For the past ~2 years, I’ve been working on a neuroscience startup (called Matter). We have an app that—by its nature—needs to operate on private user data. Some of that data is so private that we never actually want to handle it ourselves. So we don’t.
We can’t leak data we don’t have. We can’t mishandle private info that never leaves the control of the user.
We’re taking a mostly-unique approach to user data at Matter, and I’ve finally written a little bit about it:
Lots over there getting hung up on worst case scenarios. “What if someone buys you and they’re evil and they change the code and I still update?!”
Our goal here is to be better, not to try to be naïvely bulletproof-perfect. If your threat model really is “it doesn’t matter what their intent is, if my data gets out, I’m finished”, you should probably not be using a computer at all. Or at least not running software that anyone else wrote… including the software that’s in your CPU. Good luck.
@sean I dunno man, people give their data to FB, Twitter, Instagram, LinkedIn etc. They watch scrapers and language model builders suck up the rest.
Someone else comes along, not sucking up their data, trying to be good… and that, That’s what they choose to be annoyed with.
Sean Coates
Paul Reinheimer
nicksloan