I suspect they were related and the earlier CSAM tech was announced because of plans to enable e2ee like they announced today (otherwise the tech didn't make sense since they could just scan the unencrypted photos on their servers).
The apple one was only hash matches with known ncmec images over a certain threshold. Even then there was a human in the loop iirc. So it wouldn’t have been an issue.
Google was doing novel image detection though and mistakes there are more possible and more serious. As evidenced by the recent NYT article about that guy.
Interesting - their approach to hash checking prior to iCloud upload (only when upload was enabled) seemed intended to allow them to support e2ee which they announced today (without the tradeoff of enabling CSAM).
I guess they decided it wasn’t worth it and it’d be good to ship e2ee anyway despite that meaning it may protect CSAM. I had figured they’d enable e2ee first and then reannounce the hash check if you enable it but I guess not.
I can’t see the rest of the article due to the paywall, but I wonder what changed. Maybe the risks with imperfect hash matches or the recent news of Google reporting that guy to police made it not worth it?
It’s in someways to Apple’s benefit to do this, they can’t see the data and are no longer responsible for policing its contents. They can’t accidentally leak it or get social engineered to restore access to a fraudster (how some celebrities got their images leaked).
Whatever it is, it’s nice when the incentives are aligned around user privacy. It’s cool (and a little surprising) to see real consumer device facing e2ee for cloud storage, even if behind an advanced flag.
You can compare the pop-up language Apple shows for an app like FB and the one they show for their own personalized ads to see what I'm talking about. They've also run misleading ads and have made comments that confuse people about what's actually going on.
I'm no apologist for ads, but Ben Thompson is right to point out that this hurts small companies that rely on these targeted ads in order to exist a lot more than it hurts large players like FB.
For example - a grocery store doesn't want to manage 'first party' user data to track what you purchase (and you probably don't want them to), they're bad at that and more likely to do it poorly. They'd rather rely on an ad company they can use instead. This applies to most small businesses that rely on targeted advertising to get their business in front of users that would want it. In Apple's model Amazon doesn't need to say they track you because all purchase data happens on Amazon, but FB does because others use FB to target third party ads. The data doesn't leave FB though so a reasonable person could argue why is this worse?
My personal opinion is that we'd be better off in an equilibrium where these ad driven models are not viable because the models that would replace them would be better on net with incentives more aligned between user and product.
There is a problem here with how user data is handled in some cases, but Apple is also being at best misleading about the issue in a way that benefits themselves and reasonable people could think they're doing the wrong thing.
It sounds like you know more about this stuff than I do - I’ve spent hours searching online but it’s hard to find good info.
Why are mics like the SM7b better than mics like the MV7 or elgato one that just work over usb? I don’t understand how they actually work or why all the extra equipment (audio interface, gain booster, cables) is worth it - what’s the tradeoff being made when you use USB?
+1 the reasoning is the entire point and all of these examples are good.
Just telling people to make changes without telling why doesn't teach them anything - especially when sometimes what's being requested isn't actually better (or is at least subjective).
That small minority seem to have disproportionate power (at least in the case of Lia Thomas).
I’m not sure how small it is and I don’t want to enable it.
There are people forcing/pressuring participation in the public display of pronouns/say your pronouns thing (you may just live somewhere insulated from it).
I’ve had it happen in Silicon Valley interviews and I know some groups at Stanford require it.
I don’t do it because I think most of the time it’s more about signaling political affiliation than actually solving a problem.
It also tends to then extend to things like pretending there are no sex differences and that everything is socially constructed (which I think is nonsense). I used to think this was a strawman, but the Lia Thomas story shows it isn’t imo: https://overcast.fm/+vpWbPtN0o
This is obviously a third rail internet topic and people can do what they like, but I won’t put pronouns in my bio and I won’t submit to pressure to do so in order to pass some sort of political morality test.
https://www.lesswrong.com/posts/CqyJzDZWvGhhFJ7dY/belief-in-...