Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Problem with Perceptual Hashes

rentafounder.com

Post reply on HN
Print view View original

331–340 of 440 posts

Re: The Problem with Perceptual Hashes

#331

Post by jeromegv »

l33t2328 wrote: post #297
gpm wrote: post #156

Earlier quoted context omitted.

With you up to here, but this is jumping the shark > I bet you, after the next election in the US your device will be reporting you for spreading far right or deep state lies, depending on who wins. The US is becoming less stable, sure [1], but there is still a very strong culture of free speech, particularly political speech. I put the odds that your device will be reporting on that within 4 years as approximately 0…

I don’t see how the US is becoming “less stable” in any meaningful sense. Can you elaborate?

An attack on the capitol on January 6. A former president that spent weeks trying to delegitimize the election, trying to get people fired when they were just following the process to ratify the election, etc.

Re: The Problem with Perceptual Hashes

#332

Post by bccdee »

sobriquet9 wrote: post #303
CuriousCosmic wrote: post #266

Earlier quoted context omitted.

FYI Illiberal values aka non-liberal values (clarifying because the I is hard to read) use the word liberal in the traditional sense. Liberal values are liberty/freedom, consent of the governed, and equality before the law. All other liberal values build off of these three as a base. This implies that Non-liberal (or illiberal) values are the opposition of liberal values through censorship, gun control, etc like you…

Karl Marx quote on the right to keep and bear arms only applies to the proletariat. If you are a programmer and own the means of production (your laptop), you are not proletariat. All socialist and communist countries have strict gun control.

Proletarians make a living off of wage labour. The Bourgeois make enough to live on off of investments in capital.

Owning a laptop is perhaps a very tiny investment in capital, arguably, but it certainly won't provide enough passive income to replace your job.

Re: The Problem with Perceptual Hashes

#333

Post by marcellus23 »

tjoff wrote: post #329
wingspar wrote: post #321

Earlier quoted context omitted.

The operative separator is “Next” Apple is announcing 3 new ‘features’. First one scans iMessage messages / photos on device / warns kids and partners. Second one is the CSAM photo hash compare in iCloud upload feature. Third one is the Siri search protection/warning feature.

Stand corrected on the first part. But surely iCloud upload feature is on the device. And if it was only in the cloud they wouldn't need to mention iOS or iPadOS at all.

But what’s the practical difference between scanning photos when they’re uploaded to iCloud on a server, or on device?

Re: The Problem with Perceptual Hashes

#334

Post by jeromegv »

visarga wrote: post #204
Klonoar wrote: post #134

Earlier quoted context omitted.

>Both are bad, and not the responsibility of a company selling phones Apple is not just a hardware company and there is no obligation for them to host offending contents on their servers - just as Dropbox, Google, and so on would maintain with theirs. >As for your suggestions to just "stop using iCloud Photos", how about we get to enjoy the features we bought our devices for, without stuff we didn't ask for and don't…

With many more images, many more false positives. One has as a consequence a message or account being deleted, the other - being reported to the police. Very different!

They were reporting to the authorities before as well with what was found on iCloud photos.

Re: The Problem with Perceptual Hashes

#335

Post by jeromegv »

dannyw wrote: post #176
Klonoar wrote: post #137

Earlier quoted context omitted.

This example changes with regards to emotional weight if you remove "a man" and leave it at just "a scanning machine". There is no human scanning your photos on an iPhone, so let's compares apples to apples here. If that scanning machine didn't reveal the contents of my mail, and then ensured that it wasn't able to be given out in-transit? Yeah, I'd potentially be fine with it - but I'll leave this answer as a hypoth…

There actually is a man involved: enough similarities and a human will review the photos. Every algorithm, especially perceptual hashing, will have false positives, and at Apple's scale, some people's private and intimate photos will be false positives and be exposed to a man looking at it.

I think the point of the OP is that it was already the case before when you were using iCloud photos. The scan was server side.

Re: The Problem with Perceptual Hashes

#336

Post by jeromegv »

siscia wrote: post #277

What I am missing from all this story, is what triggered Apple to put in place, or even think about, this system. It is clearly a no-trivial project, no other company is doing it, and it will be one of the rare case of a company doing something not for shareholders value but for "goodwill". I am really not understanding the reasoning behind this choice.

One theory is that they are getting ready for E2E encryption of iCloud photos. Apple will have zero access to your photos in the cloud. So the only way to get the authorities to accept this new scheme is that there is this backdoor where there is a check client-side for sexual predator photos. Once your photo pass that check locally, it gets encrypted, sent to the cloud, never to be decrypted by apple.

Not saying it will happen, but that's a decent theory as of why https://daringfireball.net/2021/08/apple_child_safety_initia...

Re: The Problem with Perceptual Hashes

#337

Post by tjoff »

marcellus23 wrote: post #333
tjoff wrote: post #329

Earlier quoted context omitted.

Stand corrected on the first part. But surely iCloud upload feature is on the device. And if it was only in the cloud they wouldn't need to mention iOS or iPadOS at all.

But what’s the practical difference between scanning photos when they’re uploaded to iCloud on a server, or on device?

A world of difference. Both in practical terms and principle.

To start, once you upload something to the cloud you do - or at least are expected to - realize that it is under full control of another entity.

Because of that you might not use iCloud or you might not upload everything to iCloud.

Re: The Problem with Perceptual Hashes

#338

Post by MontagFTB »

siscia wrote: post #277

What I am missing from all this story, is what triggered Apple to put in place, or even think about, this system. It is clearly a no-trivial project, no other company is doing it, and it will be one of the rare case of a company doing something not for shareholders value but for "goodwill". I am really not understanding the reasoning behind this choice.

Legally, I believe, they are responsible for distribution of CSAM that may wind up in their cloud, regardless of who put it there. Many cloud companies are under considerable legal pressure to find and report it.

Re: The Problem with Perceptual Hashes

#339

Post by josefx »

judge2020 wrote: post #274
avnigo wrote: post #271

> These cases will be manually reviewed. That is, according to Apple, an Apple employee will then look at your (flagged) pictures. I'm surprised this hasn't gotten enough traction outside of tech news media. Remember the mass celebrity "hacking" of iCloud accounts a few years ago? I wonder how those celebrities would feel knowing that some of their photos may be falsely flagged and shown to other people. And that we…

They wouldn't be falsely flagged. It doesn't detect naked photos, it detects photos matching real confirmed CSAM based on the NCMEC's database.

Hashes, no false matches, pick one.

Re: The Problem with Perceptual Hashes

#340

Post by wongarsu »

judge2020 wrote: post #274
avnigo wrote: post #271

> These cases will be manually reviewed. That is, according to Apple, an Apple employee will then look at your (flagged) pictures. I'm surprised this hasn't gotten enough traction outside of tech news media. Remember the mass celebrity "hacking" of iCloud accounts a few years ago? I wonder how those celebrities would feel knowing that some of their photos may be falsely flagged and shown to other people. And that we…

They wouldn't be falsely flagged. It doesn't detect naked photos, it detects photos matching real confirmed CSAM based on the NCMEC's database.

It will flag pictures that match a perceptual hash of pictures of child abuse. Now what legal kinds of pictures are most similar in composition, color, etc. to those offending pictures? What kinds of pictures would be hardest to distinguish from offending pictures if you were given only 16x16 thumbnails?

I'm going to bet the algorithm will struggle the most with exactly the pictures you don't want reviewers or the public to see.

Post reply on HN