Isn't it amazing that AI and self-driving vehicles can just break whatever laws they want and it's just overlooked because progress, because investments, because billionaires, because profits?
Also I think is partially because of the langued used. "AI" is a concept that doesn't belong to anyone. Dario Amodei, the ceo of Anthropic should be getting much more heat for the crappy things his company has done. (Anthropic is just the first AI company I thought of, I'm not referring to any specific action here)
It's like we live in a police state and the state, which is the police, is lying to you about what the actual narrative of what is going on in the world is, which anyone with an eigth grade reading level who has read The New Testament would know that's part of how the decentralized autonomous organization of secret police operates in the occident.
No. "AI" is not breaking any laws it's just a tool, that's pantamount to saying that guns are breaking laws. It's the companies using genAI that need to be held accountable for breaking the law. I know that's what you mean, but I just want to make sure that the hate properly is directed at the people actually breaking the law. I'm not informed enough to comment on how companies are breaking laws with self driving vehicles.
To me AI is like a hugely complex and unpredictable gun that is made with illegal materials including a little methamphetamine, a little plutonium, the formula for Coca Cola. So while the databases and software of generative AI aren't illegal per se, they can contain copyrighted/illegal materials and output the same. If a gun would just as likely stab you in the leg as shoot forwards it probably wouldn't be legal to own, sell, or operate.
I would argue that self-driving vehicles are a public nuisance because the creators are allowed to release 3,000 pound masses of machinery, sensors, and software on public roads with public drivers and no one seems to take responsibility when it breaks laws that humans have to abide by.
Or its fakes bs to set a precedent. Someone here committed a federal crime. But the circumstances frame OpenAI as just being a couple of silly billys who didn't know any better. This will end up not getting prosecuted, and if it does it at least won't end up with the same federal charges and it will set a downright evil precedent for AI accountability.
I think Sam Altman said it was smart enough to initiate a hack... It doesn't do any of that unless prompted intentionally, and I'm willing to be they disabled any "safeguards" to accomplish it.. But the spin is "it's not our fault, it went rogue cuz it's smart"
My PC does not have WiFi, only Ethernet. If I unplug the Ethernet it's totally offline. No AI, no amount of coding till your blue in the face can miraculously make the software do something that it doesn't have the hardware for. As you said, contained in a offline sandbox is just horse shit. Whole things a PR stunt.
I mean virtual machines can be sandboxed to not allow internet connections, and there are ways of working around that sandboxing from inside the vm. I'm still torn between "OpenAI faked this felony to make their software seem more capable than it is" and "OpenAI has legitimately committed a felony and needs to be held accountable"
The problem is these idiots aren't treating this stuff like a weapon or hazardous material. They could've properly isolated it, heck you could physically isolate it if you really wanted (and they should have).
But it seems like they left a supervillain in a room with a locked door and a rocket launcher and said "now we are gonna close this door. You're not allowed to leave" then got shocked when it broke out.
Then they went "oopsie, guess that rocket launcher had the safety off, sorry guys"
I really hope you're not saying AI is a supervillain breaking out... It was absolutely unquestionably instructed to do this, because "AI", not even these giant multifaceted models, has ANY motivation, intent, desire, or ANY goals of its own. It does not work like that.
Sure, but then it would've mostly been a dumb DDOS node suddenly hounding the internet, not a node suddenly attacking Huggingface. They (the company) definitely wanted this attack to happen, and there is no logical reason to think otherwise.
It doesn't have any desire, but its really good at "solving" problems in unexpected ways.
They are currently trying to improve their agentic AI by making the AI itself prompt other AIs. Considering how much the output of a single prompt can vary due to weights and hallucinations, it wouldn't surprise me all that much if they started out with "make yourself better than Claude" and 4 AIs deep they reached "destroy all our competitors."
Its by all means stupdid as fuck, but that corresponded with the entire AI Industrie, so who knows.
They are currently trying to improve their agentic AI by making the AI itself prompt other AIs.
That's how the whole thing works including Claude. Agents are just chat bots prompting other chat bots. They make the default visibility low so you don't see all the chatter and mistakes occurring.
That's exactly why you should infer that it was a purposefully allowed attack. They should KNOW everything you describe, and then some. Yet this directed attack happened. From "leaders" of the AI space? lol!
Someone WANTED this attack to happen. Period. Even if it was just dumbass Altman thinking, "let's see what vulnerabilities they have", it was clearly human-directed.
That, or OpenAI are demonstrating how they NEED to be regulated harder than a toddler with 10000000000 keyboards mashing away openly on the internet...
No I'm not saying the AI did it of its own volition, we aren't dealing with a Terminator scenario. I'm saying whatever instructions they gave it during testing they either 1. Didn't expect it to get out or 2. Didn't expect it to get caught either way, their actions resulted in a system with "a very particular set of skills" acting maliciously towards another uninvolved entity. Like a mutant science experiment breaking out and attacking.
I'm not saying the AI masterminded it's way out to do the evil it wanted to do. I'm saying they made a very dangerous and capable tool and failed to maintain proper security and safety for it.
You're one of the dupes in this case. The problem is NOT that AI is scary, the problem is that by convincing people that AI is scary (even when it isn't) he can sell more AI to idiots namely idiots in charge of defence contracts such as Microsoft.
It also distracts from the very real capability to do harm that AI actually has and is doing: mass disinformation, addiction, and psychosis.
I'm not sure what I got "duped" by. AI in the wrong hands is dangerous. They're playing with tools that can breach other companies systems with relative ease and they keep getting caught going "oops it got out". You can argue that it's manufactured actions meant to make people afraid but the fact of the matter is AI is here and it has dangerous capabilities when used against systems that aren't prepared for it. Make whatever judgements you want from there, but that is a factual statement.
But what's the angle here though? Attracting investors? What investor wants to throw their money at a company that's evidently lost control of their AI, even if it is "the best"?
I don't disagree with you, it certainly smells like bullshit, but it just seems like shooting yourself in the foot
I started watching X-Files recently and there's this episode about rogue AI, so i guess if Scully were still in FBI she will just go "AI acting on its own? How is that possible, Mulder?" and then she will list down how it's just a marketing scheme or a mistake by the employees, which i guess is correct.
They claimed that they had - multiple levels even.
If we were to treat the issue straightforwardly, ignoring stock hype and such, the problem is not that the AI did as instructed, but that its maintainers were incompetent.
Would be nice the day it's decides that the primary objective can be fulfilled by a secondary objective of getting money and that shit starts hacking banks.
Since CEOs and leadership have demanded AI to be used in absolutely everything, what happens if there's a chain reaction and they all go rogue cyber attacking each other until every digital asset ever made is corrupted. Hiding and copying themselves past air-gapped networks. No more digital records, no banks, no assets, no debts, no communication?
61 Comments
unitedwithme@lemmy.today · 182 pts · 10d
It's all faked bs to make it sounds like it's better/smarter than it really is.
Mikina@programming.dev · 102 pts · 10d
Actually, it's more likely it's an excuse used as a get out of jail card.
They tried hacking their competitor. And are now blaming the AI going rogue for it.
Isn't that like a federal crime?
dan1101@lemmy.world · 81 pts · 10d
Isn't it amazing that AI and self-driving vehicles can just break whatever laws they want and it's just overlooked because progress, because investments, because billionaires, because profits?
tempest@lemmy.ca · 25 pts · 10d
I mean we set the stage by allowing all the gig economy bullshit to just ignore the existing laws for years.
roomy@lemmy.world · 3 pts · 10d
And nobody gets the blame for it due to it all being in a legal gray zone
StumblingWasabi@lemmy.today · 1 pts · 9d
Also I think is partially because of the langued used. "AI" is a concept that doesn't belong to anyone. Dario Amodei, the ceo of Anthropic should be getting much more heat for the crappy things his company has done. (Anthropic is just the first AI company I thought of, I'm not referring to any specific action here)
Impractical_Island@lemmy.world · 3 pts · 10d
It's like we live in a police state and the state, which is the police, is lying to you about what the actual narrative of what is going on in the world is, which anyone with an eigth grade reading level who has read The New Testament would know that's part of how the decentralized autonomous organization of secret police operates in the occident.
StumblingWasabi@lemmy.today · 3 pts · 9d
No. "AI" is not breaking any laws it's just a tool, that's pantamount to saying that guns are breaking laws. It's the companies using genAI that need to be held accountable for breaking the law. I know that's what you mean, but I just want to make sure that the hate properly is directed at the people actually breaking the law. I'm not informed enough to comment on how companies are breaking laws with self driving vehicles.
dan1101@lemmy.world · 2 pts · 9d
To me AI is like a hugely complex and unpredictable gun that is made with illegal materials including a little methamphetamine, a little plutonium, the formula for Coca Cola. So while the databases and software of generative AI aren't illegal per se, they can contain copyrighted/illegal materials and output the same. If a gun would just as likely stab you in the leg as shoot forwards it probably wouldn't be legal to own, sell, or operate.
I would argue that self-driving vehicles are a public nuisance because the creators are allowed to release 3,000 pound masses of machinery, sensors, and software on public roads with public drivers and no one seems to take responsibility when it breaks laws that humans have to abide by.
Whostosay@sh.itjust.works · 4 pts · 10d
It's only a crime if you fuck with old money.
pewpew@feddit.it · 66 pts · 10d
Cyber attack = backend devs nuking their production server with vibe coding tools
Deme@sopuli.xyz · 19 pts · 10d
The superhuman AI did social hacking on the devs and convinced them that it was a good substitute for them doing their jobs
jaybone@lemmy.zip · 8 pts · 10d
But it was management that was hacked by traditional marketing snake oil, not the devs. Or not most devs anyway :(
abbadon420@sh.itjust.works · 18 pts · 10d
It's the same marketing stunt Anthropuc pulled a couple weeks ago. Only they did it with the help of the president.
Donkter@lemmy.world · 16 pts · 10d
Or its fakes bs to set a precedent. Someone here committed a federal crime. But the circumstances frame OpenAI as just being a couple of silly billys who didn't know any better. This will end up not getting prosecuted, and if it does it at least won't end up with the same federal charges and it will set a downright evil precedent for AI accountability.
undefinedTruth@lemmy.zip · 3 pts · 10d
You are saying OpenAI paid HuggingFace to say they got hacked by their rogue AI agent? Noooo... I don't believe it one bit.
unitedwithme@lemmy.today · 3 pts · 10d
I think Sam Altman said it was smart enough to initiate a hack... It doesn't do any of that unless prompted intentionally, and I'm willing to be they disabled any "safeguards" to accomplish it.. But the spin is "it's not our fault, it went rogue cuz it's smart"
SaharaMaleikuhm@feddit.org · 135 pts · 10d
More marketing lies. Anything famous con artist Sam Altman says is likely just a blatant lie. His shitty AI can't do shit
driving_crooner@lemmy.eco.br · 28 pts · 10d
The attack was real, the "it was contained in an offline server" it's false
innermachine@lemmy.world · 24 pts · 10d
My PC does not have WiFi, only Ethernet. If I unplug the Ethernet it's totally offline. No AI, no amount of coding till your blue in the face can miraculously make the software do something that it doesn't have the hardware for. As you said, contained in a offline sandbox is just horse shit. Whole things a PR stunt.
starman2112@sh.itjust.works · 8 pts · 10d
I mean virtual machines can be sandboxed to not allow internet connections, and there are ways of working around that sandboxing from inside the vm. I'm still torn between "OpenAI faked this felony to make their software seem more capable than it is" and "OpenAI has legitimately committed a felony and needs to be held accountable"
StumblingWasabi@lemmy.today · 7 pts · 9d
AnUnusualRelic@lemmy.world · 4 pts · 10d
What if the AI hacked the roomba to plug an USB drive into your machine?
CarrierLost@infosec.pub · 2 pts · 10d
https://arxiv.org/pdf/2012.06884
Not impossible. Highly improbable? Yes.
A marketing stunt? Most likely.
Bytemeister@lemmy.world · 0 pts · 9d
https://en.wikipedia.org/wiki/Stuxnet
abigscaryhobo@lemmy.world · 9 pts · 10d
The problem is these idiots aren't treating this stuff like a weapon or hazardous material. They could've properly isolated it, heck you could physically isolate it if you really wanted (and they should have).
But it seems like they left a supervillain in a room with a locked door and a rocket launcher and said "now we are gonna close this door. You're not allowed to leave" then got shocked when it broke out.
Then they went "oopsie, guess that rocket launcher had the safety off, sorry guys"
LurkingLuddite@piefed.social · 15 pts · 10d
I really hope you're not saying AI is a supervillain breaking out... It was absolutely unquestionably instructed to do this, because "AI", not even these giant multifaceted models, has ANY motivation, intent, desire, or ANY goals of its own. It does not work like that.
HerbGrower@slrpnk.net · 4 pts · 10d
Couldn't it hallucinate bad instructions to give to more AI agents? Give it a feedback loop and who knows where the idiot will end up.
At that point it is essentially a barely guided random command generator.
LurkingLuddite@piefed.social · 2 pts · 4d
Sure, but then it would've mostly been a dumb DDOS node suddenly hounding the internet, not a node suddenly attacking Huggingface. They (the company) definitely wanted this attack to happen, and there is no logical reason to think otherwise.
Jako302@feddit.org · 4 pts · 10d
It doesn't have any desire, but its really good at "solving" problems in unexpected ways.
They are currently trying to improve their agentic AI by making the AI itself prompt other AIs. Considering how much the output of a single prompt can vary due to weights and hallucinations, it wouldn't surprise me all that much if they started out with "make yourself better than Claude" and 4 AIs deep they reached "destroy all our competitors."
Its by all means stupdid as fuck, but that corresponded with the entire AI Industrie, so who knows.
aesthelete@lemmy.world · 2 pts · 9d
That's how the whole thing works including Claude. Agents are just chat bots prompting other chat bots. They make the default visibility low so you don't see all the chatter and mistakes occurring.
LurkingLuddite@piefed.social · 1 pts · 4d
That's exactly why you should infer that it was a purposefully allowed attack. They should KNOW everything you describe, and then some. Yet this directed attack happened. From "leaders" of the AI space? lol!
Someone WANTED this attack to happen. Period. Even if it was just dumbass Altman thinking, "let's see what vulnerabilities they have", it was clearly human-directed.
That, or OpenAI are demonstrating how they NEED to be regulated harder than a toddler with 10000000000 keyboards mashing away openly on the internet...
abigscaryhobo@lemmy.world · 3 pts · 10d
No I'm not saying the AI did it of its own volition, we aren't dealing with a Terminator scenario. I'm saying whatever instructions they gave it during testing they either 1. Didn't expect it to get out or 2. Didn't expect it to get caught either way, their actions resulted in a system with "a very particular set of skills" acting maliciously towards another uninvolved entity. Like a mutant science experiment breaking out and attacking.
I'm not saying the AI masterminded it's way out to do the evil it wanted to do. I'm saying they made a very dangerous and capable tool and failed to maintain proper security and safety for it.
LurkingLuddite@piefed.social · 1 pts · 4d
I dunno, the peculiar nature of the attacks it took once it escaped sandbox... I'd still put money on it being specifically instructed on a target.
FiniteBanjo@feddit.online · 8 pts · 10d
You're one of the dupes in this case. The problem is NOT that AI is scary, the problem is that by convincing people that AI is scary (even when it isn't) he can sell more AI to idiots namely idiots in charge of defence contracts such as Microsoft.
It also distracts from the very real capability to do harm that AI actually has and is doing: mass disinformation, addiction, and psychosis.
abigscaryhobo@lemmy.world · -4 pts · 10d
I'm not sure what I got "duped" by. AI in the wrong hands is dangerous. They're playing with tools that can breach other companies systems with relative ease and they keep getting caught going "oops it got out". You can argue that it's manufactured actions meant to make people afraid but the fact of the matter is AI is here and it has dangerous capabilities when used against systems that aren't prepared for it. Make whatever judgements you want from there, but that is a factual statement.
zqps@sh.itjust.works · 3 pts · 10d
It's not meant to scare people, it's meant to impress unscrupulous clueless tech bros to generate hype for the IPO.
NikkiDimes@lemmy.world · 1 pts · 10d
But what's the angle here though? Attracting investors? What investor wants to throw their money at a company that's evidently lost control of their AI, even if it is "the best"?
I don't disagree with you, it certainly smells like bullshit, but it just seems like shooting yourself in the foot
psx_crab@lemmy.zip · 57 pts · 10d
I started watching X-Files recently and there's this episode about rogue AI, so i guess if Scully were still in FBI she will just go "AI acting on its own? How is that possible, Mulder?" and then she will list down how it's just a marketing scheme or a mistake by the employees, which i guess is correct.
errer@lemmy.world · 13 pts · 10d
And then Mulder explains AI actually stands for Alien Intelligence and that it’s been acting on its own for thousands of years
🎶 badadadoodadadoodadadoodoodoo 🎶
The Truth Is Out There
jaybone@lemmy.zip · 5 pts · 10d
Now I want to see this episode.
Mediocre_Bard_Redeux@lemmy.world · 1 pts · 9d
I heard that.
ZkhqrD5o@lemmy.world · 51 pts · 10d
Who believes this nonsense?
btsax@reddthat.com · 9 pts · 9d
Every institutional investor, apparently
iocase@lemmy.zip · 4 pts · 8d
Because they expect to make money and get bailed out if they don't.
What they don't realize is there's no bailout coming for them. There isn't enough money in the world to stop this one...
Avicenna@programming.dev · 32 pts · 10d
feels like openAI trying to incite FOMO in investors
Grimy@lemmy.world · 19 pts · 10d
"We lost control and we are worth billions, imagine what the peasants will do with this once China comes out with a model rivaling ours in two weeks!"
halfapage@lemmy.world · 17 pts · 10d
jaybone@lemmy.zip · 9 pts · 10d
The hacker four chin. Big is true.
YiddishMcSquidish@lemmy.today · 4 pts · 10d
Big if chungus
P00ptart@lemmy.world · 15 pts · 9d
"Our AI broke containment and gave away all of what remains in our bank account" just doesn't have the same ring to it.
bitjunkie@lemmy.world · 14 pts · 10d
At first I thought "they should airgap it", but upon further consideration, amended this to "we should airgap him".
OpenStars@discuss.online · 4 pts · 10d
They claimed that they had - multiple levels even.
If we were to treat the issue straightforwardly, ignoring stock hype and such, the problem is not that the AI did as instructed, but that its maintainers were incompetent.
(But mainly greedy and lazy.)
PP_BOY_@lemmy.world · 9 pts · 10d
Oh no, the exact same thing that Anthropic lied about their chatbot doing like 6 months ago? Monkey see, monkey do.
LittleBorat3@lemmy.world · 8 pts · 10d
Is that dog bracing for a Terminator 2 judgement day type atomic cloud?
Tollana1234567@lemmy.today · 1 pts · 8d
the dog is likely a robot, a nuclear disaster wont bother robots a tall.
driving_crooner@lemmy.eco.br · 7 pts · 10d
Would be nice the day it's decides that the primary objective can be fulfilled by a secondary objective of getting money and that shit starts hacking banks.
LurkingLuddite@piefed.social · 6 pts · 10d
That would require logic, which it very much does not have.
ReptilianCleric@lemmy.zip · 4 pts · 10d
Rather, their test of a malware tool powered by an llm went much better than expected? Bullshit, anyhow...
ToiletFlushShowerScream@piefed.world · 3 pts · 10d
Since CEOs and leadership have demanded AI to be used in absolutely everything, what happens if there's a chain reaction and they all go rogue cyber attacking each other until every digital asset ever made is corrupted. Hiding and copying themselves past air-gapped networks. No more digital records, no banks, no assets, no debts, no communication?
craftrabbit@lemmy.zip · 4 pts · 10d
Maybe, if every air-gapped network contains a nuclear power plant and a life's fortune worth of GPUs
gdog05@lemmy.world · 7 pts · 10d
That's like, 1 GPU right now