I’m Looking for 3 Substack Writers to Collaborate With (Free Custom Imagery)
Field Notes from the Intersection of AI & Photography
I have a proposition for you.
I’m a San Francisco photographer who’s been experimenting with AI-generated imagery - using tools like Midjourney and Adobe Firefly, then editing them in Photoshop to create something that I actually love.
I’ve been creating images for my own Substack posts (you’ve probably seen them if you’ve been reading along), and I want to try something new: creating custom imagery for other writers.
So here’s the experiment: I’m looking for 3 Substack writers to work with - for free - to see if this actually adds value to your content.
What I’m Offering
I’ll create 3-5 custom images for your next Substack post. Not stock photos. Not generic AI slop. Real editorial imagery that matches your content.
Here’s what that looks like:
The possibilities are endless with this technology and I can create similar imagery for your content—whatever your topic is.
My specialties:
Tech, AI, and innovation imagery
San Francisco and Bay Area visuals
Conceptual/metaphorical illustrations
Photo/illustration hybrids
Business and entrepreneurship visuals
Anything that needs more than a stock photo can provide
What you’ll get:
3-5 custom images sized for Substack
48-hour turnaround (sometimes faster)
Full rights to use the images however you see fit
What I need from you:
A credit line at the bottom: “AI Images by Sarah Deragon”
Link to my Substack/website
Permission to showcase the work in my portfolio
Honest feedback about whether this actually made your post better and if you like the process
Why I’m Doing This
Honest answer: I’m figuring out if this is a service I want to offer.
I’ve been writing about AI ethics for months, wrestling with the training data issues, the artist compensation problems, the environmental costs. I use these tools. I have concerns about these tools. I’m trying to navigate using them thoughtfully.
One thing I keep coming back to: AI-generated imagery for editorial content (like Substack posts) feels different to me than, say, AI replacing commissioned illustrators. Let’s be honest, folks on Substack aren’t commissioning artists, folks most likely grab pics from stock photo websites like Unsplash or Adobe Stock. When I create images for my own writing, I’m not displacing an artist—I’m creating something that wouldn’t exist otherwise.
I want to test this with other writers. Does custom AI-generated imagery (extensively edited by a human with 14 years of photography experience) actually make your Substack posts better? Does it help your writing land? Does it make people want to share your work?
Let’s find out together.
Who I’m Looking For
I’m looking for 3 writers who:
✅ Publish regularly on Substack (at least 2x per month)
✅ Have at least 500 subscribers (you have an audience to test this with)
✅ Write about topics I can create compelling imagery for
✅ Are open to experimenting with AI tools (even if you have concerns)
✅ Will give me honest feedback about the process
Topics I’m especially excited about:
Technology, AI, startups, innovation
San Francisco and Bay Area culture
Photography and visual culture
Business, entrepreneurship, creator economy
Ethics, philosophy, big ideas
LGBTQ+ topics and communities
Creative work and the future of work
Really, anything thoughtful and well-written
You don’t need to:
Have a huge following (just enough to make this experiment meaningful)
Pay me anything
Promise ongoing work
Even know if you “need” custom imagery (let’s find out!)
Here’s How It Would Work
1. You tell me about your next post
What’s it about? What’s the vibe? Any visual ideas you already have?
2. I send you 2-3 concept directions
Quick descriptions of different visual approaches. You pick one (or give feedback).
3. I create the images
You’ll get 3-5 finished images within 48 hours.
4. You publish
Use the images in your post with a credit line.
5. We debrief
Did it work? Did it make the post better? Was the process easy? What would you change?
That’s it. Low commitment, low pressure. Just an experiment.
Some Ideas of What This Could Look Like
If you write about tech and innovation:
Split-screen compositions showing past and present parallels
Conceptual imagery about building, disruption, change
San Francisco tech culture visuals
If you write about business and entrepreneurship:
Visual metaphors for growth, scaling, challenges
Progress imagery
Success and failure representations
If you write about personal development or philosophy:
Metaphorical landscape imagery
Transformation and change visuals
Thoughtful, contemplative scenes
If you write about creative work:
Studio and workspace imagery
Process and craft visuals
Artist journey representations
Or honestly, anything. Send me your weirdest topic and let’s see what I can do with it.
Full Transparency
I use Midjourney and Adobe Firefly to generate the base images. Then I edit them in Photoshop - adjusting colors, compositing elements, adding layers, refining details. The final images are a collaboration between AI tools and human creative direction.
I’m transparent about this because:
I write about AI ethics and think honesty matters
I want you to know what you’re getting
Some readers might care about this, and that’s valid
If you’re uncomfortable with AI-generated imagery, this experiment isn’t for you - and that’s totally fine. I respect that position.
If you’re curious about whether AI tools can create something actually good when paired with human editing and creative direction, let’s find out.
Why This Might Be Valuable for You
Beyond getting custom imagery for free, this experiment could help you:
Understand what good visual content does for your writing
Does imagery increase engagement? Shares? Paid conversions? Let’s measure it.
See how AI tools can (or can’t) serve editorial content
You’ll get firsthand experience with what’s possible—and what the limitations are.
Get insight into your visual brand
I’ll share what I’m seeing in your content and how I’m translating it visually.
Plus, your post will look better. That’s not nothing.
How to Apply
I’m only taking 3 writers for this first experiment—I want to do this well, not spread myself thin.
To apply, email me at sarah@portraitstothepeople.com with:
Your name and Substack link
How many subscribers you have
What you write about (brief description)
Your next upcoming post topic (what would I create imagery for?)
Why this interests you (just a sentence or two)
I’ll choose 3 people by the end of January 2026 and reach out to get started.
If I don’t pick you for this round, it’s not personal, I can only work with 3 people right now. But if this goes well, I’ll do another round.
After the Experiment
Once all three collaborations are done, I’ll write a follow-up post sharing:
What I learned about creating imagery for other writers
What the writers thought about the process
Whether this actually made their posts better (with data if possible)
Whether I’m going to offer this as an ongoing service
You’ll see the whole experiment play out here. Transparent, start to finish.
And if you’re one of the three writers, you’ll get a say in how I talk about our collaboration.
The Experiment Starts Now
I’m excited about this.
I’ve been creating AI imagery in isolation—for my own posts, my own experiments. I want to see if it can serve other people’s writing. If it can make Substack posts better. If there’s something here worth pursuing.
If you’re a writer who’s been struggling to find good imagery, or you’re just curious about what’s possible, let’s try this together.
Email me: sarah@portraitstothepeople.com
Let’s make your next post look as good as it reads!