OK, so this is just the general anti-AI image generation argument where you believe any image generated is in some meaningful way a copy of every image analyzed to produce the statistical model that eventually generated it?
I’m surprised you’re going the CSAM route with this and not just arguing that any AI generated sexually explicit image of a woman is nonconsensual porn of literally every woman who has ever posted a photo on social media.
was seeded with the face of a 15yr old and that they really are 15 for all intents and purposes.
That’s…not how AI image generation works? AI image generation isn’t just building a collage from random images in a database - the model doesn’t have a database of images within it at all - it just has a bunch of statistical weightings and net configuration that are essentially a statistical model for classifying images, being told to produce whatever inputs maximize an output resembling the prompt, starting from a seed. It’s not “seeded with an image of a 15 year old”, it’s seeded with white noise and basically asked to show how that white noise looks like (in this case) “woman porn miniskirt”, then repeat a few times until the resulting image is stable.
Unless you’re arguing that somewhere in the millions of images tagged “woman” being analyzed to build that statistical model is probably at least one person under 18, and that any image of “woman” generated by such a model is necessarily underage because the weightings were impacted however slightly by that image or images, in which case you could also argue that all drawn images of humans are underage because whoever drew it has probably seen a child at some point and therefore everything they draw is tainted by having been exposed to children ever.
97% of the internet has no idea what Matrix channels even are.
I’ve been able to explain it to people pretty easily as “like Discord, but without Discord administration getting to control what’s allowed, only whoever happens to run that particular server.”
A more apt comparison would be people who go out of their way to hurt animals.
Is it? That person is going out of their way to do actual violence. It feels like arguing someone watching a slasher movie is more likely to make them go commit murder is a much closer analogy to someone watching a cartoon of a child engaged in sexual activity or w/e being more likely to make them molest a real kid.
We could make it a video game about molesting kids and Postal or Hatred as our points of comparison if it would help. I’m sure someone somewhere has made such a game, and I’m absolutely sure you’d consider COD for “fun and escapism” and someone playing that sort of game is doing so “in bad faith” despite both playing a simulation of something that is definitely illegal and the core of the argument being that one causes the person to want to the illegal thing more and the other does not.
…and most of the people who agree with that notion would also consider reading Lemmy to be “trawling dark waters” because it’s not a major site run by a massive corporation actively working to maintain advertiser friendliness to maximize profits. Hell, Matrix is practically Lemmy-adjacent in terms of the tech.
eventually they want to move on to the real thing, as porn is not satisfying them anymore.
Isn’t this basically the same argument as arguing violent media creates killers?
They have to deal with old men masturbating to them getting raped online.
The moment it was posted to wherever they were going to have to deal with that forever. It’s not like they can ever know for certain that every copy of it ever made has been deleted.
it says “this hidden site”, meaning it was a site on the dark web.
Not just on the dark web (which technically is anything not indexed by search engines) but hidden sites are specifically a TOR thing (though Freenet/Hyphanet has something similar but it’s called something else). Usually a TOR hidden site has a URL that ends in .onion and the TOR protocol has a structure for routing .onion addresses.
Even then, a common bit you’ll hear from people actually defending pedophilia is that the damage caused is a result of how society reacts to it or the way it’s done because of the taboo against it rather than something inherent to the act itself, which would be even harder to do research on than researching pedophilia outside a criminal context already is to begin with. For starters, you’d need to find some culture that openly engaged in adult sex with children in some social context and was willing to be examined to see if the same (or different or any) damages show themselves.
And that’s before you get into the question of defining where exactly you draw the age line before it “counts” as child sexual abuse, which doesn’t have a single, coherent answer. The US alone has at least three different answers to how old someone has to be before having sex with them is not illegal based on their age alone (16-18, with 16 being most common), with many having exceptions that go lower (one if the partners are close “enough” in age are pretty common). For example in my state, the age of consent is 16 with an exception if the parties are less than 4 years difference in age. For California in comparison if two 17 year olds have sex they’ve both committed a misdemeanor unless they are married.
It also likely gives you the best $ spent/children protected rate, because you know the producers have children they are abusing which may or may not be the case for a viewer.
An LLM is roughly as smart as the corpus it is summarizing is accurate for the topic, because at their best they are good at creating natural language summarizers. Most of the main ones basically do an internet search and summarize the top couple of results, which means they are as good as the search engine backing them. Which is good enough for a lot of topics, but…not so much for the rest.
https://tubitv.com/movies/332938/repo-the-genetic-opera
I guess that could be the specific dystopia we’re aiming for. It’s a hell of a race to beat the other dystopias/apocalypses there though. But what else would you expect from a type 13 planet in it’s final stage?
I don’t think this is the kind of neuroplasticity we’re supposed to be aiming for.
Not quite. KiwiFarms is older than 8Chan. KiwiFarms was originally called CWCki and was dedicated specifically to Chris-chan because Encyclopedia Dramatica wasn’t quite what some of that person’s harassers wanted it to be. 8chan wouldn’t open for another 8 months after CWCki was founded but before it was renamed KiwiFarms.
Let’s be clear here - Chris-chan has been systematically harassed for literal decades, has a bunch of problems at least some of which are consequences of that. Chris-chan is also probably the single most documented human being in history as a consequence of the massive and dedicated “fan” base.
Pretty sure there’s at least one quote of Davis himself using the n-word to refer to federal agents before describing how bright they glow…
I mean, it would suck for any mod using F4SE. The answer for what to do is the same as every other update - recommend people not update until F4SE is updated.
Did they have their own plugin DLL or were they just using F4SE as is? If the former that would make it suck for them even worse since they’d potentially need to find some new hook addresses of their own and wait for the new F4SE and then reconcile their DLL.against that and then test it all again to make sure nothing broke.
So basically what happened with the AE release for Skyrim SE where Bethesda switched to a new compiler version and the tool the script extender team was using to find the correct offsets couldn’t handle it so they had to track down the offsets manually like before they’d written the tool, leading to a longer than usual time for the script extender to update than usual?
Which if it’s anything like SSE means that mods that didn’t use F4SE were basically unaffected, mods that use F4SE had to wait for it to update which took longer than usual after which they would mostly work unmodified, and mods that involved a plugin DLL for F4SE had to at the very least be recompiled against the new versions of the game and F4SE. Nothing about that specifically targets Fallout: London though from what it sounds like.
The tippy part is they go off the OG lore so Cyrodiil is a more tropical/Mediterranean climate which is fun.
Fucking Thalmor denying the power of Talos of Atmora.
Seriously though, the canon explanation for Cyrodiil being the way it is now as opposed to original lore is that when Talos achieved CHIM he changed it, because that’s a thing you can do with the secret syllable of royalty. All part of the path to mantling Shor/Lorkhan via one of the Walking Ways and forging an empire.
I’ve joked in other places in the past that CHIM stands for “Character Has Installed Mods” because what it allows is roughly on par with the character opening the modding tools and changing what they want to change.
Both very soon after release by an update which was specifically designed to break the mod.
I’m now curious about this from a technical perspective - how did the update specifically break their mod in particular? Were they doing a bunch of custom DLL hooks or something?
I know with Skyrim SE modding it’s usually that any update breaks SKSE and a tiny handful of other mods that directly hook DLLs or the executable (these mods are usually scripting engine extensions and are a dependency for a variety of other mods), and depending on the update sometimes it takes longer than average to get a new version of those running (the AE update was one of those because they switched compiler version and that broke the method SKSE was using to find hooks). But in general that only breaks 1) mods using those (think SkyUI) until a new version comes out, after which most of those mods start working again without the individual mods needing an update and 2) mods that include their own plugin DLL, (think SkyClimb) which have to wait on an update and then compile a new version of the DLL for the new version of both the game and the other mod, because addresses and functions they are hooking may have changed. Mods not using SKSE or similar generally run just fine between versions of SSE (including AE).
To be clear, when you say “seeded from” you mean an image that was analyzed as part of building the image classifying statistical model that is then essentially running reverse to produce images, yes?
And you are arguing that every image analyzed to calculate the weights on that model is in a meaningful way contained in every image it generated?
I’m trying to nail down exactly what you mean when you say “seeded by.”