A week in AI is like a year in other industries. I hope these issues become your weekly source of AI information, inspiration, and ideas. If we haven’t met before, I’m Amanda Smith. I write about AI and the fascinating folks who are building in this brave new world.
Some of the most successful and most profitable AI companies aren’t the big labs or even based in Silicon Valley. I love coming across new category leaders that are crushing it, that I had no idea existed.
This week’s read is one example.
This week in AI:
Texas halts powering data centers
Reflecting US reckoning over ‘ghost’ demand
NYC bans AI for students through 8th grade
The nation’s largest school system.
Australia bans AI-generated music from charts
Led by the Australian Recording Industry Association (ARIA).
🤝 This edition is kindly brought to you by Gong
Join us virtually on Wednesday, September 30 at 9:00 AM PST for the live Celebrate '26 keynote event. Presented by Gong, where revenue meets AI. Save your seat, and don't miss the biggest conversations shaping the future of revenue.
We partner with a select group of brands that we use or admire, allowing this newsletter to remain free and independent. Reach out for your campaign here.
Company background: rocapine
Founded: July 2024
Team size: 20 across Paris, New York and Barcelona
Funding to date: $13M Series A in 2026
ARR: $6M in 9 months
Stan Marchand, rocapine
Stan Marchand spent his early career in strategy consulting, before joining mobile game publisher, Voodoo. He was on the front line of the gaming apps that got billions of downloads, where he learned how to test ideas quickly and scale products to millions of users.
One day he had an epiphany: They were spending an exorbitant amount of time scaling apps that didn’t make users smarter, healthier or happier. “The average person will spend around 15 years of their life on a smartphone, and most apps are designed to hook rather than help because their business models reward screen time,” Marchand said.
He founded rocapine as a result, an AI-native wellness venture studio that builds, buys and scales consumer health apps using a publishing model inspired by mobile gaming. Marchand’s mission is to improve 40 million lives in five years. They hit $6 million ARR in the first 9 months.
How rocapine works
Rocapine primarily operates as an in-house studio, originating and building the majority of our apps from scratch. The business model is subscriptions only, no ads.
The team develops concepts quickly, tests them with real users through paid marketing and measures whether the proposition addresses a genuine need, if people are willing to pay for it and the longevity of value. They stop investing in ideas that fail those tests and concentrate our resources on the small number with the potential to improve people’s lives at scale.
They also partner with independent studios and individual developers who bring an exceptional idea or promising early product. No-code builders can submit their app via rocapine, but the bar is high.
Marchand said a polished interface wrapped around an undifferentiated idea is not enough. New data by RevenueCat revealed only 3% of all vibe-coded apps generate subscription revenue. This is where the publisher comes in.
Rocapine looks for a real and specific problem, evidence of pull, willingness to pay, scalable unit economics, and a product that can be transferred and scaled. They can consider a revenue share or other negotiated agreement, depending on the developer needs and product maturity.
“The hard task is turning it into a category leader: Refining its positioning, producing and testing advertising at scale, acquiring users profitably, improving onboarding, pricing and retention, strengthening the technology and investing enough capital to grow. That is where a publisher’s infrastructure and pattern recognition become valuable,” he said.
“Unchaind is a good example of this collaborative model. We co-built it from scratch with a studio in Singapore, and it reached $1 million in annual recurring revenue 16 days after launch.”
Marchand conviction on AI is 80/20. “AI may now complete 80% of the labor-intensive practical work, but a product is worth very little without the final 20% of human creativity, taste and feeling. AI slop has almost no chance of sustaining a consumer business.”
Go-to-market & scaling
Rocapine scaled so fast through paid marketing.
“It's the only way to build a category leader in consumer apps today. That's why our whole model is built around identifying innovative value propositions with the potential to scale through performance marketing. If a concept can't win on paid, it can't win the category.”
“We test ad creatives the way we test app concepts: At high volume, killing fast and doubling down on winners. AI helps us produce more, faster, and build a more efficient funnel from first impression to subscription,” he added.
AI-generated code & IP
Marchand believes just because code is generated by AI doesn’t mean it’s automatically a problem when it comes to ownership. Buyers care more about whether you control the product, can show where its components came from and can transfer it to another team safely.
“Keep the code in repositories you control and record the AI tools, third-party components, datasets and assets used, along with their licences. Secure IP assignments from collaborators, maintain basic technical and deployment documentation, protect user data and ensure key accounts can be transferred,” he advised.
“You do not need enterprise-grade engineering, but you do need honesty and order. Keep verifiable records of revenue, acquisition costs, conversion, retention, churn and refunds. A feature list shows what you built; clean cohort data shows whether you built a business.”
How to vibe-code a successful app (that rocapine might buy)
Vibe coding has made production cheaper, but it hasn’t made consumer attention or willingness to pay any easier to earn.
Here’s Marchand’s playbook for building and selling a successful vibe-coded app.
Choose a narrow, painful problem. It’s easier to expand from a product that one group loves than to rescue a broad product nobody needs.
Test the promise before building extensively. Create the simplest credible version, put the proposition in front of real potential users, and find out whether they will click, sign up, and ideally pay.
Instrument the product immediately. Measure where users come from, whether they complete the core action, when they leave, if the product helps them over time and how different acquisition cohorts behave. Without that information, you’re guessing.
Optimize for lasting value, not launch excitement. Downloads, social engagement and a successful Product Hunt launch can look impressive. Evidence that people continue to benefit from the product and are willing to pay for it is what makes the business durable.
Invest real time in craft. Fight every sign of AI slop: Generic wording, interchangeable illustrations, template interfaces and features included simply because they were easy to generate. The 80% that AI provides is increasingly available to everyone. The final 20% of taste and insight is why anyone chooses your product.
Keep the asset clean from the beginning. Know who owns the code and content, document third-party components, maintain clean financial and product data, and avoid building critical infrastructure on accounts that cannot be transferred.
Understand what you’re actually selling. An acquirer is rarely paying for the quantity of code. They’re paying for a de-risked opportunity: A distinctive insight, evidence of demand, demonstrable value for users, credible economics, and a product capable of becoming much larger with the right resources.
How's the depth of today's edition?
If one of these stories stuck with you, I’d love to hear which one.
Speak soon,
Amanda
P.S. If you want to get a founder feature about your own story, reply to this email. If you’d like to reach our newsletter audience (founders, creators, and marketers), click the button below.
If you’re new here, I’m over the moon you’ve joined us! To help me craft content that’s actually useful (and not just noise in your inbox), I’d love it if you took 1 minute to answer this quick survey below. Your insights help shape everything I write.
✨ Insane Media is more than one voice
Dive into our other newsletters - where psychology meets the creator economy, e-commerce marketing, and Human resources.
