I have talked about it with a high-ranking french policeman. That person is mostly active in fighting sex-crimes on children, which is the angle I will mostly be referring to. From what I understood, it is very clear to them that even if these laws comes to pass, a good amount of criminal activities will move to other safer options. However, the general criminal is not technically competent. Currently, with WhatsApp providing end-to-end by default, access to pedophilic content is extremely simple. By suppressing these simple means of end to end encryption, the goal is to reduce the amount of people accessing these networks due to a higher entry bar.
What's of course concerning is that it renders anyone using encryption suspect, which includes pedophiles and narcotrafficants, but also activists and co.
Also, if we're only targetting pedophile networks, one option that comes to mind to me is the following : Most of those images are known and have been circulating for a while. By hashing any sent images and comparing them to the checksum of known ones, one could easily flag suspicions senders and proceed to access the phones of those users. Does that seem feasible to you or am I missing something?
It seems so, yes. While I do trust eff, it seems to me that their article barely skims over the explanation of why this is a problem, although the second one does mention the ability to arbitrarily decide what triggers the filters.
I would however like to point out that in France, the police cannot arbitrarily arrest people for more than 24 hours, after which they need an investigating judge's approval to prolongate detention. They also need those judges' permission to access a device. Free access to any channels of communication has never been on the table, but extrapolation of that technology to other kinds of governments with more liberal law-enforcement remains the obvious issue.
Still, I kind of fail to see how full privacy as a default is a necessity, if and only if it remains a possibility.
Furthermore, by using non open source messengers such as WhatsApp, we are blindly trusting Zuckerberg, a random dude who got lucky and rich and wishes to remain on good terms with Trump, to keep our data as safe and as unreachable as he pretends.
First, the French state has zero interest in access to these messages to help children. The proof: NEITHER the French police NOR the secret service investigate child abuse cases. The police only investigates them if they absolutely can't avoid it, because mostly people with access to children commit child abuse. You will find child abuse requires a child. In other words, who does that? Teachers, sports teachers and coaches and child welfare workers (sports clubs are almost exclusively government funded in France). The large majority of the perpetrators, of course, are government employees. Child services investigates child abuse cases, and WILL NOT get access to these messages. So there is zero intention to give access to messages in child abuse cases.
I mean how ridiculous is this argument. They want access to such messages to investigate child abuse cases, so they demand French spies, and tax investigators get access to everyone's messages ... Child abuse investigators are not even mentioned.
And it's not just that.
Next, France is famous in Western Europe for being one of the only EU countries where access to, uh, hentai comics, is legal and they're sold in newspaper stands.
If the French state cared about fighting sex-crimes on children they would fund taking care of the children they do "help", rather than catching criminals. Instead, this is what they do:
(at the very least they let it happen, but in practice they also hire people that will do this job at a very cheap wage because it provides access to vulnerable children)
Without fixing this FIRST, the only thing catching criminals will do, obviously, is make the situation of children worse. The French state fails this test.
The situation with French schools, both the immigrant situation AND the constant decline in teacher quality (for at least 3 decades now) show how much the state cares about children's future in general. Again, the state fails the test completely.
And I haven't even mentioned the refugee situation in Paris. Obviously that situation is producing a flood of child prostitution. Again, the state is showing itself unwilling to help children. Again, the French state is exposed as not doing shit to help children, or at the very least, they're totally ineffective.
So no, and sorry to state the obvious, but your suggestion is completely beside the point.
Most of what you said is may be true but not the part about hentai ? France does block, without asking a judge, most of the big hentai websites with a problematic catalog, and you can't sell those IRL either !
I actually saw this in a bookshop (more of a very big model newsstand, really) in Lille, about 4 years ago. At that time, it certainly was still there.
I would love to take the time to inspect your answer more thoroughly as soon as I have the time, but your premise is wrong. Most CSAM creators are neither teachers nor coaches, but parents i). The French state is not some cynical entre-soi that protects child abusers who are on their payroll. In my personal experience, which, while it isn't generalizable does prove the existence of consequences, I've come across three teachers accused of incorrect behavior with children ii). All three disappeared from the schools within months.
i)> Research suggests that a significant proportion of child sexual abuse material (CSAM) is produced and distributed by parents who victimise their children. An online convenience sample of 150 adult survivors of CSAM found that, of those abused by a single perpetrator, 42% identified their biological or adoptive father or stepfather as the offender; and of those abused by multiple perpetrators, 67% identified their biological or adoptive parents or step-parents as the primary perpetrators (Canadian Centre for Child Protection (CCCP), 2017). https://bravehearts.org.au/research-lobbying/stats-facts/onl...
ii) the first one had us do sexual education at age 8 and was gone five weeks after he began, the second had been on the radar for racism and was gone two months after complaints of staring at girls skirts, the third, a sport teacher, disappeared at the end of the semester for systematically correcting girl's stances while squatting and such. In the two last cases, the schools were relatively big and rumors of worse offenses were around, but I don't know if those were true.
> i)> Research suggests that a significant proportion of child sexual abuse material (CSAM) is produced and distributed by parents who victimise their children. An online convenience ...
In case you seriously don't understand how you're misleading people: this is an organization that is FORBIDDEN by law to investigate cases of CSAM where the parents aren't involved. And CSAM is not their focus. Their focus is placing children.
In other words what have you proven with this study?
If you exclude all perpetrators except the category you want to accuse, then only exactly the people you want to blame are guilty. Proving all ravens are white by excluding any black ones, in other words.
The (wrong) summary study you linked to by the way, when it mentions non-family CSAM studies always has thousands to tens of thousands of cases, but when mentioning parental involvement they have 1 study with 150 examples. The other case of CSAM they mention in Canada is about "sextortion", ie. going further than just CSAM, but using CSAM to force a minor to do sexual acts. It happened exclusively at schools and talks about 23000 cases ...
Let me do the math with those numbers, even if I realize this is not a great way to compare and the numbers are not remotely complete, but let's use the numbers we have: 150 parents involved, 23000 = 0.6%. If that number can be an indication of the ratio, even with 500% error ...
Sometimes online people use these sorts of numbers as justification for that we should just totally stop investigation of biological parents in child welfare because even if you totally prevented ALL biological parent child abuse, you would have barely made a dent in child abuse as a whole. In fact there's studies claiming that because of the focus on biological (especially single female) parents child welfare agencies cause numbers of child abuse cases to go UP, not down.
Note: the study you pointed to clearly states that when it comes to single female parents, NOT A SINGLE ONE was even involved (not even unknowingly involved) in CSAM. Not one.
This is despite that being the main focus of C3P. Child abuse by single mothers is what they spend most of their attention on, how most children get placed, despite not being able to show a SINGLE case of it leading to CSAM.
So there's the problem: states force investigations away from actual child abuse. Here's how it works: Police is legally forced to refer child abuse cases to child welfare agencies, without investigation (ie. they CANNOT decide to investigate by themselves, unless there are other serious crimes). Child welfare agencies are explicitly forbidden from investigating schools, and especially forbidden from investigating the child welfare agencies themselves (despite reporters exposing a "prostitution ring" or the like inside child welfare agencies in every country every few years). You will have no difficulty finding stories of ex-"placed" children who were forced into prostitution, on facebook, tiktok, ... nor will you have problems finding reporters exposing child welfare prostitution with other government employees involved (e.g. Rotherham) regularly.
Another thing about child welfare agencies that get caught prostituting children they're supposed to protect: the numbers are absurd. Rotherham exposed that UK child welfare employees and city hall employees had organized and received payment for 1400 confirmed rapes. Claims that the total number was over 20000 are rife online. One might remark that ALL child sexual abuse cases in London in 10 years is less than 5000. That ONE case of child services sexual abuse made more victims than all other cases in the entire London area (15 million people) for 3 to 20 years COMBINED.
You know how the government refers to the perpetrators, by the way? They refer to child welfare employees and city hall employees by these words in all articles about it "Pakistani men" (and, of course, only about half were actual Pakistani, by the way. Especially in the city hall employees, no shortage of Brits. Also the police officers involved were British)
> The French state is not some cynical entre-soi that protects child abusers who are on their payroll
I've gone to school, not in France, but close enough to hear stories.
> All three disappeared from the schools within months.
Exactly. Well, they were mostly moved from one school to another, I take it? That's the case for the stories I heard. Note that you explicitly DID NOT say punished or persecuted, despite of course committing crimes. You don't even mention they were investigated at all. And they weren't.
One of the stories I heard was about a gym teacher having placed a camera in the women's showers by the way. He was caught, and indeed disappeared from school after a second incident. But ... what do you suppose happened to the videos? Nobody even went looking for them. What happened to all the videos of the ones who weren't caught?
Oh, and, three is a rather large number for one student to encounter, don't you think?
I hope you can at least agree there's a problem here. AND, that giving spies and legally-forbidden-to-investigate police officers access to private messages is very unlikely to help in those cases.
I have talked about it with a high-ranking french policeman. That person is mostly active in fighting sex-crimes on children, which is the angle I will mostly be referring to. From what I understood, it is very clear to them that even if these laws comes to pass, a good amount of criminal activities will move to other safer options.
However, the general criminal is not technically competent. Currently, with WhatsApp providing end-to-end by default, access to pedophilic content is extremely simple. By suppressing these simple means of end to end encryption, the goal is to reduce the amount of people accessing them due to a higher entry bar.
What's concerning to me is that it renders anyone using encryption suspect, which includes pedophiles and narcotrafficants, but also activists.
Also, if we're only targetting pedophile networks, one option that comes to mind to me is the following :
Most of those images are known and have been circulating for a while. By hashing any sent images and comparing them to the checksum of known ones, one could easily flag suspicions senders and proceed to access the phones of those users. Does that seem feasible to you or am I missing something?
Phrases like "actual understanding", "true intelligence" etc. are not conducive to productive discussion unless you take the trouble to define what you mean by them (which ~nobody ever does). They're highly ambiguous and it's never clear what specific claims they do or don't imply when used by any given person.
But I think this specific claim is clearly wrong, if taken at face value:
> They just regurgitate text compressed in their memory
They're clearly capable of producing novel utterances, so they can't just be doing that. (Unless we're dealing with a very loose definition of "regurgitate", in which case it's probably best to use a different word if we want to understand each other.)
The fact that the outputs are probabilities is not important. What is important is how that output is computed.
You could imagine that it is possible to learn certain algorithms/ heuristics that "intelligence" is comprised of. No matter what you output. Training for optimal compression of tasks /taking actions -> could lead to intelligence being the best solution.
This is far from a formal argument but so is the stubborn reiteration off "it's just probabilities" or "it's just compression". Because this "just" thing is getting more an more capable of solving tasks that are surely not in the training data exactly like this.
Thanks, they seem like more than just random splashes of color.. possibly I'm anthropomorphising but it feels like it was straining to draw something specific like a young child would.
I've found another[1] on a blog post[2], captioned as follows:
Frontispiece 1. Art drawn by chimpanzee Ai using sharpies(Saito, 2008)[p.19]
Frontispiece 2. Art styles of 4 adult chimpanzees(Saito, 2008). Guess which one was by Ai[p.20]
Not sure what the background of the author is, but this essay/lecture note discusses ego or literal self-awareness of apes contrasted against human children, using quotes from books. Apparently apes don't exhibit explosive growth of vocabulary, show use of syntax etc etc, and are therefore not able to acquire language. The post later also argues their ego may be on the edge of formulating but must be weak/incomplete.
There's also magazine excerpt[3] on a page on relevant Kyoto University research center comparing an inpainting task done by a chimpanzee and a human child of 3 years old, showing that chimpanzees can only recognize and trace existing patterns, whereas kids go and complete the face with eyes, nose and mouth.
I agree there is intent there, but it doesn't look like an effort to draw a still life, more like the chimp was fascinated with the patterns and techniques it could manipulate.
Yes, same with Koko. I think they do not fully understand art and
abstraction, nor profits made by good art. It is too abstract.
They can, however had, understand sign language and symbol language,
and basically that art is also an abstraction. Will probably take a
while before we can identify abstract art by apes.
What's of course concerning is that it renders anyone using encryption suspect, which includes pedophiles and narcotrafficants, but also activists and co.
Also, if we're only targetting pedophile networks, one option that comes to mind to me is the following : Most of those images are known and have been circulating for a while. By hashing any sent images and comparing them to the checksum of known ones, one could easily flag suspicions senders and proceed to access the phones of those users. Does that seem feasible to you or am I missing something?
reply