3 ms·

It’s not a distinction without a difference. ‘Kind of like’ is going to be read by most people as ‘easily fooled’. That is a reflection of the false positive

by zepto 5y ago

It’s not a distinction without a difference. ‘Kind of like’ is going to be read by most people as ‘easily fooled’. That is a reflection of the false positive rate. It all depends on the quality and tuning of the algorithm and how it is deployed. If you are going to imply false positives are common, then you need to back it up. Nobody is saying false positives are impossible. Apple is saying false positives are on the order of one in a trillion per user account per year. That doesn’t sound like something that matches images that are only ‘kind of similar’. Yes - cryptographic hashes have much lower false positives rates even than that, but that is a distinction without a difference since both make the risk negligible. > The way perceptual hashing works is that an image is shrunk down to a 8x8 or 26x26 image and then Which is it for Apple’s hashes? There is no point in reading old articles about perceptual hashes if the conclusions don’t apply to Apple’s neuralhash algorithm. If they don’t then reading about other hashes is just a distraction. What can you tell us about the likelyhood of Apple’s hashes to create false positives?

heavyset_go 5y ago
> It’s not a distinction without a difference. ‘Kind of like’ is going to be read by most people as ‘easily fooled’. I'm not really concerned with what you're afraid most people will think. Two images that kind of look like one another will have the same or similar hashes. There are literal examples of this in the links I posted above. And it's literally the point of perceptual hashing, to find images that look similar to a source image by comparing hash similarity. > If you are going to imply false positives are common, then you need to back it up. I just did with two links I posted above. Twice. > Apple is saying false positives are on the order of one in a trillion per user account per year. Sounds like a claim that wasn't replicated or independently verified. Of course Apple is going to say their system is nearly perfect, that's what all companies do. The onus is on Apple to prove that their marketing claims reflect reality. > There is no point in reading old articles about perceptual hashes if the conclusions don’t apply to Apple’s neuralhash algorithm. If they don’t then reading about other hashes is just a distraction. The onus is on Apple to demonstrate that their methods are remarkably different from the rest of the science and industry. This is like saying the normal principles of computing don't apply to new Apple products because they might have invented a new brand computing paradigm that isn't anything like any classical or quantum computer mentioned in scientific literature at all. Yeah, maybe they did, but it's unlikely and the onus is on Apple to prove it.
- zepto 5y ago
  > I'm not really concerned with what you're afraid most people will think. What matters is not what I think, but whether you care about making misleading comments. > > If you are going to imply false positives are common, then you need to back it up. > I just did with two links I posted above. Twice. No, you posted some links that are not about Apple’s system, and you can’t explain how they apply presumably because you don’t understand what Apple is doing. > Sounds like a claim that wasn't replicated or independently verified. Of course Apple is going to say their system is nearly perfect, that's what all companies do. The onus is on Apple to prove that their marketing claims reflect reality. So this tells us you don’t know what algorithm Apple is using… …And are accusing Apple of lying, when it is clear that you haven’t read about how they avoid false positives. I think the onus is on you to prove your accusation. > This is like saying the normal principles of computing don't apply to new Apple products because they might have invented a new brand computing paradigm that isn't anything like any classical or quantum computer That just silly. It’s doesn’t take breaking the laws of quantum or classical computing to build a system with a low false positive rate. One obvious way would be to leverage multiple images rather than just one. Increasing the sample size of a population sample generally reduces the false positive rate. Have you considered that someone could build a system this way?
  - zekrioca 5y ago
    Very patronizing of you. You clearly never learned that both classical (like your "sample size" example) and advanced statistical models are susceptible to first-order evasion attacks (or “Adversarial Examples”) that fool models at run-time. But go trust Apple's 1/1.000.000.000.000 claims.
    - zepto 5y ago
      Obviously such an attack has nothing to do with the false positive rate. But, setting that aside, can you explain how a first order evasion attack can be used against Apple’s mechanism? They are a real kind of attack in the lab, but it’s not obvious how they could be used to exploit Apple’s CSAM detection. If you have reason to think they are a real threat, I’m sure you can explain.

by zepto 5y ago

heavyset_go 5y ago

> It’s not a distinction without a difference. ‘Kind of like’ is going to be read by most people as ‘easily fooled’. I'm not really concerned with what you're afraid most people will think. Two images that kind of look like one another will have the same or similar hashes. There are literal examples of this in the links I posted above. And it's literally the point of perceptual hashing, to find images that look similar to a source image by comparing hash similarity. > If you are going to imply false positives are common, then you need to back it up. I just did with two links I posted above. Twice. > Apple is saying false positives are on the order of one in a trillion per user account per year. Sounds like a claim that wasn't replicated or independently verified. Of course Apple is going to say their system is nearly perfect, that's what all companies do. The onus is on Apple to prove that their marketing claims reflect reality. > There is no point in reading old articles about perceptual hashes if the conclusions don’t apply to Apple’s neuralhash algorithm. If they don’t then reading about other hashes is just a distraction. The onus is on Apple to demonstrate that their methods are remarkably different from the rest of the science and industry. This is like saying the normal principles of computing don't apply to new Apple products because they might have invented a new brand computing paradigm that isn't anything like any classical or quantum computer mentioned in scientific literature at all. Yeah, maybe they did, but it's unlikely and the onus is on Apple to prove it.

zepto 5y ago
> I'm not really concerned with what you're afraid most people will think. What matters is not what I think, but whether you care about making misleading comments. > > If you are going to imply false positives are common, then you need to back it up. > I just did with two links I posted above. Twice. No, you posted some links that are not about Apple’s system, and you can’t explain how they apply presumably because you don’t understand what Apple is doing. > Sounds like a claim that wasn't replicated or independently verified. Of course Apple is going to say their system is nearly perfect, that's what all companies do. The onus is on Apple to prove that their marketing claims reflect reality. So this tells us you don’t know what algorithm Apple is using… …And are accusing Apple of lying, when it is clear that you haven’t read about how they avoid false positives. I think the onus is on you to prove your accusation. > This is like saying the normal principles of computing don't apply to new Apple products because they might have invented a new brand computing paradigm that isn't anything like any classical or quantum computer That just silly. It’s doesn’t take breaking the laws of quantum or classical computing to build a system with a low false positive rate. One obvious way would be to leverage multiple images rather than just one. Increasing the sample size of a population sample generally reduces the false positive rate. Have you considered that someone could build a system this way?
- zekrioca 5y ago
  Very patronizing of you. You clearly never learned that both classical (like your "sample size" example) and advanced statistical models are susceptible to first-order evasion attacks (or “Adversarial Examples”) that fool models at run-time. But go trust Apple's 1/1.000.000.000.000 claims.
  - zepto 5y ago
    Obviously such an attack has nothing to do with the false positive rate. But, setting that aside, can you explain how a first order evasion attack can be used against Apple’s mechanism? They are a real kind of attack in the lab, but it’s not obvious how they could be used to exploit Apple’s CSAM detection. If you have reason to think they are a real threat, I’m sure you can explain.