429
Teen deepfake victim pushes for federal law targeting AI-generated explicit content
(www.nbcnews.com)
this post was submitted on 18 Jan 2024
429 points (95.2% liked)
Technology
59569 readers
3825 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
There isn't ways to limit certain models without limiting all AI tech, which is what the first comment above from another user was saying. That corporations want to be the only ones using it by keeping it out of the hands of regular people, and this plays into that.
Something this powerful should absolutely be democratized, we should all have our own open source models, and unfortunately that means those smart glasses the guy on the bus is wearing could be undressing everyone in real time.
There's nothing to be done about it, and trying to do something is worse. It's like the war on drugs. Folks who want to do it are gonna do it. Fighting it is only going to make the world worse. Unfortunately there are victims here, but societally I think we're just going to have to get over it.
Other than vague slippery slope fearmongering I don't see how banning the creation and distribution of deepfake porn is going to make AI monopolized by corporations. If have your own personally trained and run AI model, you have complete control of what sort of content it's generating. Why would you have issues with deepfake porn laws if you are not generating and hosting that content?
It just doesn't add up, there's some logical leap here that seems almost on the level of conspiracy theories. As much as governments do tend to favor corporations over regular people there is nothing so far even vaguely suggesting that AI would be so profoundly restricted that only corporations could use it. In fact, what has been described of what is proposed so far does not target the technology at all, only the users who engage in this kind of bad conduct.
But I profoundly disagree with this "nothing to be done about it". How would fighting it be worse than letting people suffer for it? It's not like drugs where the main person who might have issues is the user themselves, this affects unrelated vulnerable people.
If it is identified who is making deepfake porn and where it's being hosted, it can be taken down. You could argue that not every single responsible person will be identified, but it might still be enough to diminish the prevalence and number of victims. And to the point that the remaining ones will have to be sneaky about it, that still might lead to less harassment to the victims.
You compare it to the war on drugs. Meanwhile I think of the rise of the automobile, with people crying that seat belts and traffic lights were ruining their freedom and "there's nothing to be done" about people dying in car crashes.
If everyone could create their own, and just run it locally, explain how the laws could be enforced?
Aguing that since you do a crime with a tool, outlawing the crime outlaws the tool is a bad argument. Outlawing murder doesn't outlaw knives.
As far as enforcement, it may be enforced with varying degrees of success but the argument that someone may get away with the crime also isn't a reason not to make it a crime.
If someone created deep fakes using locally run models, rubbed one out and then deleted everything they probably wouldn't be caught..but largely who cares that they didn't? It's the harm to others that it causes that you would largely like to prevent, and if a person didn't distribute the image at all them "getting away with it" doesn't matter much.
Edit: I think the argument that existing laws already cover this is more compelling than any of the above arguments as far as why this new law shouldn't be passed.
You conceded that no one cares if someone makes images locally then deletes them. But that's how they're all going to be made shortly.
Currently folks are sharing them because not everyone has the means to create them, some folks do, and share what they've made.
Once litterally every can just make them the moment they want to, no one will be sharing. Everyone will fall under that use case that you admitted no one would care about, which is exactly what I've been saying. It's 1. futile to try to stop, and 2. going to become so wide spread that we as a society will stop caring about it.
I do not think this is true. There are reasons to generate and distribute these other than to have a personal wank off gallery.
Like what? Why share something when anyone curious to see it can instantly generate their own?
I'm curious as to why you cannot come up with any yourself, but here are a few from the top of my head: to pass them off as authentic (likely for clout purposes), to have a laugh with the boys about it, to collaborate with others on them, and to distribute them to harass, ridicule, or disparage the target of them.
Degenerates exist in lots of shapes and forms, and not all degenerates will have enough of a sense of shame to be degenerates privately or to even know they are being degenerates at all.
I don't think you're properly understanding the paradigm shift that's coming with these models being open source and widely available while wearable AR smart glasses get better.
"You know Sharon is HR, look at this scandalous photo of her."
"Uh, I'm seeing a live generated porno of everyone in this room right now, why would I care about that."
And I don't think you're fully understanding that the above is some type of fantasy you have, and will not actually be what the future is like at all.
It's probably a bit of an exaggeration, but my point stands. It's going to be so easy for anyone to see ai gen material of anyone else, no one is going to care anymore.
I don't even think that's necessarily true. If you make it illegal and/or platforms ban it, you're already taking a step toward making it more difficult to do.
I think throughout this thread you're mistaking the technically possible for the probable or likely.
By making it illegal, you essentially eliminate the commercial incentive for making it easy. Every barrier to doing something makes it more unlikely that people will do it. I understand that there is an inherent motive for people to do it anyway, but, every hoop they have to jump through (e.g. setting up their "own, local AI") reduces the likelihood of them doing it.
People don't even run their own email servers, music servers, video servers, etc. etc. etc....most people don't even "jail break" devices...many don't even store a local cache of regular porn...why the hell would most people bother themselves with setting up a local generative AI instance for this purpose?
Outlawing it and banning it from platforms makes it much more within the realm of the creepy basement weirdo rather than something that is as inevitably ubiquitous as you're saying it will be.
Policy is very often about reduction of harms rather than elimination of harms. It's not the black and white realm that you're trying to make it out to be.
It's not illegal to to work on, sell, or distribute the models. And making that illegal is what the first commenter said would be dangerous to do, since then regular people wouldn't be able to compete with corporation's abilities.
Once the models and portable hardware are good enough, and it's just a matter of time, I think you're underestimating how ubiquitous it will become.
Every teenage boy will have a pair of nudie glasses in the form of their smartphone running open source models, and you think they're just going to not use them?
I think you again vastly overestimate how many people are going to run their own AI versus using a sanitized, policy-driven, managed platform version that's cloud based (e.g. Dall-E and ChatGPT right now).
It's possible today (and usually better) to do a lot of things locally, but yet still almost everything routes through an app to a platform on your smartphone and the few remaining things that don't route through a platform using your phone's browser.
When it becomes one click to see the chick across from you naked, tell me how many 16 year old boys won't. You are far too naive to be having this conversation.
It's not naive to think that corporations will continue to win the "AI" war. It's actually pretty naive to think otherwise.
I also dunno why you think that all of the resources in oss AI will focus their efforts on making it easy to generate excellent, likely already illegal deep fake porn of random teenagers in "one click".
I've been using oss for decades and almost nothing is that easy to do even when it could be. Why would people focus their efforts on this?
Also also, I don't get why you think that generating AI porn of people around you is:
A) so much better than just watching the millions of hours of already available porn
B) anything even remotely similar to "seeing someone naked"