Earlier quoted context omitted.
> There are types of hashes that aren’t for finding exact matches. My thesis in school way back when was building something for image near-duplicate detection using a sort of hash made out of the image features detected. The output of a neural net run on an image is also a sort of hash in a (maybe) semantically meaningful vector space. And then you can look for similar or nearby vectors. This isn’t just useful for fi…
Boy, there's a number of misconceptions here. I was considering not responding, but you're so wrong here that I just can't walk away and let you continue to spread this kind of blatantly wrong handwaving. > If for every one billion people(hypothetically of course) that are caught with child abuse images, one innocent person is asked to unlock their phone to prove that it's a false positive to a court, I think that's…
I agree, but it’s just as had waving when I say maybe it’s low chance that there are false positives, as it is when you’re freaking out at it being a hypothetically high number of false positives.
> That's not the problem here. If people want to continue to store photos of child exploitation on their iPhone, they can do it with the flip of a switch. Once you disable iCloud, Apple stops client-side scanning altogether.
If it’s so much under the users control why do you care or feel like this is an invasion of privacy then? I’ll say I feel like it’s still useful because criminals make dumb mistakes that incriminate themselves all the time — and this is a fine balance between being overly invasive and catching some people that store abuse images.
> Oh, like they already[0] do[1]? iMessage has been wiretapped since 2013, it's an open secret at this point that Apple is deep in bed with domestic intelligence agencies. Denying this is just refusing to acknowledge a truth that has reared it's ugly head again and again and again.
Yes, everyone knows this stuff, but just because it’s happening does not mean it can be used to incriminate someone in court unless there is justified prior suspicion which a judge issues a warrant for. That’s where the transparency is, at the court level.
> Countries like China…
This is a silly example, it makes me sad to say, but people in China are going to have their personal freedoms violated in very harsh ways regardless of Apple operating there of not. Unfortunately that’s the reality of living under totalitarianism — I don’t know why you’d even bring this up when I am clearly stating that what really protects people is living in a free democratic country with a well functioning court system.
> If they actually cared about those things, they'd divert some of the national budget to addressing those issues. Instead, they take that money and invest it in surveillance. Billions of dollars get poured into black budgets every year, with margins that constantly expand. The NSA regularly backdoors encryption standards, the FBI regularly designs exploits and incorporates them into US infrastructure.
I don’t wholly disagree, but I don’t see the connection with the original point. Also, you’re talking about governments as a monolith, which is a mistake. Any government is made of may departments and people, when you say ‘if they actually cared’, who’s the ‘they’? I think it’s really silly to say no-one in any part of the FBI cares about stoping child abuse for example. Sure I get that the CIA does not care, but that’s just not their job either.
> If any of our surveillance programs want to backdoor an iPhone or introduce vulnerabilities into iPhones, Apple has no recourse.
Maybe, maybe not. What is your hypothetical scenario for this? Can you explain how you think they’d do this where Apple has no recourse? Also who is the ‘they’? When you just say ‘they’ in a handwaving general way you sound conspiratorial. Are you saying they’d do it and apple could not detect it? Or that they’d coerce apple and apple would have no legal way to stop them?
> A much more appealing effort is to keep tabs on everything you do, everywhere you go and everyone you talk to. That way, if you become a legitimate threat to the state they can zero you before you cause any real damage. Make no mistake, every government is only out to save their own skin. You may disagree with it, or find it silly, but that's realpolitik for you.
This is where we depart the most. I’m European, In my opinion, America, with its two party non proportional system is a very poor example of democracy that is quite susceptible it corruption because of it’s concentration of powers. Just because you guys don’t live in a well functioning democracy does not mean that’s the case for the whole world — you’re projecting a bit here. I guess we agree in our mistrust of the US. Hopefully you guys don’t slid more on that totalitarian scale even more I guess.
> There is no way that we're living in a world where our government lacks the competence or resources to effectively monitor us at every corner.
I guess you don’t know about off grid log cabins that have no cell signal ;-)
> Apple is given a black-box list of irreversible image signatures that the government themselves consider harmful, and there is literally zero transparency that goes into this process. We have quite literally no idea how any of this works,
I think this is the point you’re wrong here, if there is a false positive I believe even in the US that you’ll be able to simply exhibit the source image that caused the false positive in court, if it even gets that far, which I think probably not, and prove innocence. The hash match is not the proof that you have child abuse images — it just starts a process of review. I think in practice it’ll start a process where a manual review will happen where someone will look at your source image and say, oh yeah, that’s clearly a false positive, and then it’s done. I’m sure we’ll know more where the first people get cause with this mechanism because it will still have to be prosecuted in a public court and at that time we won’t have to speculate so much.