4 ms·
I'm happy your iPhone shows you the notification in all situations, but that's not true for older iOS versions or any version of Android -- aside from maybe v11
by bryans 5y ago
I'm happy your iPhone shows you the notification in all situations, but that's not true for older iOS versions or any version of Android -- aside from maybe v11. So to outright claim that it's not possible just because you, personally, don't experience it, is plain silly. Especially because I, personally, can open any number of apps on my Pixel 3 right now and have them record in the background without a microphone icon ever being visible.
And as far as real world audio being a challenge, I think you might be slightly behind the times on the power of perceptual hashing. It's disturbingly accurate with very little context or effort.
- eightysixfour 5y agoIt has been that way for a long time on iOS, I just checked and even a phone call requires the icon. Yet another reason not to trust Android devices. I am broadly aware of perceptual hashing and its capabilities. Where I feel like you are underestimating the amount of work involved is in performing perceptual hashing without anyone noticing and then matching without anyone noticing. If you take the OPs example of “warheads,” where does the matching happen? Are you suggesting the social apps are storing the hashes of everything people said and then sending them to servers to match against a list of perceptual hashes? Because then you’re just hashing the audio and sending it, which would be obvious. The other option would require a list of perceptual hashed to be stored on device and a local match performed, this would, again, be obvious because now the social apps would have to store a giant set of hashes locally. For example, Roku and TVs use the former - they sample your content and send hashes off to a server for matching, this is identifiable in their network data and would, similarly, be identifiable in mobile app network data. Importantly, they’re sampling controlled data and can simply send an image hash every X-seconds, you can’t do that with the “warheads” example, you need to capture and send it all to catch the key words.
- bryans 5y agoI hear what you're saying, but I think you're making a few assumptions that aren't necessarily always true or guaranteed. As to the obviousness of sending hashes to a server, I can think of a dozen ways to make that look like routine API data that's getting synced during a routine polling process. The app is a black box until data hits a networking layer, and it can be made to look like anything they want before it gets there. Would they need to be acting deviously to accomplish this? Of course. Is it in these companies' nature and history to act deviously? You bet it is! As for the complexity of matching without anyone noticing, I think you may be overestimating the amount of data needed to accomplish this task. They wouldn't need to record full conversations or phrases, or even have any context at all. As you and others have noted, they already have so much contextual information about us, that all it would take is successfully capturing a few words once in a while to really fine-tune they're targeting. Some very simple audio processing could easily chop up "probable" words based on any number of factors. It could throw out any recording that didn't meet some minimal requirements or seemed noisy. And at that point, you're just sending some strings to the server occasionally, no longer than the typical cookie. And again, they already brazenly do this with Messenger and WhatsApp and everything else in the world that isn't Signal. So, that it's both technically possible and they have already invested heavily into such systems, makes this all a real possibility. At the very least, it's something that should continue to be investigated and actively monitored for, because even if the biggest of the companies aren't doing it, that doesn't mean nobody is doing it -- or doing it on their behalf.