Lushane

I’m Lush. I build software and products, lately mostly with AI. Underneath that I’m chasing one question: what is the simplest thing that can teach itself to become smarter, and is it the same thing life has been doing all along?

Cells do it. Animals do it. We’re starting to build machines that do it. My hunch is that it’s one principle showing up in different materials. I want to understand it, and use it to reduce suffering.

Me as a kid in a red jacket and a Pokémon shirt, standing in tall grass

I always loved the outdoors. Especially summers in Norway.

How I got here

I was a science kid. Stones, physics, how things worked. Somewhere growing up I lost it. I studied Global China Studies, went into tech, became a product manager, then a founder. It worked well enough. It always felt slightly off.

Two ideas brought it back. Richard Sutton and David Silver arguing that the next leap in AI comes from agents that learn from their own experience, not from our data. And Michael Levin showing that cells and tissues are agents too: they solve problems, pursue goals, and repair themselves, running on bioelectric signals most of biology ignored. Same question, from two sides.

I’m not a trained scientist. What I’m good at is building: turning half-formed ideas into working systems people use, and talking to the people on both sides. In a field this young, that’s a lot of the work.

A butterfly resting on my outstretched fingers in a summer field

Now

Agents that learn only from their own experience. No pretraining, no human data, never reset. Does it keep getting better without breaking?
Bioelectric signals in an open dataset of cancer cells. Does a cell’s voltage follow its own history or its neighbours’?
A virtual cell built for control, not prediction: a model that learns a cell by experimenting on it.
The foundations, properly. Python, statistics, the maths underneath, primary papers.
A few product builds at a time through my studio.

I write up what I learn here as I go. The essays are the honest record of how far along I am.

Me outside a street café on film, sunglasses on my head Me at an evening picnic in the park, looking back over my shoulder

Where this goes

A farm, close enough to a city with real talent in it that I can be in a room with the right people by lunch. Mild climate all year, if such a place exists. Summers in Scandinavia. A lab on the property. I wake up, I go to the lab, I work on this.

Red farmhouses at the edge of a green field, pine forest behind, big sky

A gravel road through pine forest under a blue sky Cumulus clouds over rooftops and a birch tree

Past

When	What
2026 – now	Co-founder, Hour and Codelitics.
2023 – 2026	Founder, Feynman Labs. Independent product studio, 0-to-1 products for SaaS, fintech, crypto, and AI clients. The patterns I kept hitting there became the thesis for Codelitics.
2021 – 2023	Growth product manager, ARIVE. Acquisition, activation, and retention for the consumer product.
2021	Product consultant, VitaDAO. Longevity research funding.
2020 – 2021	Product operations, Donut.
2019 – 2020	Entrepreneur in residence, Mapify Travel.

Laughing on a sailboat, cliffs behind A big group in national costumes from everywhere, posing in an office Close selfie with headphones on A crowd of us dressed in white on a hillside above Oslo Presenting a slide on trust to a room A polaroid of me and a friend indoors, cake in hand Dinner for three on a restaurant terrace at sunset Me and a friend at a startup event, in black and white

Elsewhere

If you work on bioelectricity, AI for biology, or anything near it, I’d like to hear from you.

Email · GitHub · RSS