They already scan for bombs and hazardous materials, so yes the line is drawn somewhere between ‘let anything get sent’ and ‘track everything’

Not sure I'm going to go all in on unqualified support, but it seems like comparing an image to a series of known hashes is qualitatively somewhat different than the postal inspector open all your mail. Though those convenient little scans they'll send you if your mail if you sign up suggests they do have some interest in who sends you mail already.

A closer comparison would be a machine opening your mail, scanning the contents, then using the same perceptual “hash” with an unknown database of “hashes”

If whoever controls that database wants to prohibit say a meme making fun of a politician, they need only to add the picture to the DB.

I think it would be horrible if they opened all mail... However, if they had a system that could detect CSAM in unopened mail with an extremely low false negative rate, then I'd be fine with it.

With many packages this already happens in the search for drugs and weapons and I have no problem with that either.

I believe millions more nudes are sent via snap/watsap/etcetc every day then are sent in the mail in a year.. with "extremely low false negative rate" - if that means there are a bunch of geeks and agents staring at your wife/gf.. your daughter/ etc.. is that okay? Some people will say yes - but I don't think those people should get to decide that all the other people in the world have to hand over all the nudes and dick pics everyone else has taken.

when they open mail with drugs and weapons it's most likely not a sculpture of your wife's genitals / daughter's tits, pics of your hemorrhoids / erectile implant - whatever other private pic stuff that no one else should be peering at.

If they were opening and reading every written word sent and received - I can guarantee you people would be up in arms about it.

> I believe millions more nudes are sent via snap/watsap/etcetc every day then are sent in the mail in a year.

I know for sure this is true, but this BS scenario was still used as a rebuttal against my comment, so I figured I'd give them an actual answer.

> when they open mail with drugs and weapons it's most likely not ... other private pic stuff that no one else should be peering at.

People order a lot of stuff that is very private online. From self help books, to dilos (which show up clearly on x-ray), to buttless chaps. If anything, packages are probably more private than mail these days, since people only mail postcards and bills.

> If they were opening and reading every written word sent and received - I can guarantee you people would be up in arms about it.

They would. But this is completely incomparable to what's happening on iOS devices now.

Maybe some crossed wires in the discussion - I thought your initial reply was in regard to what gp wrote about letters - and then you were saying about xrays and weapons and drugs.. which generally are not an issue with letters..

think I brought too many examples in basically trying to say flat paper.. yeah I think most people assume boxes/packages are xrayed/scanned by usps - and people do still send stuff they would not want public that way.

we are in agreement in that "They would. But this is completely incomparable to what's happening on iOS devices now."

I suppose one important distinction here, if this makes any difference, is that drugs and weapons (to use your examples) are physical items, and these could arguably cause harm to people downstream, including to the postal system itself and to those working in it. In contrast, photos, text, written letters, and the transmissions of such are merely information and arguably not dangerous in this context (unless one is pro-censorship).

The proliferation of CSAM is extremely harmful to the victims in the photos and more people seeing them might encourage more CSAM production in general.

I suppose I should clarify my point that I was referring to those dangerous items in the mail in the sense of their capacity to directly cause physical harm to those handling or receiving them, rather than in the more general sense of the societal effects of their proliferation, which is something else altogether (to your point).

To be clear, I'm not disagreeing with you in regards to the harm caused in the specific case of CSAM, but I can't help but see this as a slippery slope into having the ability in the future to label and act on other kinds of information detected as objectionable (and by whom one must also ask), which is itself incredibly harmful to privacy and to a functioning free society.