In the latest in a string of privacy abuses from the chatbot, Grok provided porn performer Siri Dahl's full legal name and birthdate to the public, information she'd protected until now.
I used to work in the porn industry years and years ago as a developer (I made the websites) so I can assume how grok got this info.
in the US at least they have to keep information of all their talent that they use to produce content for a law called “2257”. Essentially what would happen is if we were doing photo sets or videos before the content there would be a couple photos or short video of the talent holding their ID to the camera. We’d then take that stuff and put it into a DB and if need be file it with the US government to comply with 2257. Basically this just ensures that the talent is of legal age. For one company I worked for we literally had two ladies that spent every day all day filing this information and sending it off. Because of this they were also the ones that would approve if we could use the content or not.
So you can now safely guess how Grok was able to get access to this info. For whatever reason Grok, and X (probably via DOGE) has access to the 2257 database. Why it would need/want access to this? well…come on we know the reason.
You’re saying that Musk training his AI on doge data, and then using that training for Grok is unbelievable? Or do you want concrete evidence?
I mean Musk’s complete disregard for rules and regulations, combined with his own and his cronies’ incompetence, makes me think that, while no evidence exists, the scenario would definitely not be out of the realm of possibilities.
I’m simply asking if there any reason to believe this. You know, beyond, “Oh yeah, that’s totally something Musk would do?” As crimes go, it would be like writing your own name and address on a bank’s wall.
Apparently the answer is no, we have no evidence to support this hypothesis.
If you were an arrogant narcissistic maniac who knew for absolutely certain that you could rob that bank in full daylight, and NEVER be punished for it, of course you would write your name on the wall.
I’m not going to fabricate stories, but if the explanation for something is “AI was used irresponsibly” then I would tend to not dismiss the idea. And I’d be prone to propagate the idea far more. It’s not a conscious choice mind you, that statement is a result of self-reflection.
The way people unquestionably, irresponsibly, and without the slighest contemplation of the ramifications, are using AI for the most mundane tasks, worries me. Why do you need to make a photo of someone, who haven’t accepted the use case, into a drawing? Why do you need to have an AI transcribe our meeting and summarize the points? Nobody’s going to read the summary anyway, and now some unknown entity knows what other people said with an otherwise ordinary expectation of privacy.
One thing is that tech overlords have already scraped every image, sound bite, video clip, and written word publicly, and semi-publicly, available. Don’t give them more material.
The personal risk isn’t in talking to AI, it’s in listening to AI. And our environment is already becoming crowded with it. There is an argument that AI is disproportionately expensive and exploits externalities, but it’s hard to take that argument seriously from meat eating motorists.
Hey hey now, calm down, there’s no need to get personal here :-) We both dislike AI, can’t we just agree to hate AI and leave it at that? Like Legolas and Gimli in LotR? I’ll even let you toss me…
My main concern about AI is that it’s rotting our cognitive abilities, yours seem to be founded in an environmental stance. I can respect that, and I think it’s a valid argument.
The personal risk isn’t in talking to AI, it’s in listening to AI.
I think both are bad, but can we agree to “AI bad. Don’t use AI”?
Yeah, both are owned by Musk. Go ahead and tell me that I’M the idiot for believing that he’d share his DOGE-stolen databases with his other enterprises.
I’m not saying that your argument isn’t plausible. But I have another route for this info.
Grok is training on tweets. Some person tweets “I just found out that my neighbor, Jane, is actually the pornstar Siri Dahl”. That tweet never gets traction, so only 5 followers reads it, but so does grok. Then in another tweet same person combines the firstname “Jane” with lastname “Doe”.
If your explanation is correct, as to where the info got leaked, I’d guess we’d see way more pornstars doxed.
Alternatively, it could be using image databases. Someone somewhere could have obscurely tweeted a photo of this performer off-duty with a note “here I am with my friend/relative/etc.” and included her legal name, and the image-recognition algorithms could have connected that with her performance images and went from there.
Of course, I mean to Trump a 17yo is at best a MILF. Besides she’d be used goods. Well maybe Epstein could have used the girl for less discerning perverts like Andrew Mountbatten-Windsor, but it would seem that both Einstein’s and Andrew’s days of raping kids is over…
Although one could hope that Andrew’s involvement in rapes aren’t over. Now that he’s in jail he has plenty of opportunities to experience spontaneous unwanted violent love making.
Sorry but whenever I hear this wording, I always imagine the porn lobby had a meeting one day to decide on a better word for ‘meat’ and they ended up on ‘talent’
I used to work in the porn industry years and years ago as a developer (I made the websites) so I can assume how grok got this info.
in the US at least they have to keep information of all their talent that they use to produce content for a law called “2257”. Essentially what would happen is if we were doing photo sets or videos before the content there would be a couple photos or short video of the talent holding their ID to the camera. We’d then take that stuff and put it into a DB and if need be file it with the US government to comply with 2257. Basically this just ensures that the talent is of legal age. For one company I worked for we literally had two ladies that spent every day all day filing this information and sending it off. Because of this they were also the ones that would approve if we could use the content or not.
So you can now safely guess how Grok was able to get access to this info. For whatever reason Grok, and X (probably via DOGE) has access to the 2257 database. Why it would need/want access to this? well…come on we know the reason.
Nothing to do with DOGE ripping every data base under the sun… Of course not.
sorry, you have credible reason to believe that X trained Grok using databases provided by DOGE?
You’re saying that Musk training his AI on doge data, and then using that training for Grok is unbelievable? Or do you want concrete evidence?
I mean Musk’s complete disregard for rules and regulations, combined with his own and his cronies’ incompetence, makes me think that, while no evidence exists, the scenario would definitely not be out of the realm of possibilities.
I’m simply asking if there any reason to believe this. You know, beyond, “Oh yeah, that’s totally something Musk would do?” As crimes go, it would be like writing your own name and address on a bank’s wall.
Apparently the answer is no, we have no evidence to support this hypothesis.
If you were an arrogant narcissistic maniac who knew for absolutely certain that you could rob that bank in full daylight, and NEVER be punished for it, of course you would write your name on the wall.
deleted by creator
Now you’re getting it. MAGA dragged us into a post-truth world, so now they have to live with it, too.
I understand people’s concerns about AI but it’s getting harder to take them seriously when half of their concerns are just made up speculation.
AI clearly needs stronger regulation, however all these people constantly spreading their just-so story criticism of it aren’t helping.
I’m not going to fabricate stories, but if the explanation for something is “AI was used irresponsibly” then I would tend to not dismiss the idea. And I’d be prone to propagate the idea far more. It’s not a conscious choice mind you, that statement is a result of self-reflection.
The way people unquestionably, irresponsibly, and without the slighest contemplation of the ramifications, are using AI for the most mundane tasks, worries me. Why do you need to make a photo of someone, who haven’t accepted the use case, into a drawing? Why do you need to have an AI transcribe our meeting and summarize the points? Nobody’s going to read the summary anyway, and now some unknown entity knows what other people said with an otherwise ordinary expectation of privacy.
One thing is that tech overlords have already scraped every image, sound bite, video clip, and written word publicly, and semi-publicly, available. Don’t give them more material.
The personal risk isn’t in talking to AI, it’s in listening to AI. And our environment is already becoming crowded with it. There is an argument that AI is disproportionately expensive and exploits externalities, but it’s hard to take that argument seriously from meat eating motorists.
Hey hey now, calm down, there’s no need to get personal here :-) We both dislike AI, can’t we just agree to hate AI and leave it at that? Like Legolas and Gimli in LotR? I’ll even let you toss me…
My main concern about AI is that it’s rotting our cognitive abilities, yours seem to be founded in an environmental stance. I can respect that, and I think it’s a valid argument.
I think both are bad, but can we agree to “AI bad. Don’t use AI”?
fashion, what can you do, it’s a power stronger than any of us
Yeah, both are owned by Musk. Go ahead and tell me that I’M the idiot for believing that he’d share his DOGE-stolen databases with his other enterprises.
I’m not saying that your argument isn’t plausible. But I have another route for this info.
Grok is training on tweets. Some person tweets “I just found out that my neighbor, Jane, is actually the pornstar Siri Dahl”. That tweet never gets traction, so only 5 followers reads it, but so does grok. Then in another tweet same person combines the firstname “Jane” with lastname “Doe”.
If your explanation is correct, as to where the info got leaked, I’d guess we’d see way more pornstars doxed.
Alternatively, it could be using image databases. Someone somewhere could have obscurely tweeted a photo of this performer off-duty with a note “here I am with my friend/relative/etc.” and included her legal name, and the image-recognition algorithms could have connected that with her performance images and went from there.
They want to locate the ones who are not of legal age.
To prosecute the people exploiting them, right?
…right?
You fucked that up. First panel should be locate not prosecute.
Edit: dammit you’re right
Here you go:
Of course, I mean to Trump a 17yo is at best a MILF. Besides she’d be used goods. Well maybe Epstein could have used the girl for less discerning perverts like Andrew Mountbatten-Windsor, but it would seem that both Einstein’s and Andrew’s days of raping kids is over…
Although one could hope that Andrew’s involvement in rapes aren’t over. Now that he’s in jail he has plenty of opportunities to experience spontaneous unwanted violent love making.
Sorry but whenever I hear this wording, I always imagine the porn lobby had a meeting one day to decide on a better word for ‘meat’ and they ended up on ‘talent’
all performers across most fields are referred to as ‘talent’.
It’s standard film & television terminology. Not everyone sees sex work as demeaning, you know.
Why not the more real, laborer?
probably because “talent” is more specific:
that would be like “why not call them dogs?” when talking about poodles…yes, it’s technically correct, but not really representative, is it?
Depends on what the circumstances are. But it’s more of a reminder that while some of the talent in the arts are super rich they are still laborers.
I think this term is used across all industries, Dua Lipa is talent, same as Brad Pitt
I always used actor, singer, …
Their way makes it seem like they’re the only ones there with talent lol