"help improve apple intelligence" - sounds suspect

[stevemiller]

this launch window saying that creating images helps improve apple intelligence makes it awfully nebulous as to whether they are training on your personal data like images.

can anyone clarify what 'improving' apple intelligence means in this context? and even if its not using any personal data, shouldn't we at least push back on companies using vague wording like this that very much leaves the door open to them using your photo library to train models if you access it from within image playground?

like they can say 'no we aren't' right now... but then change their mind later, and the user facing copy could stay identical and they could claim 'well we notified you here and you agreed to it'.

at the end of the day anything i'd put in image playground would be anything different than what i'd likely be posting online to be scraped anyway. this is more the principle that this stuff is getting more and more invasive and it feels like they use 'hey just try this fun thing we made' to try to make people more and more comfortable with it.

 

[kitKAC]

There's a Report a Concern function in the app, you can choose whether you share the image that was created or any diagnostic data used to generate the image.

 

[stevemiller]

There's a Report a Concern function in the app, you can choose whether you share the image that was created or any diagnostic data used to generate the image.

that is definitely the innocent reading of it. but i just find it unsettling that the very first thing you're presented with when opening the app is that you're helping improve apple intelligence by creating images (which often involves a base image of your own).

most apps don't introduce themselves as a call to action for bug reports. i get that this is a 'beta', but i also feel like with all the concern around training ai models on user data (and companies being EXTREMELY shifty about admitting it), this wording can easily also be read as some form of broad consent.