Why now
In short
The world is accelerating in every direction at once — agents that act for people, consumer AI that remembers and buys, courts examining how products are designed, an organised movement forming around keeping AI under human control.
None of this proves anything about Life GPS. It describes the environment a system for a whole human life now has to be built in — and we are building and updating a map of that world, so that a person's route does not go stale while they are walking it.
Every item below was re-checked against primary sources on 16.09.2026. Where only secondary reporting exists, it says so. Our own earlier research notes were not treated as evidence. This page dates quickly — the published date and the last-updated date are shown separately for that reason.
What we are building toward
Life GPS holds two maps: the map of a person, and the map of the world around them. The second one is the one nobody keeps for themselves — services change, tools change, what is possible this year was impossible last year, and a route built on last year's world quietly stops being the best route.
Status. Vision / in progress. The environment, partners and favourites parts of the world map are in the beta. We are building a continuously updating map of the world — we do not yet claim it keeps itself current.
September 12, 2026 — pacing the frontier becomes an explicit industry debate
Anthropic CEO Dario Amodei called for deliberately pacing the growth of frontier AI capabilities so that safety, evaluation and oversight can keep up. He framed pacing as slowing capability growth, not stopping technical progress. Sam Altman and Elon Musk publicly supported the direction. Other industry leaders continue to argue that safety and speed need not be a binary choice.
The question this raises. If capability growth is the thing being paced, what is supposed to be catching up?
Our position. We don't think progress should stop. We think the human side has to catch up — and that the catching up is a thing someone has to build, not a thing that happens when everyone agrees to wait.
Industry leaders remain divided over coordinated pacing. On 16 September, Meta's Mark Zuckerberg argued that AI labs already have sufficient incentive to build safely and set that against coordinated slowdown. Others hold the opposite. The disagreement is live, and this page will age.
What this does not prove. That the industry has adopted any shared slowdown. And nothing at all about whether Life GPS is the answer.
Sources: Dario Amodei, "We must pace the frontier", 12.09.2026 · Reuters, 12.09.2026 · OpenAI's own published pauses tied to safeguards.
September 15, 2026 — pro-human AI becomes an organised movement
More than 50 organisations across technology, policy, research, investment, philanthropy and advocacy launched the Pro-Human AI Coalition, an independent nonprofit focused on keeping AI under human control and broadening who benefits from it. Co-founded by Project Liberty, the Future of Life Institute and Ergo Impact. One of its five workstreams is explicitly called a "Pro-Human AI Stack".
The question this raises. The principles are now written down by serious institutions. Who builds the consumer layer where a person actually lives?
Our position. The movement formulated the principles. Our claim is narrower: Life GPS is an attempt to turn them into a working layer around the whole life of one person.
What this does not prove. The Coalition does not endorse Life GPS, and we are not stating membership. "50+" is the number the Coalition publishes; we do not invent a precise count.
Sources: prohumanai.co · launch release, 15.09.2026.
September 8, 2026 — consumer agents move from answering to acting
Meta launched Muse, a consumer personal AI agent that turns long-term goals into plans, works across connected apps, continues tasks in the background, fills forms, negotiates and makes purchases with user approval. Meta says it runs in a dedicated secure VM with an approval layer and an audit trail. Free tier, with paid plans reported at $20 and $100 per month.
The question this raises. When execution becomes widely available and cheap, what is left that is scarce?
Our position. Execution is becoming widely available. Direction is not. Knowing how to do something, and knowing whether it is the right thing for this life this month, are different problems — and the second one requires holding the whole life, not a task.
What this does not prove. Not that execution is already a commodity — only that it is becoming more available. Not that Muse understands a person's life the way this system tries to. And not that we are differentiated — that has to come from our mechanics, not from another company's launch.
Sources: Meta launch page, 08.09.2026 · prices reported by Reuters, TechCrunch and Axios — not published on Meta's launch page.
August 26, 2026 — attention-oriented design becomes a material legal risk
U.S. state attorneys general announced a multistate settlement with Meta of up to $17.1 billion over claims concerning harms to children and teens, together with binding product-design and child-safety requirements: time limits, nighttime blocks, notification restrictions, parental tools, age assurance, an independent auditor, and a non-personalised feed option for minors. California describes the settlement as proposed and subject to court approval.
The question this raises. If design choices carry this scale of liability, what does it mean to design deliberately in the other direction?
Our position. The mechanics being restricted there are the ones our constraints prohibit by name — written down before the code, not after a settlement.
What this does not prove. This was a settlement of litigation, not a fine, and not a court finding that every engagement mechanism is unlawful. It is not an admission by Meta. And it is not evidence that our design is safer or more effective.
Sources: California, New Jersey and Colorado Attorney General releases, 26.08.2026.
March 2026 — the fight moves from content to product design
A Los Angeles jury found Meta and Google negligent over platform design in a case focused on alleged harms to a young user, awarding $6 million; post-trial motions were denied in June. The disputed design features included infinite scroll, autoplay, algorithmic recommendation and notifications.
The question this raises. The legal fight moved from what content was shown to how the product itself was designed. Every product holding a person's attention now sits inside that question.
Our position. A system that holds a person's whole life has more power to steer than any feed — which is why the limits on it have to be explicit, public, and written first.
What this does not prove. No court has ruled infinite scroll a defective product; we looked for that exact finding and could not verify it, so we do not write it. The verdict is not final after appeals, and it proves nothing about Life GPS.
Sources: Reuters, 25.03.2026 and 10.06.2026 · Los Angeles Superior Court, JCCP5255.
February 9, 2026 — conversation becomes a commercial surface
OpenAI began testing ads in ChatGPT in the U.S. for logged-in adults on Free and Go. On August 31 it said ChatGPT Ads had reached a $1 billion annualised revenue run rate in under 200 days. OpenAI states that advertisers do not receive conversations or personal details, that ads are separate from answers, and that paid plans remain ad-free.
The question this raises. Once intent lives inside a conversation, the question is not whether it gets used, but who is allowed to buy influence over it.
Our position. Business should enter a person's route when it is relevant to a real need — and should never be able to buy influence over the route itself. Quantity belongs in creation; quality belongs in consumption.
What this does not prove. This is a product and business distinction, not an accusation. An annualised run rate is not annual revenue, and it is not "a billion dollars earned in 200 days".
Sources: OpenAI, 09.02.2026 and 31.08.2026 · OpenAI Help Center.
Also on the map
Robotics moving into ordinary settings · tissue and organ engineering changing what a human body's timeline looks like · personal AI arriving in every consumer device · agentic commerce being built into the payment layer by Visa, Mastercard and Stripe · AI companions and the regulation now following them · education and work changing faster than the institutions that certify them.
These are directions we track and account for, stated without claiming precision on any single item. Each becomes a dated entry above when it is verified to that standard.
How to read this page
Six steps, every time: the event with a source and a date · what the source actually says, without compression that changes its meaning · why it matters to our question · our position · what Life GPS does and what status that has · and what the event does not prove about us.
An external event is context. It is never evidence that our product works.
Life GPS · lifegpsos.com · published 16.09.2026 · sources re-verified 16.09.2026
Published 2026-09-16 · Updated 2026-09-16 · Every claim on this page carries a status