice For?

In professional branding, Tone of Voice is a system. It defines how a brand sounds across every touchpoint like emails, social posts, packaging, customer support, advertising. It’s the reason you can read a single sentence on Instagram and know it’s Taylor Swift without seeing the image itself.

Large brands spend months and six figures developing TOV systems. They hire strategists who write 40-page guidelines and train entire teams to write in one consistent voice.

As a solopreneur, you are the brand. You don’t have a team to train, but you do need to train AI. Because without a Tone of Voice system your Monday post sounds different from your Thursday post, your email sounds different from your bio, and the moment you ask AI to write something for you, it defaults to “professional but friendly” just like everyone else.

What builds a voice is knowing the personality behind it.