A number
with context.
When a number is supported, see the population, time frame and calculation behind it. An estimate for a group isn’t a prediction for a person.
A new way to put risk in perspective
Ask what’s on your mind. Add the details that matter. Get sourced evidence and useful next steps, with room for uncertainty.
An early iPhone beta. Built around your question.
“How does my trip
change the picture?”
A little context. A big Ah!
01 FROM QUESTION TO CONTEXT
One question. A few relevant follow-ups. An answer you can look inside.
Start with the question you already have.
Add the details that belong to this question.
The estimate, its limits and useful next steps.
Each time you board a plane counts once. A nonstop round trip is two boardings. Connections add more.
Try changing the count. The source stays the same.
Passenger death from aviation incidents
About 1 in40 million
across 2 boardings
Commercial airlines in the study’s first-tier countries, including the US. Data: 2018–2022.
A population estimate, not a prediction for you or a particular flight. Assumes independent boardings with the same historical rate.
THE SOURCE
The study reports approximately one passenger death per 80 million boardings for its first-tier-country group in 2018–2022.
Read MIT’s account of the study (opens in a new tab)THE CALCULATION
1 − (1 − 1 / 80,000,000)2
≈ 0.0000025% across 2 boardings.
This is an approximate historical average. The source gives no uncertainty interval for this rate. It does not account for the details of a particular flight or infection during travel.
An interactive example from the Ah! library. No live research request or personal data is sent.
02 THIS IS WHERE IT GETS PERSONAL
Who it’s about. Your routines. A trip. A known condition. Go deeper when it helps, and skip what you don’t know.
Some details change an estimate.
Others make the next step more useful.
Ah! explains the difference.
“What about the car
I actually drive?”
Explore the kind of details you can add.
WHAT THE DETAIL ADDS
Match recall information to the make, model and year; use the official lookup for a specific car.
NHTSA: Check for Recalls (opens in a new tab)Context and guidance. No numerical adjustment.
This example doesn’t ask for or save your vehicle details.
03 AN ANSWER YOU CAN LOOK INSIDE
Open the sources. See the reasoning.
Know where the evidence ends.
When a number is supported, see the population, time frame and calculation behind it. An estimate for a group isn’t a prediction for a person.
Go from information to something worth doing. Your “Worth your effort” shortlist brings practical actions together, with the evidence behind them.
A detail to check. A habit to consider. A question to take to a professional.
Some questions don’t have a reliable number. Ah! explains the gap. That’s part of a useful answer, too.
An early beta can make mistakes. Ah! is not for diagnosis, treatment or emergencies.
THE NEXT CHAPTER
We’re testing an early iPhone beta.
Be part of what comes next.
STAY IN THE LOOP
Leave your email for news and invitations. Joining the interest list doesn’t give you immediate access.
We’ll use your email for beta updates and invitations. No immediate access or automatic invitation.
Ah! stores your email, contact consent and signup time with Netlify for beta updates and invitations. We do not sell this list. We retain it while the beta interest program operates, or until you ask us to remove your entry. Email hello@getah.co for contact or removal requests. Netlify also processes technical request information to operate and protect the site. This signup is separate from the app’s data-sharing choices and does not collect health answers.
A LITTLE MORE CONTEXT
No. A relevant detail can change the comparison group, add a practical next step or leave the estimate unchanged. A numerical adjustment needs evidence that applies to the question.
No. Profile details are optional. Start with one question, answer relevant follow-ups and skip anything you don’t know. You can save useful context for yourself and your household.
Your profile and saved answers are stored on your phone. You choose which profile categories to share for live research. Questions, follow-up answers and short titles of recent questions are sent to our research service, which uses Anthropic’s AI. Structured location sharing needs consent for that question; information you type can also contain personal details.
The test build has a separate usage disclosure. If accepted, uploaded events use a random install identifier and can include questions and follow-up answers. Structured profile answers and location fields are excluded; personal details in text you type can remain. Offline mode or turning off “Send my questions” stops future usage uploads. The beta-interest form on this page is separate; it saves only your email and consent.
No. Ah! is an early beta for understanding risk evidence and context. It can make mistakes and is not for diagnosis, treatment or emergencies. For symptoms or health decisions, consult an appropriate healthcare professional.
Ah! is being tested through an early iPhone beta. Wider availability hasn’t been announced. You can register interest above; invitations are a separate step.