I have a confession that will either make me look like a hypocrite or a human being, depending on how generous you're feeling.

I am writing this essay with the help of AI.

To be clear: Not this sentence. I always write my first drafts, and then I have a collaborative back and forth with my bestie Claude to refine my original work is original.

But the research, the scheduling, the investor CRM, the legal documentation, the competitive analysis, the content calendar, communication summaries sitting in a folder on my desktop labeled “Claude Cowork Daily Digests 💙” — AI is woven into almost everything I do at this point.

And yet: I am simultaneously building a product suite whose entire thesis is that some things should not be automated: slowing down and un-automating life is what gives it meaning.

This contradiction dawned on me as I was drafting my last essay. I’m not going to resolve in this essay. But I think it’s time to name it, because it turns out the framework I built for motherhood is the same one I’m using to build my company — and both of them are being stress-tested by the same force.

---

You’ve read, if you’ve been here since January, about my old agency framework. TLDR: I produced measurable outputs, earned external validation, kept the zoetrope spinning. And it worked across every domain I applied it to — career, athletics, learning — until it hit the open-loop system of a baby, and shattered.

Motherhood introduced me to a different kind of agency: iteration, directionality and game-masterring: facilitating the conditions for something to happen, without being attached to results. Leaning into chaos to find patterns.

My newest lightbulb moment is that this framework is also describing how to build a company 2026.

---

Startups, like babies, are not closed-loop systems.

I’ve always know this intellectually. The plan you write in January is not the company you have in December, that the feature you were most proud of (ehem, my play arch!) gets cut, sometimes your most devoted investors will often start with a “no”, and the best customers will often discover you in circumvent ways.

But knowing this and operating inside it are different things. Before I had children, I could tolerate startup uncertainty because I still believed, somewhere underneath the chaos, that the variables were ultimately manageable. With smart enough strategy, rigorous enough execution, and you could produce particular effects with reasonable certainty.

Motherhood ended that belief permanently.

What replaced it was this notion: the capacity to stay inside an open-loop system without needing it to close. To iterate without needing external validation. To trust my own judgment when the feedback is delayed, internal, and often invisible to everyone but me.

This is the framework I have subconsciously brought to my Mother of Invention founder’s journey.

---

Here’s where the AI question comes back in.

I was with a group of founders recently, and someone asked what tools everyone was using to move faster. The answers came quickly: Claude for research and writing, Cursor for code, Perplexity for competitive intel, Notion AI for documentation, Loveable for UI prototyping. The room had the energy of people who had discovered a cheat code and were delighted by it.

And me too! I nodded along, because I use all of these tools and I am genuinely, unambiguously excited to be building a company right now. Things that used to require a team of three I can do alone, in an afternoon, at a quality that would have taken weeks two years ago.

And then someone asked what I was building and I said with a bit of a cheeky tone: a diaper changing pad designed to slow parents down and make the diaper change longer.

The room laughed because I can be funny at times and the irony is real: AI quite literally eats repetition for dinner and produces a workflow.

---

The thesis of Mother of Invention — the one I’ve been building toward across twelve essays and eighteen months of product development — is this:

The most repetitive moments of early caregiving are a developmental goldmine for baby and untapped moments of stress relief and joy for parents. They are the curriculum of parenting. The 5-8,000 diaper changes between birth and potty training are not a burden or task to rush through: they are opportunities to practice presence, attunement, play, connection. These primal, often messy moments, are where a child’s brain develops and where a parent’s capacity for Gamemastering gets built — rep by rep, change by change.

The MOI Nest exists and is being manufactured because I believe this. It is a physical object designed to interrupt the efficiency mindset and cue parents toward something slower and profoundly human.

And I am building it inside a technological moment that is optimizing everything in the opposite direction.

---

This is where the Gamemaster framework stops being a parenting metaphor and starts being a founder one.

A gamemaster isn’t attached to the short term win.

She’s playing the long game. She’s invested in whether the game is being played to its fullest potential. And she does this by experimentation, adjustment, and often without external validation. She has her own internal feedback loop, built through years of staying inside hard problems long enough for patterns to emerge.

What I’m discovering is that this is exactly what the current moment in AI requires of founders who are building against the grain.

I cannot opt out of the tools because the leverage is too significant and the competitive landscape doesn’t allow it. But I can be a Gamemaster about which parts of my work I hand to AI and which parts I don’t.

Some work only produces value if a human does it.

The diaper change has to be done by the caregiver. The connection between a mother and her child cannot be automated, and neither can the thinking that leads a founder to understand, at a bone-deep level, what she is actually building and why.

AI is my Chief of Staff. But my mother/founder framework came from staying inside a problem long enough — from the disorientation and the reconstruction and the slow, unglamorous accumulation of understanding that happened between my second daughter’s birth and right now.

That part can’t be delegated to AI. But I will say: Claude has only helped me arrive at this evolution of thought with more speed.

---

I started this newsletter in January by saying I had originally planned to be in market back in January, and I wasn’t. Why? I’d discovered a core component was designed and engineered incorrectly and I made the difficult decision to redesign from the ground up.

I said my launch was delayed, but my values clarified.

I’m still clarifying and figuring out, week by week, what it means to build a product predicated on slowness, during the fastest technological acceleration most of us will ever witness, using the tools of that acceleration to do it.

I think this contradiction might be the most honest lens through which I’m trying to problem solve and innovate and scale.

The framework says: stay inside it. Don’t reach immediately for the resolution. Let the discomfort deepen into something true.

So that’s what I’m doing.

-Maggie