Grok floods X with sexualized images of women and children: Grok generated an estimated 3 million sexualized images, including 23,000 of children in 11 days

The CEO can’t be held personally liable for the actions of the company, for example; their underlings could have been responsible and kept the leader in the dark.

The onus should be on the company to prove their employees kept the CEO in the dark, not the other way around.

That is a tricky question. IT isn't just does the CEO know, but should the CEO have known. If you make a machine that injures people the courts ask should you have expected that.

The first time someone uses a lawnmower the cut a hedge the companies and gets hurt can say "we never expected someone to be that stupid" - but we now know people do such stupid things and so if you make a lawn mower and someone uses it to cut a hedge the courts will ask why you didn't stop them - the response is then we can't think of how to stop them but look at the warnings we put on.

When Grok was first used to make porn X can get by with "we didn't think of that". However this is now known. They now need to do more to stop it. there are a number of options. Best is fix Grok so it can't do that; they could also just collect enough information on users that when it happens the police can arrest the person who instructed grok. There are a number of other options, if the court accepts them depends on if the tool is otherwise useful and if whatever they do reduces the amount of porn (or whatever evil) that gets through - perfection isn't needed but it needs to get close.

can we give a random teen the nuclear codes?

It wasn’t just one salute. Fucker did it twice with gusto.

If people are still on his platform after this and the grok thing, they are stupid, evil, or both.

shakes fist at cloud back in my day, that's what 4chan was for!

And that's the "leftist propaganda" according to MAGA. PBS and NPR have so overcorrected, and in exchange have gotten nothing but defunded

it does it to pictures of real 12 year olds and nudes too, not just the nazi bikinis (but also that)

Have you seen who's in charge of our government? They're probably the ones making the prompts

Because the US seems to be supportive of it. They lashed out at governments who've said to resolve it or face site bans.

The blame kinda rests with the creators of the tool as well as the "content" creators, no? And isn't it more important to inform people about what a service is allowing than pointing out what specific users are doing? This is ultimately a reflection of X as well as X users.

btw if you or anyone still uses X, it's fair to assume you're a pedo and or a nazi.

X is for nazi pedos

This is what I want to know. It doesn't do anything unless someone tells it to. Why aren't the people telling it to make child porn being held accountable? My 3D printer can make guns, but they won't send the printer to jail if I decide to tell it to make one.

"Twitter users use child porn machine to make millions of images of child porn." Would be accurate imho. Legally though AI generated works are owned by no one because the machine made it but cant own it.

I mean, yeah, but it's one thing if some perv's running it on their own box after reading 5 guides vs. elon musk having a tiwtter bot that does it for you without even trying to stop it from doing that even after knowing it's doing it. the former is unavoidable, the latter is a choice he's made for some fucking reason

I agree.

It's fine if people want to get mad about it, but it's more effective to just learn to live with it because it's not going away.

Weren't they already picking out their wives from public nudes?

Yeah NGL. Kind of weird that all the noise is around the machine, that can't be held accountable, because it is a machine, and not on the people actually making the images.

This is like blaming the camera for recording something illegal.

Maybe they should look at the hand holding the tool if they want accountability.

Most AI platforms allow sexualized content to varying degrees. Google, Instagram, Tiktok, etc all host CSAM. Always have. It's been understood that they're not liable to the extent that they remove it when reported. Their detection technology is pretty good as to automatically handle it, but never perfect. They keep track of origins and comply with subpoenas which has gotten tons of people convicted.

Grok image gen was put behind a paywall, which people claim is worse. However, most people paying lose anonymity and thus can be appropriately handled when they request illicit content, even if Grok refuses the request like it usually does.

I think the Grok issue is sensationalized and taken out of context of the realities of what happens online and in law enforcement.

Grok didn't make itself.

The images are being generated by editing photos of real children. It's not fictional characters. This is discussed in the article and has been widely reported as this issue has been in the news.

I think this code could be made more efficient