To each their own.

For twenty years the bargain has been the same. Software gets better when it knows you, knowing you requires your data, and handing over your data meant handing over control of it. So your mail, your calendar, your documents and your photographs went to companies who wrote the rules for what happened next — and kept the right to rewrite them.

AI does not extend that bargain so much as complete it. To be genuinely useful, an assistant needs the whole picture: what you wrote, who you answer quickly, what you keep putting off, what you said in a meeting last spring. Handing that over produces the most complete portrait of a person that has ever existed outside their own head — and under the usual arrangement, it becomes raw material for someone else’s product.

We think the premise is wrong, not the technology. Holding someone’s data does not have to mean owning it. A service can keep your data encrypted, use it only to serve you, refuse to share it, and hand all of it back the moment you ask. That is a choice a company makes, and it is the one we have made.

So we are building it that way. The models are open-source ones anyone can inspect, moving onto our own hardware as the waitlist grows. Your data is encrypted and never sold. Nothing is used for training, or anything else, unless you say yes. And every piece of it can be downloaded to your own machine, in open formats, whenever you like. What we sell is the software — not you.

AI is moving faster than most people can take it in. The largest labs are racing each other, and the rest of us are asked to keep up — with our jobs, our savings and the planet absorbing the cost of the speed.

We cannot call off that race. But much of the money behind it comes from ordinary subscriptions, and every subscription is a choice. Moving yours here moves it away from the biggest labs. If enough people do that, it slows them down. If it does not, it still sends a message they can count: people want this to go slower, and more carefully.

We want the good version of this future — the one where AI makes life genuinely better for nearly everyone. We think the way there is gradual: keeping the pain of lost jobs and economic strain as small as it can possibly be, rather than treating it as the price of progress, and protecting the environment along the way rather than spending it.

That shapes how we build. We choose efficient models sized to the job, not the largest one available. We grow at a pace people can actually adopt. And we would rather be useful than inevitable.