You don't need to build a prototype to know it will be abused. That's obvious. By the way, once they start hashing and scanning every piece of content that goes through, do you think they won't keep that data? They'll have a record not only of everything being shared in real time, but of everything anyone has ever shared on their platform.
And will be able to go back and retroactively be able to see what all a person was looking at. Say they get ahold of a pdf that's anti-govrenment. They just hash that and see who all was reading it months before, before their system even knew about it. How much data does it take for the system to send a hash of every file you open? A few KB a day.
Exactly. It can even go deeper. There are algorithms for getting a lossy fingerprint of an image such that you can find similar images in a resolution independent way by comparing the fingerprints--it's what services like tineye do. So it doesn't even have to be the exact image.
Or apply it to racist content when they decide that racism is obscenity not covered under the first amendment. It's easier to classify than CSAM anyway.
Also, with Pegasus they can put any content they want onto anybody's phone. Do you really want your phone pinging the FBI when you don't have 100% control over every file that's on it?
As much as people downloading that content sucks I still firmly believe in the concept that software run by a user should 100% pursue the interests of that user. Software on your phone pinging the FBI is never in your interest and you shouldn't run software by companies that write software that isn't in your interest.
I think the same metric should be applied to it that should have been applied to covid. People dying sucks. It infinitely sucks. But if the tools used to mitigate that suck give unprecedented powers to organizations and governments that have proven that they can't be trusted, then it's not on the table. We need to stop experimenting with giving them powers whose implications can't be known. It's far more dangerous than dealing with the suck that is the unfortunate but luckily rare human experience.
Of course they are going to use an issue that stirs emotions to gain power. They will do it every time. If we let that pattern work they will get any power they want any time they want, or at least the next step in full control. They always have a button they can push to get what they want.
As much as people downloading that content sucks I still firmly believe in the concept that software run by a user should 100% pursue the interests of that user.
Government is the biggest threat to mankind. It's a large concentration of power that can be controlled by a small number of people, for their own twisted purposes. Any group of people that's a threat pales in comparison to those same kind of people plus government power. So, to give such a dangerous power to government just to stop a group of people sending photos is not worth it.
As much as people downloading that content sucks I still firmly believe in the concept that software run by a user should 100% pursue the interests of that user.
I think almost every developer would agree with this, and almost zero for-profit companies implement it. Almost all paid software treats the user as, most benignly, a mark for future software sales. At a middling level of evil, the user's device is the product -- the way most malware operates. At worst, the user is the product.
(post is archived)