NO use of AI is good.
These companies advertising methods are becoming a bit too much!
Cool. Now target the student loan borrower databases.
That’s how you get doorkickers instead of a lawsuit.
I think it was a joke, but the student loan borrower databases are backed up in salt mines.
Time to bring a garden hose
It’s always funny to see the priorities. People were asking how the bunker owning billionaires were expected the AI/climate apocalypse to go for them and we had to point out that while your wealth, data, and all might be screwed up and lost, they have records of their ownership and your debt that will survive a nuclear war.
LLMs are hacking tools actively used to commit cyber crimes. They should be shut down and deleted now.
And then only the criminals would have them. You know because criminals tend to break the law so making a tool that they used to break the law illegal wouldn’t really impaire them.
There’s a lot of people in here who don’t know about open weight models that can run on an old 7900XTX.
The data center part of LLMs is how they’re going to try to make money off of LLMs. The technology to use it for nefarious uses is already loose in the world.
still built on theft. stop acting like running your own model means it wasn’t built on top of giant fucking theft.
I mean yeah, you’re right. Guess you missed the point…
No, you’re simply oblivious to the fact that it’s all built on theft. someone has to compute the inference, requiring tremendous amounts of resources, built on theft. you’re not going to keep the llm and destroy the data center, even if you steal the fucking model someone had to compute it.
but cute distractions kiddo
Your reading comprehension is lacking
right up there with your critical thinking skills.
lemmy users will miss any point you make, no matter how obvious, if it lets them jockey for attention.
it’s fascinating that you’re ignoring the premise while validating their ignorance.
childish, but fascinating.
Tripling down on that whoosh.
Thats a level of dedication you don’t see often these days.
Of course every criminal has easy access to large data centers, and stuff. Except for the federal three letter criminals, of course.
Criminals don’t own datacenters running the models that actually pose a threat.
No does anybody else, but apparently they have access to them well we wouldn’t even be talking about it.
Mid-high gaming pc money will have you running near-frontier level models locally.
train tho?
yeah it’s always “we can run them on our gpus!”
yeah uh who trained it? how many gigawatts and thousand gallons of water did it take for inference?
near frontier levels of stolen shit, which is fine, it’s all stolen, but holy hell don’t act like it didn’t require idiotic amounts of precompiling and a moral ambivalence to the stolen corpus.
Or maybe they’re too stupid to know lol
this would be amazing… if you could convince people to use it lol.
as is, it’s a retort without meaning, it doesn’t change the situation if nobody uses it.
mhm
Mathematics is a tool used to compute missile trajectories and weapon yields. Math should be shut down and deleted now.
LOL
Guns are a tool used to kill people, guns should be shut down and deleted now.
That was my first thought. I think the difference is that AI isn’t a literal killing machine. AI is more like a hammer, it can be used that way. A gun is a killing machine.
Technically, math helps us figure out the diameter and barrel length as well as the amount of accelerant to create adequate pressure to fire a bullet to accurately kill things, so in the end it’s all math’s fault.
Knives are a tool used to kill people, knives should be shut down and deleted now.
Uh oh, you can’t point out mainstream cognitive dissonance you’ll get a down vote spanking.
The idealist in me says “fuck yeah”
The pragmatist in me says “fuck, we’re gonna have to do more mandatory cybersecurity training next week”
Especially because it’s the same damn people every time that caused the security vulnerabilities in the first place. If Karen could just stop clicking on obviously spam emails we’d be fine.
Sounds like Karen doesn’t get email priveleges anymore, on account of being a constant active threat vector.
Can’t do your job without email, you say? How tragic. If you learned how to use it properly the last seven times we had this discussion and referred you to mandatory phishing safety classes, this wouldn’t be a problem.
I just flag every email from IT and HR as a phish
Thank you for your service, soldier
We’re just going to be sitting the middle of supply chain hacks until the cows come home. All that porting bullshit to rust might end up being the final answer.
As he used AI, surely no actions can be brought against him right? What’s good for the goose is good for the gander and all.
If he said his intent is to hack then yes, he can be charged. But if he say “it’s rogue AI” then it’s all clear.
It was just a test. I didn’t know it had internet access, I just connected it up to the internet how was I supposed to know?
Essentially OpenAI’s defense every time one of their AI does something that was instructed to do, but turned out to be illegal.
Even better if you put it in a default sand box environment from 2 years ago and tell it not to get out of the sandbox but then also give it every hack to get out of the sandbox.
Even modern sandbox environments have access to curl and python and who’s going to read EVERY permission prompt anyway? Just set it to auto which lets another AI model check for safety or even better, YOLO mode where everything is allowed.
I don’t understand people who just let the AI loose on their system and don’t have any safeguards in place. At the very minimum it shouldn’t have admin access and should be limited to just the project folders. Also back up the project folders to another location the AI doesn’t have access to.
Really basic digital hygiene is all that’s needed, and yet there are constantly stories about idiots having their entire hard drives wiped.
“sandbox” with no wood walls or sand
Just a few “Lincoln logs”
“I was just using AI to help me come up with dystopian cyberpunk scenarios. Whatever it chose to do with the prompt is on the AI.”
Of course not. The system is rigged. Also, his actions clearly show motivation to generate damages. So, in a sense, they’ll be a scapegoat. His case will be used as an example and allghe weight of our punitive society will fall upon him :(
Yup, and a justification to remove access to local models because “this technology is too dangerous for the population to have access to it”.
Hacker man strikes again, sitting in his dark room with his hood up just so we know that he is definitely a computer hacker.
I remember back when was just recording movies with his phone.
mr fancypants has a big enough memory card for that
I get why the guy on the left has a death stare; the guy on the right looks like he wants to kill.
Man, I remember those really shit recordings from a shaky phone of movies…they really really sucked.
While current cam videos are a lot worse than the clean data ones once it comes out on streaming, they are a lot better than they used to be. Do people set up a tripod or something now? And of course they have to wear a grey hoodie, it’s required.
Obviously the grey hoodie is mandatory. I’ve not really been interested in any new releases for the last decade or so. Or not interested enough to download a theatre recording anyway. 25 years ago, yeah, grabbed them all the time.
I will have you know that this is precisely my attire when I am programming computers.
My office gets chilly often. I may need to get a fluffy hoodie too.
Pikachu onesie helps you commune with the electrons, it’s well-known.
Has anyone done a fanfic of hackerman and hatman falling in love? Like hatman shows up in the dark room and instead of being scared, they kiss.
Be the change you want to see in the world
deleted by creator
Only the dim glow of the screen illuminated Hackerman, leaving his face in a shrouded mystery under his hood.
From the corner of Hackerman’s room, a figure crept out of the void. A nightmare silhouette crowned with a trillby hat. Hatman stood behind Hackerman, his cold strong hands squeezed Hackman’s shoulders.
The frenzy of typing ceased. Hackman spun around in his Respawn Pro chair. Their faceless expressions hide their desires.
“Are you not scared stiff?” Hatman whispered.
“Oh I’m stiff but not scared.” Hackman replied.
Hatman’s eyes moved down to Hackerman’s lap to see a tower of lust erected in his honor. He was shocked by the size.
“People only see me from the desk up, I have no need for pants.” Hackman said.
In a moment of passion Hatman spun around and dropped his shadow pants to expose his shadow love hole.
Looking over his shoulder Hatman said “No one can move because of all the benadryl and I just want someone to move me! Take me Hackman!”
Hackman stood and placed his nibble fingers around Hatman’s slender waist. He took one hand and lowered Hatman down with such force, his hat fell off. Than he thrusted deep.
“I’m in!” moaned Hackerman.
You did it you crazy sonovabitch, you did it
10/10 no notes
I am always curious how he does not see all those lit up numbers, letters or icons floating around his head.
computer thugs
deleted by creator
“It’s not script kiddie stuff.”
No, it’s slop kiddie stuff.
If it works it works
The distinction is in the understanding of what you are doing. “Script kiddie” is considered derogatory because running scripts written by someone else does not require understanding of how those scripts achieve what they do. It is a distinction between hacking a system and turning on a machine and standing nearby while it hacks the system.
Given that definition and distinction, using an AI model to do this is merely changing where the scripts came from. It may work, but the person doing it is still a “script kiddie” by the common definition.
I suspect that what was meant was that the attacks themselves were far more sophisticated than what a skiddie could do (after all, if you just run scripts, you probably don’t really learn anything from the information you get back), not so much that the person ordering the attacks was anything more than a script kiddie. Of course, they didn’t really make it clear what exactly they meant by it.
Where are all the details? Shocking journalism
Looks like this may be the original source. It’s about two articles deep from OP.
You the real MVP
Look, we know they’re a Chinese-speaking hacker, what else do you need to know? /s
China bad.
Hacker bad.
It’s modern journalism, only the headline really matters… only a small percentage of people are going to read the article anyway.
It says AI right in the title, that’s going to bring all the attention and interaction on social media.
The details in the article infer a few weak innefectual attacks. So epic.
This was inevitable
Hurray
I am conflicted
I said a few weeks back it was just a matter of time before someone did this. I hope they fuck with big companies, the rich, and governments that deserve it.
deleted by creator
ebic
