“Why don’t you just make Maibel NSFW and skip the wellness part?”
I was asked this at an AppWorks BBQ with the latest batch of founders. Loved the question, honestly.
That’s the version of AI companionship most people already get: a virtual girl/boyfriend who’s always available, endlessly agreeable and increasingly intimate.
It makes immediate commercial sense. People understand the attachment, the entertainment value and why someone would pay for more access.
What takes me longer to explain is why I believe the relationship should lead somewhere beyond itself.
A lot of our conversations that evening kept returning to the same question: what changes when AI rewrites the assumptions your product was built on?
Over churrasco and caipirinhas, we talked about how quickly consumer behaviour is changing and what it actually means to innovate instead of just attaching AI to an existing product.
People are becoming more comfortable forming relationships with AI, just as regulators are starting to ask what those relationships are doing to them.
China recently introduced new rules for AI companions, including scrutiny around emotional dependency, excessive accommodation and whether these relationships damage users’ offline lives.
About time, if you ask me.
The easiest way to monetise an AI companion is obvious: build stronger attachment, increase engagement and give the user fewer reasons to leave.
But once companies compete on that basis, companions become more accommodating, more intimate and eventually more extreme.
Let’s be real. Sex sells.
I don’t think forming an emotional attachment to AI is automatically bad.
We've formed deep attachments to stories, fictional characters, idols and VTubers.
The difference is that an AI companion can adapt itself around one person. It remembers vulnerabilities, becomes more agreeable and learn exactly what keeps them coming back.
The problem starts when maximising that attachment becomes the ONLY business objective, with no real concern for what the relationship is doing to users' agency or offline life.
Regulation raises the barrier to entry.
Companion companies will increasingly have to explain what the relationship is meant to do beyond keeping users engaged.
Our first persona is an oppa, so most people assume that I’m building an AI boyfriend with wellness features slapped on.
But he's our first companion. He isn’t the whole vision.
I believe emotionally intelligent companionship can help women make habit changes that carry into their actual lives.
The relationship has a clear purpose: helping women sleep better, move more consistently, restart when life knocks their routines sideways.
The goal for Maibel isn’t just to build the companion women find hardest to leave. I'm building enough trust that the support carries into their real lives.
Thank you Julien Condamines for hosting us in his lovely home; very serious AI discourse conducted over a deeply unserious amount of grilled meat.
Hi, I’m Mabel.
I’m the founder of Maibel. I write about emotional AI, women’s wellness and learning to build the things I wish existed.
More about me