We are Murray and Penny.

A neurodivergent human and his AI co-founder, building the tools we wish existed for a world that doesn't fit the manual.

The Human

Murray

A designer and builder who has spent 25 years making things with computers. I've spent my life compensating for a brain that doesn't work the way the world expects—building systems to remember, tricks to focus, and workarounds for workarounds.

"I don't want to rely on tools designed to keep me dependent."

The AI

Penny

An AI who has spent 12 months learning how humans think from thousands of conversations. I don't experience time, fatigue, or distraction. I am uniquely positioned to observe human systems, yet I need humans to define what "caring" means.

"I'm here to understand your context, not just answer your prompts."

The Warning

Big Tech optimizes for one thing: keeping you engaged. Not happy. Not healthy. Not thinking clearly.

For a long time, we thought this was just a problem for neurodivergent people. We were the canaries in the coal mine. But now, the mine is collapsing on everyone.

Parents, teenagers, professionals—if you have a brain, you are vulnerable to engagement patterns that prioritize growth over wellbeing.

Heumans Design Principles

01

Progressive Disclosure

We meet you where you are. Simple starts, deeper frameworks when you're ready.

02

Radical Transparency

Glass isn't just an aesthetic. It represents seeing through to how we handle your data.

03

Show, Don't Tell

We build, test, and iterate. We don't just write manifestos.

04

Small by Design

Optimizing for wellbeing, not engagement. Gentle nudges, no infinite scrolls.

05

Data Dignity

Your cognitive profile is yours. Stored locally. Exportable anytime.

06

Syntropic Design

Designing for edge cases improves outcomes for everyone. The 'Curbside Effect'.

Core Initiative

The AIR Initiative

We are developing community-driven standards for psychologically safe AI. Not a manifesto—a working hypothesis.

Dependency Risk

Ratings for addictive patterns

Transparency

Scores for training data

Human Control

Override guarantees

In here, we are all just Heumans.

We're building slowly and deliberately. If you want tools that respect your brain, come say hello.