The layer on the human side of AI. AI gets more capable every month. We don't think progress should stop. We think the human side has to catch up.
Life GPS holds a whole life in one context — money, health, work, love, education, spirituality, family, friends and self-realization — and gives one next step with the reasoning shown. The decision stays with the person.
Rebuild the week before adding work. Here is what it costs in the other spheres.
The more a system can do for a person, the less it should control that person.
Human life should remain the center of technological progress. We don't want people to spend their lives adapting to technology — we want technology to continuously adapt to the life a person actually wants to live.
One life. One context. One route.
The world is organised around categories. A person is not. Life GPS sees the whole person at once — every sphere, the time they actually have, relationships, environment and what is happening right now — and turns it into one understandable next step.
One step, with the reasoning shown. The person decides.
The system traces relevant constraints and candidate causes. It does not claim independently proven causality.
AI should not replace people. It should know when another person is more valuable.
The more capable AI becomes, the easier it is to replace a person with it. We use AI for the opposite purpose — and nobody here is ever rated.
No marketplace · no feed · no followers · no public ratings · no trust score attached to a person
A person is not a profile.
Captured once, a person is wrong within months. Life GPS keeps a living model that changes with actions, time and context. Behaviour weighs more than words. A real action outweighs the model.
Actions are not a sixth view. They are the signal that updates the others.
Technology learned how to capture human attention. We are building the opposite layer.
Life GPS protects attention, time, choice, mental state, relationships, a person from their own overload, from the pressure of their environment, from consumption that doesn't serve them — and from the pressure of this system itself.
Eighteen constraints, written before the code, published so the product can be held against them.
All of the above holds in the current beta — and each one costs us something a normal product would keep.
Eighteen constraints, written before the code.
Not features — limits. What the system may not do, even when doing it would improve every metric a normal product optimises.
No direct path from data to advice · protection outranks any goal · silence is a valid answer · identity does not become a label without the person confirming it · no conclusion is permanent, and where behaviour contradicts the model, behaviour wins · every step must explain itself · people are never scored.
The world is accelerating in every direction at once.
None of this proves anything about Life GPS. It describes the environment a system for a whole human life has to be built in.
Pacing the frontier becomes an explicit industry debate
Anthropic's Dario Amodei called for slowing the rate of frontier capability gains so safeguards can catch up — pacing, not stopping. Industry leaders remain divided over coordinated pacing.
Does not prove any shared slowdown was adopted, and nothing about whether Life GPS is the answer.
Pro-human AI becomes an organised movement
More than 50 organisations launched the Pro-Human AI Coalition, with a workstream devoted to a pro-human AI stack. The movement formulated the principles; our claim is narrower — we build the consumer layer.
Does not imply endorsement or membership. We are not stating either.
Consumer agents move from answering to acting
Meta launched Muse — goals into plans, work across apps, background execution, purchases with approval. Execution is becoming widely available. Direction is not.
Does not prove execution is already a commodity, only that it is becoming more available.
Attention-oriented design becomes a material legal risk
State attorneys general announced a multistate settlement with Meta of up to $17.1 billion over claims of harm to young users, with binding product-design requirements. California describes it as proposed and subject to court approval.
A settlement, not a fine. Not an admission, and not evidence that our design is safer.
What this changes
For a person
Less fragmentation. One path across a whole life, instead of nine products pulling in different directions.
For relationships
AI becomes a bridge to people rather than a substitute for them.
For the internet
The feed becomes a route. Scrolling becomes filtering. Content becomes action inside a life.
For business
Relevance replaces the fight for attention. We are not against business — we want a better moment between a business and a person.
For technology
More capability without removing human agency.
For education
The environment can assemble around the learner instead of forcing the learner into a fixed path.
Written as consequences of a design — not as achievements. What this changes
Every claim on this site carries a status. Where something is not ready, we say so on the line. What it does for a person · Figures and verification for investors
Publicity isn't my thing — I'm calmer behind the scenes. But someone has to tell you about the product.
Five years building this. Twenty-eight countries, four of them lived in. Dozens of ventures, seven full business cycles. Two funds offered $40M at pre-seed; the offer was declined to keep control. Two daughters, 18 and 14 — part of why this exists at all.
iOS, open beta, about fifteen minutes.
No account and no password — the first claim on this site you can check, in the first fifteen seconds.