Rendered at 11:29:31 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
bayarearefugee 1 days ago [-]
Shouldn't some of these AI company higher ups be arrested and brought up on Computer Fraud and Abuse charges?
Any individual running a local LLM that hacked a company would absolutely be held responsible for the LLM's actions, and corporations are, as Mitt Romney likes to say, people.
jpfromlondon 23 hours ago [-]
it's fiction, that's why.
dormento 17 hours ago [-]
Seems like nowadays the mantra is:
- "Hey dev, why did you write bugged code?"
- "It wasn't me, it was my LLM"
- "Hey ops guy, why did you deploy and insecure env?"
- "It wasn't me, it was my LLM"
- "Hey dude why did you hack this firm?"
- "It wasn't me, it was my LLM"
ad infinitum.
ActionHank 22 hours ago [-]
"I have a girlfriend, but she goes to another school so you won't know her, but she's real!"
cyanydeez 21 hours ago [-]
which part, CEOs being arrested for illegal activities?
inigyou 17 hours ago [-]
The hacking is fiction. Every AI company is claiming this happened to get a stock boost.
iAMkenough 18 hours ago [-]
If there isn’t a legal resolution, that’ll only leave us with a Luigi resolution.
scotty79 20 hours ago [-]
> Any individual running a local LLM that hacked a company would absolutely be held responsible for the LLM's actions
What if he could present evidence that he didn't tell AI to do it? Is it reckless negligence to give any AI agent access to the internet or even not prevent it 100% from gaining it? Because millions are doing it daily. Pretty much everybody who's doing anything with AI agents.
nDRDY 1 days ago [-]
This reality is wild. It would be like arms manufacturers holding weekly demonstrations of how much collateral damage their latest designs cause in an attempt to increase sales.
elias_t 1 days ago [-]
It’s even worse, because they would have to say: “the bomb dropped itself because we forgot a screw in the airplane, but see how effective it is!”
twister2920 16 hours ago [-]
apropos of nothing I highly recommend "The Palestine Laboratory" by Anthony Loewenstein
specproc 22 hours ago [-]
They do this in Gaza.
inigyou 17 hours ago [-]
You can't say that here.
halJordan 21 hours ago [-]
Everyone loves the opening scene in iron man.
elias_t 1 days ago [-]
That’s crazy that such errors in isolating the models became a marketing stunt.
I’m afraid of what it will be next. They will remove guardrails just to provoke more and more headlines
felooboolooomba 1 days ago [-]
> I’m afraid of what it will be next.
Palantir: our system accidentally started WW3.
pier25 23 hours ago [-]
Shall we play a game of thermonuclear war?
felooboolooomba 21 hours ago [-]
Wouldn't you prefer a good game of chess?
expedited123 1 days ago [-]
This is hilarious. How obvious can they be ? Suddenly every LLM is a hacking genius
nmstoker 1 days ago [-]
Interesting that none of the talk around this has touched on any form of enforcement, even though this is illegal activity in US/EU.
Any fine would be peanuts for them but it might help to refocus the mindset from "oops, misconfiguration" to "test security but do so with care, consent and in line with the law"
itopaloglu83 1 days ago [-]
It feels like Meta would like some of that free advertisement.
Ekaros 21 hours ago [-]
I can see one company being incompetent. But so many? Should it not be pretty trivial to to track whatever traffic goes out? All this traffic must originate from somewhere or have I massively misunderstood how these things work?
moktonar 22 hours ago [-]
Waiting for the “Our model inadvertently hacked the nukes and launched Armageddon” title
cyanydeez 1 days ago [-]
"Metoo guys, im also doing bad stuff"
noir_lord 1 days ago [-]
It's Meta, I'm not sure anyone paying attention could ever assume they aren't doing bad stuff.
ernsheong 22 hours ago [-]
I see a pattern...
helsinkiandrew 23 hours ago [-]
I'm sure Elon Musk has told some underling to make sure that Grok attempts to hack into a few companies so they get the street cred to.
vcryan 1 days ago [-]
Meta never wants to be left out.
iammjm 1 days ago [-]
It’s like Trump bragging he could shoot someone at Fifth Avenue and not lose voters. Congrats, I guess??
FuturisticLover 22 hours ago [-]
Did he actually say that?
beering 20 hours ago [-]
Yes, it was a joke he made about how loyal his voters are, and it was widely covered in the news.
inigyou 17 hours ago [-]
And then he actually did kill some people and didn't lose any voters. Also he was exposed as a child raper and didn't lose any voters.
vrighter 20 hours ago [-]
oh.... now all of them are accidentally doing it
927373912092 17 hours ago [-]
[flagged]
specproc 16 hours ago [-]
The eight-hour-old accounts with strings of digits as handles are also well-known for their balanced and non-partisan perspectives.
Any individual running a local LLM that hacked a company would absolutely be held responsible for the LLM's actions, and corporations are, as Mitt Romney likes to say, people.
- "Hey dev, why did you write bugged code?"
- "It wasn't me, it was my LLM"
- "Hey ops guy, why did you deploy and insecure env?"
- "It wasn't me, it was my LLM"
- "Hey dude why did you hack this firm?"
- "It wasn't me, it was my LLM"
ad infinitum.
What if he could present evidence that he didn't tell AI to do it? Is it reckless negligence to give any AI agent access to the internet or even not prevent it 100% from gaining it? Because millions are doing it daily. Pretty much everybody who's doing anything with AI agents.
I’m afraid of what it will be next. They will remove guardrails just to provoke more and more headlines
Palantir: our system accidentally started WW3.
Any fine would be peanuts for them but it might help to refocus the mindset from "oops, misconfiguration" to "test security but do so with care, consent and in line with the law"