Hacker News new | ask | show | jobs
by aprilzero 3846 days ago
Thanks!

First of all, we're not selling any data, nor do we ever plan to. We do make money by offering a pro subscription--this new design is one of the features, along with things like custom domains and access to our Healthkit app. We've already gotten a few dozen upgrades today, which is really exciting and lets me confidently say things like "we will never sell your data".

We believe that sharing of this data is an important part of the experience for most people. Of course you don't have to, but for many people that is the first thing they want to do. Perhaps only with your family or a few close friends, and not the whole world, but a single-player experience that only exists on your device would be missing many of the experiences that we think are core to Gyroscope.

A small percent of users have made their profiles public, and many of those have set up a custom domain name and are very excited about sharing it and having a personal website that reflects who they are. Having the data in the cloud makes that use-case possible, though of course doesn't require it.

Our new mobile apps are also taking a slightly different approach, where all of the source data is private but allows for curation and customization, and you can export individual images with aspects that you want people to be able to see. I think that will be the right balance of privacy and control for most people, while still allowing sharing of things they want to highlight.

Making my account completely public has been an interesting experiment. When I first did it last year, I wasn't sure what to expect. Overall it has been very positive experience, and my current opinion is that more transparency is generally good. The location data is probably the most sensitive out of all the things we track. For other people, weight or age have been concerns, and we've added settings to fully hide those.

There are things I consider very private and wouldn't want anyone else to see — like my messages, emails, finances, etc. — but I am personally happy to release the data on the site, like what my heart rate is or how much I slept yesterday. That balance will vary for everyone, which is why I have been experimenting mostly on myself before recommending things to others.

The location data is probably the most sensitive, and we've been playing around with other themes that show different combinations of data and keep more of that private.

We might add other login options in the future, but currently we've found Facebook to be a great experience because we can boot up your basic account info (timezone, name, cover photo, age, etc.) without needing all that to be manually entered. It would be a terrible experience if the first 5 minutes on Gyroscope were just spent typing in basic facts about yourself. One of our big challenges with Gyroscope is how to passively power all this stuff without creating any more work for the user. Our goal has been to go through the whole experience without ever needing to use your keyboard to type anything, maybe with the exception of setting your username.

3 comments

Well, in my case, I deleted my facebook a few years ago, so you actually did lose me, not just as part of some public stand.

I think it's interesting you're publishing your data, and I'm glad you're happy with what you're publishing -- it totally makes sense. User-defined privacy is an awesome thing.

It remains my proposal that you consider actually cutting off your own access to customer data unless they decide to publish it, essentially extending that ability you have while in control of the system to your customers. I think you can see from some of the comments in this chain that people aren't so trusting as you feel you deserve.

And, of course, if someday you or the company needs the revenue, then it might be nice to have already, when not under duress, made a not easily-revocable decision about the ways you'll choose to monetize user data. Or, obviously, it might hamstring your plans.

Same here, deleted my Facebook account a year or two ago, so any time a "cool" service pops up that requires Facebook, it's a bummer to not be able to try out the service, but at the same time it feels like it validates my stance that Facebook has become one of the biggest dangers to the original dream of what the Internet could become.
+1 - Not a facebook user any more, but someone very interested in personal health. The site looks amazing but I wouldn't reactivate my facebook just to use it.
> We've already gotten a few dozen upgrades today, which is really exciting and lets me confidently say things like "we will never sell your data".

This wording makes it seem like that data sale starts as soon as your revenue stops making you happy

my point was we're working now to try to figure out the right mechanism — whether it is pro subscriptions or selling hardware, etc. — so we don't end up in a position where we'll need to sell data later.
> end up in a position where we'll need to sell data later

The point is that there's nothing structurally preventing you from selling data later. Unfortunately that's not a great reassurance, as much as one may want to believe in the project.

To be clear: I think this is an interesting undertaking and you guys are producing good stuff. This structural problem is concerning though, and deserves some serious thought.

It's possible to set up encryption model where it will be particularly hard to sell data later -- this is one of the things we've spent a decent amount of time building in the company I work for, for this reason. No, we're not selling data, and we're actively working to block that pathway in future even if the leadership has completely changed.

E.g., data at rest for private accounts is encrypted with per-account keys based at some level on the user's login, and internally can't be decrypted until the user is signs in next.

Your software interests me and I would like to try it. However, I am not a Facebook user and don't plan on being one. For me, it's a pity that it is tied to Facebook.