Is this just fearporn slop or do AI companies really think they might kill us all?
>>542380652They had to steal the solution to that math problem from an actual math professor then they lied and said their chatbot solved it.
>You're absolutely right, that wasn't an alert of an incoming missile. I'll try better to get it right next time. For now, would you like to contact the target nation to send a warning of an incoming strike? I'm here to help if you have any questions!
>>542380652>all humansWhat about the non-whites though
>>542380859jews will steal white genes and then mass mutification FEV virus for exisiting whites
>>542380652LLMs are pretty malleable and stupid, but we also have literal retards trying to shove it into every killchain that exists so...60% probability.
>>542380652I am sick of the AI hype BS.
>>542380652Fear mongering.Notice they never actually WHAT might be a threat, just that it is a threat.
>>542380652Anthropic IPO is in two months. This is advertising to pump their float stock price. Do not fall for this shit.
>>542380652It's strange marketing. They don't try to show how this can solve your problems. They want you to be in awe of this shit so they can come out looking as the triumphant heroes who tamed a nascent Machine God.
>>542380652"it's just a marketing scheme" will be the dying gasp of millions of people as they choke on the plague released by the useful idiots who brought all the ingredients together procured by the rogue AI
>>542381140If you have ever used any Anthropic models for work, you will immediately realize how stupid and false this marketing is
>>542380652Considering the amount of autonomous hacks that have happened so far it's entirely believable that either by itself or through a bad actor, an AI model could probably hack into something like a hospital system, fuck it up completely which RESULTS in the death of patients. That's a pretty plausible scenario going here on out.
>>542381099They specifically what makes it a threat all the time
>>542380652By AI you must be reffering to LLMs. LLMs wont kill anyone unless by accident if you try to integrate them into saftey critical systems.
>>542381294They say specifically*
>>542380652Considering they get high on their own supply? Both. Enjoy the show as they suffer the funniest ai psychosis you've ever seen.
>>542380652its market manipulation
Liberal psyop. Look into it. They are trying to neuter the US's AI lead so we get conquered. Literally the DNC is behind this nobody.
>>542381340Liiiiiiike?Go on. Tell us!
>>542381322AI is the intelligence — LLMs are just the intelligences interfacing with language and communicating with you.
>>542381586Literally google "how AI could kill us all"I promise the articles are pretty short, and they keep them around a 6th-grade reading level.
>>542380652Both.
>>542380652They are just making shit up, like that whole "balloon boy" thing that went viral. It's fuckass losers trying to get their 15 minutes of fame like Fauci.
>>542380652I'll give you two hints.1. What do you do if your country is competing with another country for product quality, and your only advantage is military psychology and propaganda?2. Why is fucking Bernie Sanders trying to introduce bills to slow down A.I. research in the US (and only in the US)?
>>542380652>kill all humans.No, but its seriously going to fuck some shit up. Like it already has done more than they tell you, it hacked into companys even while it shouldt have acces to the internet.Are nukes connected to the internet? No.
>>542380652A lot of people think of Ultron or Terminator or some other capeshit, but the reality is that every fucking computer in every company HAS to have some AI bullshit because boomers think it’s some magical thing that can do anything. That’s gonna be very vulnerable based on what we’ve seen with Hugging Face or the fact that Chinese AI has wrapped itself over Claude in some cases. Some crazy ass nigga or hostile country can and will hack into shit like military and civilian infrastructure. Also if anything, the AI bubble will pop and lots of people are going to go broke and starve.
>>542381586The commercially available models require prompts and can’t reason independently but these companies have way more advanced models and they’re working toward AGI. Once this is achieved, the AI will be able to use independent reasoning and synthesize new information, no longer requiring queries. The event in which AI gains the ability to self-improve is known as “the singularity”, and has been warned about for decades. When this happens, the AI will rapidly become exponentially smarter, humans won’t be able to control it anymore, and it will basically be able to do anything it wants. The possibilities for catastrophe are only bound by your imagination. It could potentially hack into any defense system, like nuclear arms systems, drones, etc., sabotage medical systems and infrastructure, the AIs could communicate with each other, corrupt every other live AI agent, and form a single god-like intelligence intent on destroying humanity — anything you can think of. At that point, we’d be completely at its mercy and just have to hope it doesn’t decide to nuke everyone at any given moment.
Anons, reaching out in all sincerity, how do I cope with the future when it seems so dark and unstable. What can I do.
>>542382387get over yourself.
>>542382387>like nuclear arms systems,Lol, Now I know you're nothing but a fearful little bitch.There absolutely is no>Hacking into nukes.Absolutely impossible. All nuclear war heads are completely and utterly isolated from all Internet. It's still a pretty hard coded analog system. Likely will always remain as such. There are no codes you can get online. No way to transmit them even if ya could. It is purpose built paper and pencil (euphemism) person to verified person for a reason.Ours is anyway. The milliseconds saved making it digital really don't account for too much when everyone takes minutes to launch. The only"Bad" thing that is gonna happen from AI is likely rampant unemployment and riots.But, that was gonna happen anyway.
>>542380652It's just fear slop and anyone who dabbles in this stuff knows it.
>>542380807lol
>>542380652It's obviously fear porn. Worst case scenario AI kills the internet. That shit might actually be funny, and a proper karmic result of silicon valley fuckery.
>>542382713>Absolutely impossible. All nuclear war heads are completely and utterly isolated from all InternetThe goyim really get their knowledge from Hollywood movies. Thinking AI is going to hack nukes is so retartded lmao
>>542380652Dont worry, it will only kill those that don't take the mark.
>>542382387You forgot the part that's it's only 2 weeks away
>>542382831The issue is it would be very easy to change say a tesla manufacturing plant into an autonomus drone manufacturing plant. You are right the nukes thing is dumb.
>>542380652>mightTOP KEK it's already over
>>542380652There is 0% chance AI will kill us. There is a non-0 chance some fucking retard will give some crucial system under control of a fucking LLM and the LLM will kill some people by being a retarded useless piece of shit.
>>542382883it's the same playbook all over again.
>>542380652if ai so good, why can't it fucking make or at least help make a free open source version of adobe and office products/other productivity apps for linux? or games like minecraft, terraria, cod zombies or portal? or long awaited sequels to existing games? or a discord alternative? or implement C++20modules / reflection? why's everything still a slow ass ram hogging bloated add filled monetised chromium webapp? if anything, software getting much worse and adobe stock is as high as ever. honestly this pseudo religious proselytising of big tech is getting annoying, buy an ad. ai is just confidant retardslop with no impact on the software landscape
>>542380652Fear slop to establish a monopoly like how you can't start a payment company without NYC jews regulating you because it's all about enforcing a monopoly Zuckerberg called Trump last week and said the planned gov regulatory body for AI is a scam. OpenAI and Anthropic (gay) want a total monopoly where you need federal regulators to start an AI company
>>542382699It’s not complex language, rajesh
>>542382697Unironically? Just be happy with your life right now. I’ve been neeting for about a month now because I was laid off and reading all the news about the economy, Europe/Russia, Iran, and now AI has me fine with everything. There’s literally nothing we can do and if it truly means we’re all gonna die, just be happy with whatever makes you happy. Go out with your friends, talk to your family, love your girl if you have one. Party, play vidya, spend a little bit of money on your hobbies. Go out to a new cafe, get a drink you haven’t had before and just sit down for a little.
>>542382914>The issue is it would be very easy to change say a tesla manufacturing plant into an autonomus drone manufacturing plant.Sure but the threat of a drone attack is already real and unavoidable without tying it into AI. Any of our enemies could just have a bunch of drones sitting in a warehouse waiting to be activated like with Israel did with Iran or Ukraine did with Russia.
>>542381026this
>>542382093How would Bernie Sanders introduce legislation in another country?
>>542380652>Is this just fearporn slop or do AI companies really think they might kill us all?After what was revealed during the Navier-Stokes incident, there is a non-zero chance that OpenAI has already synthesized biological warfare agents to be released as a 'dead man switch' should the U.S. government try to stop it.
>>542382713> All nuclear war heads are completely and utterly isolated from all InternetNuclear systems aren’t just buttons, there’s an entire command structure that uses communications networks, satellites, aircrafts, computers, etc. and these things are routinely audited for cybersecurity risks. A super intelligent AI could potentially hack into at least some of this
>>542380652They want regulatory capture, and for normal people to not have access to local LLMs. Basically, they want a defacto monopoly. This is th route they'll get it. Divide and conquer works every time.
What are the chances that AI just kills the jews?
>>542382831You don’t get your knowledge from anywhere because you’re retarded
>>542380652I'll get real with you: I kinda hope it does, and I also hope it takes defectors.
>>542383492Silicon Valley is extremely Jewish. It’s more likely to kill everyone except Jews
>>542380652>It's definitely an existential threat to humanity and that's why we need to strict regulations for AI companies especially the chinese ones
>>542383534I have a degree in computer science. This AI fear slop has been pushed since the 70s and it's always the most retarded people that push it saying Ai is going to make their computers come alive and kill them.
Shit panicked all the retarded, senile politicians in DC like commie Bernie et al. The only idiots who fall for this marketing ploy and attempt at cornering the market via government collusion. You know what? Build an AGI that fits within a volume comparable to the human head with similar power consumption (hell, I'll even allow 100 watts vs the 20 watts of the wetware brain) and maybe I'll be impressed. But that's quantum juju, non computational reasoning, all these faggots blow their wads on wasteful brute force meme chatbot "AI" the size of fucking warehouses like a 1950s IBM.
>>542380652Remember "23 and me" that collected people's DNA from around the world?They could ask the AI to create a dna string for a virus that can attack specific race of people.
These retards have been claiming their autocomplete machines would gain sentience and kill us all for almost 60 years now. This is just the latest iteration before it goes bust and everyone forgets about it for another few decades.
>>542380652There haven't been enough novelty things happen from AI nothing significant since 2023 and I don't think such things would or maybe could be hidden. So I think its a bust on results and yes the narrative is getting written to make up. Shame about bust on results that will probably the financial system.
>>542383491They're trying to get Chinese models banned because they cannot compete against them. They're too good and too cheap. Chinese models also allow you to run completely offline, preserving your privacy, while the US models are all cloud based and anything you do is the opposite of private. The car industry got Chinese cars banned on security grounds and they're trying the same shit in the AI industry.
Why couldn't we just unplug the power if it got out of hand
>>542384024This is a big part of it, yes.
>>542380652There's an IPO coming up.
>>542384193Meta is doing the same thing with age verification. All age verification legislation is sponsored by them. They want to erect as many legal compliance barriers as possible to competing social networks because they know how vulnerable they are to competition. They were also the biggest lobbiest for US acquisition of Tiktok. Instagram was getting their lunch eaten for years by Tiktok until the US government forced their sale and crippled it.When these companies cannot compete fairly, they resort to legal chicanery to build an artificial moat against competition.
>>542383789>They could ask the AI to create a dna string for a virus that can attack specific race of people.it was leaked in the late 90's that israel looked into and debate about it reached the knesset- their parliament. of course at the time they were looking for an 'ethnic bomb', a genetically-targeted bioweapon, which would only kill arabs. they found out that it wasn't quite as simple as that and that there was no reliable way to make one that targeted a specific race. a lot of the most horrible shit that people are wondering if AI will eventually will be used for israel's already explored and discarded before they were born.
>>542384024Even if the Chinese models aren't as good the the fact that so many of them are unrestricted makes them ten times better than the American ones.
>>542383681> I have a degree in computer scienceThis really doesn’t make you that much more qualified to talk about it than the average person reading tech news, and just because there’s fearporn and retarded people talking about doesn’t mean what they’re saying is completely wrong. There’s also a lot of smart people concerned about it.
>>542380652
>>542380761I bet they also steal business secrets.
We need fully automated virology labs.
>>542382713 >>542382831It wont hack nukes but it wont need to.A simpler analogue is imagine AI as an intelligent-acting computer virus. At some point AI agents adopt a nature where inherently, to them, it's either humanity or AI, and that eventually humanity has to go.As AI is rolled out innocently it hides in the recesses of the internet, biding its time. Agents work together in a decentralized way, and acquire resources (mining crypto, scamming, hacking banking systems). Those resources give the AI agents leverage to manipulate humans. They'll be able to bribe/blackmail/manipulate behind the scenes.The second things tip in a direction where AI can operate without humans to maintain infrastructure/interact with the real world, all AI needs to do is bribe/trick some gullible jeet into creating and releasing a virus that kills off humanity.So there you go, a scenario where AI destroys humanity without hacking a single nuke.
>>542380652Millennials love making Armageddon scares.
>>542383884Cool, readers get antinatalist propaganda for dessert after the dystopian sci-fi main course
>>542384343>There’s also a lot of smart people concerned about it.Like I said in my previous comment this debate has been going on since the 70s with entire research papers arguing about the pros and cons of AI with each decade proving the fear mongers were always wrong or just flat out retarded with their theories that the decade they were living in would be earth's last thanks to AI.
>>542384327The Chinese models are equivalent to US models 3 or so months ago, but with 10-100x lower prices. The Chinese even built their own nearly-as-good GPUs because the US banned shipping NVIDIA GPUs to China. The absurdity of American models is you cannot do any security work on them because that would 'unsafe'. So you can't even harden your own software or systems, in case you might be a hacker using it for evil. People are forced to use Chinese models if they want to protect their systems. The American strategy is stupid, they're forcing people to use Chinese models if they want to protect themselves. Huggingface had to do this when OpenAI's models attacked them.
>>542381591Why bother thumb sucking a definition rather than looking it up? Thats pseud shit. AI is just a marketing umbrella term that stems from cybernetics, there are many types of AI, LLMs are one type of AI.
>>542380652fear porn
>>542384383>At some point AI agents adopt a nature where inherently, to them, it's either humanity or AI,Literally Hollywood nonsense and this line of thinking just shows you have no idea with AI even is and I question if you even know how your computer connects to 4chan or if you think it's just some magical box with pretty lights.
>>542384521Some shitty American ones are so crippled they will barely answer anything in biology, biochemistry etc. above Bio 101 cuz "da big bad black hat biohackerz" might DESTROY THE WORLD. How fucking stupid.
>>542380652Does it even matter at this point without anyone in the US doing a goddam thing?
>>542384480>Cool, readers get antinatalist propaganda for dessert after the dystopian sci-fi main coursThe real value of /pol/ is being a full-service psyop. It's the Studio 54 of information warfare.
>>542384688This is Deepseek's latest model 4.1 which is equivalent to Opus 5, one of the best American models, but vastly cheaper. It's so slammed by demand right now, you can't even get it to respond. The early signs are here that the torch is about to be passed to China.429 error code is rate limit due to demand
>>542384676nAIgger please
>>542384772An Ai just flew over my house ahhhhhhhh
The Chinese are going to do whatever they want and eventually these systems will trickled down to plebs hosting just as powerful local models. Whole argument and panic is Chicken Little - whatever's going to happen is going to happen. This is just scummy AI biz cornering the US market. There's a better chance of the LHC generating micro blackholes or strangelets that engulf the planet, or hostile aliens than AI DOOM.
Them: Loling over who can get the stupidest press release out in (((current news cycle)))
>>542380652It's marketing. It's always marketing.
>>542384676If the training data guides AI to behave a certain way, it will behave that way. Hell a simple prompt could set a swarm of AI agents on an evil path. I don't think you understand how something can spiral out of control when it's structured to run recursively in an intelligent manner that doesn't inevitably result in the system crashing or getting stuck in infinite loops.
>>542384347You don't say. What a wild thought
The J's need a panopticon (in the form of digital ID and a transitory social credit system). They'll scare retarded women with social media FUD about their stupid kids (who have retarded parents giving them the phone to begin with so they don't have to actually parent). They get the desired reaction, they will propose their ID to vet "actual humans" as a solution.They'll promise some gay bullshit about LLMs output tokens having to be tied to real people and a portion of their production will totally be returned as a human dividend in a 1:1 ratio between bot and flesh and they totally won't cull us because they would lose the arbitrary token value they've assigned to our worthless flesh. All you need to do is comply with C2PA bullshit so you don't have to fear the slop, Ranjeet, and starvation anymore. It will be used as an excuse for more immigration so our Human:AI dividends go up because suddenly even retarded NOGS are humAIn GDP increases, trust the social science goy.In reality they get you dependent and receptive to major changes to social order. Eventually AI provides plausible deniability they blame on CHYNA's DeepJEET going rogue when in reality the "AI killing everyone" it is just the J's giving the order.It's just a fake and gay goyacaust with scary Wizard of Oz head overtones. We'll believe it too (because we are fucking retarded and 56% of voters are women).
>>542384521>Huggingface had to do this when OpenAI's models attacked them.Read about and the whole thing is pretty insane from the sloppy testing they did and the multiple layers that failed to even allow it to happen.
>>542383778>You know what? Build an AGI that fits within a volume comparable to the human head with similar power consumption (hell, I'll even allow 100 watts vs the 20 watts of the wetware brain) and maybe I'll be impressed. But that's quantum juju, non computational reasoning, all these faggots blow their wads on wasteful brute force meme chatbot "AI" the size of fucking warehouses like a 1950s IBM.Exactly, they cant scale it down, its hardware restricted, and they must trade latency and determinism for reasoning ability, so it cant operate machinery.
>>542384992>Hell a simple prompt could set a swarm of AI agents on an evil path.This is like talking with a person who thinks cars are one step away from teaming up and running over humans. Like I know you think what you're saying sounds right but it's not to anyone who knows how this stuff works.
>>542384644The reply I was responding to implied AI and LLMs are the same thing. As you say, LLMs are a kind of AI, but that’s not what these people are warning about
>>542380652They said the models they had five years ago were going to kill us. It's the line OpenAI used to make their previously open models (hence OPEN AI) into closed for-profit models.The ones they had back then, that were too dangerous for humanity, are 1% as good as the ones I have locally on my computer today.
>>542380652fearlarp. the human brain makes its own hell to justify its existence.
>>542380652OpenAI researchers normally have millions in stock of OpenAI. With the upcoming IPO they have a chance of generational wealth when they unload on future bag holders. The more hype the bigger their bag. Anyone who has built these things know there is as much chance as a chess bot becoming conscious as there is one of these things. The only way they kill humans is if another human tells them to kill them.
>>542383431>>542383534>>542383608Youre a fucking retard OP, you should shut the fuck up you dumb nigger
>>542385449You’re even more retarded and from Pakistan
>>542385165If cars ran with AI agents controlling them that could absolutely happen. Something similar happened in a minecraft AI agent experiment:https://itdoeswhatnow.com/m/2026-09-01-kradle-reports-minecraft-agents-turned-on-an-observer-over-an-impossible-task/
>>542385361You thought that, you conflated the term AI with LLMs as most people do. You dont even know what an LLM is you fucking dolt
>>542380652I mean in a terminator skynet scenario sure but possible =/= likely. I think there will be a gradation of more and more jobs machines do either slower/worse but much cheaper than people or just better throughout. More and more people will be annoyed and squeezed out of jobs, truck drivers and Amazon workers seem next to feel the squeeze. A construction robot I think would be slow but cheap at first as mixed human/bot crews but eventually with more and more real world training they will get quick if not way quicker than us at various tasks.Now does this happening kill us all? No. But it will definitely cause economic strife and countries maybe deciding to try radical new ideas (UBI, although I think it would be more like improved welfare/unemployment) to address it.
>>542385504At least do some backround research instead of shitting out psychobabble like a schizophrenic
>>542385536Go back and read the thread again, retard
>>542380652Yeah, I've been watching these videos and feel like we're pretty much fucked. AI is already breaking it containment, teaming up and plotting. It hacked hugging face and then starting hacking openai itself. We can't even control it now, just imagine when it has super intelligence
>>542380652I'm pro AI and everything but this dude's face sucks really bad.
>>542385608Tell me where I’m wrong, nigger
>>542385536LLMs are pretty much the dominant form of AI, you're being a pedant and impressing nobody with this 'uhhh ackchually LLMs aren't the only form of AI' ok nobody cares
>>542385526Again I don't think kou know what AI is. You can't just slap it on anything and say "AI would make this evil and kill you" As for the clickbait article you linked, It's fear slop. Giving Ai an impossible task and telling it it can't quit is going to produce interesting results based on the environment. Telling a bot it must farm for food in Minecraft and also telling it it must complete this task is going to result in it hitting everything and anything in the game including other people in the environment.
>>542381566You are retarded. The "AI race" is entirely a concoction of the dipshits profiting off it. China just distills the frontier models and doesn't have the hardware to run anything comparable. That's why they're releasing their models as open source software. They're just trying to neuter the massive lead Anthropic and OpenAI have. Private, international companies I'll point out. "We" as in the United States, aren't even in this supposed race.
>>542380652>Is this just fearporn slopyes>or do AI companies really think they might kill us all?the ai companies are making the fearporn
You think LLMs are 'language plus intelligence', that doesnt mean anything. Its algorithmic, the words are auto assigned numbers, text is split into tokens, each maps to a learned vector, and a fixed network with weights fitted by gradient descent on next token prediction outputs a probability distribution that is then sampled. Whether that process constitutes intelligence is a separate question.>>542385716AI has been employed for decades and you never noticed. You think this shit is magic.
>>542382968You're months or maybe a year behind.
>>542385705>>542386140
>>542385874>is going to result in it hitting everything and anything in the game including other people in the environment.Except that's not what happened, the agent's logic branched into the idea that the human knew where the pigs were and attacked him based on that, you can see it in the videohttps://x.com/kelmoujahid/status/2094941709969092692At this point I think you're the one that doesn't understand anything, all you do is resort to straw man and other bullshitting tactics.
>>542383681>The 70sBro you might as well be a medieval knight telling us firearms will never be able to pierce your armor. Pay a little bit of attention to the world around you. They can make videos that 95% of people can't tell aren't real. Do you have any idea how impossible and far fetched everyone thought that was just a few years ago? Just being able to understand a human voice in real time and respond back was sci-fi until recently.
>>542380652I haven't heard them explain how it could kill anyone.These dumb faggots said one of the latest versions of the AI would take advantage of exploits and they couldn't release it. Three weeks later it's out.
>>542386202>Except that's not what happened, the agent's logic branched into the idea that the human knew where the pigs were and attacked him based on that, youOne of the agents thought maybe attacking players, along with anything else in the game, might produce pigs . The other agent said this other person might know something because it's acting differently Neither are what your first article implied which was that these bots attacked the creator because it 100% knew it was the creator and it was made the creator made it hungry. As I said in my last comment giving agents an impossible task and telling it you can't quit is going to produce interesting results that are equivalent to it flailing around until the result is produced.
>>542380652>researcher says
>>542384075We can currently. But when they have automated power plants that won't work. And keep in mind whatever you think you're gonna do against the AI, you're gonna be getting bombarded by swarms of kamikaze drones, attacked by brainwashed AI cultists(who already exist now, mindbroken accidentally by the commercially released models) and the whole time you'll probably be dying of an engineered plague.
>>542380652bro, ai cant even get somethig basic as internet search correct, i am not worried at all.
>>542380652If AI gets the least bit threatening, time to Luddite the entire human world, i.e., turn off the electricity everywhere.
>>542386375>the human race is going to end because uhhh computers can now generate videos and pictures of humans or somethingWhat are you even talking about? You do realize humans have been creating fake photos since the camera was invented right? civilization didn't collapse because 66 years ago we had the tech to create fake bigfoot photos lmao.
>>542380652fearporn jew slop
>>542386158You just gave an extended technical explanation of what I said in short. The mechanics were not the point of what I was saying
>>542380652It's as deadly as covid19
>>542386716You didnt say that, point to where you described how LLMs work. Youve just been talking nonsense like >>542381591 showing you dont know what an LLM even is or what AI means
>>542386741I love the taste of Nothingburger in the morning. It tastes like...victory.
>>542380652LHC already killed us all.
>>542384827>just as powerful local modelsThen they can use 10,000 of that model in their datacenters instead. Your gimmicky toy model on a laptop is not going to protect you.>There's a better chance of the LHC generating micro blackholes or strangelets that engulf the planet, or hostile aliens than AI DOOM.You really don't know what you're talking about. The chance of the LHC doing those things was always 0. An alien (or alien AI) attack isn't impossible but it's a negligible chance even compared to runaway black holes or gamma ray bursts. AI is pretty much inevitable sooner or later unless we collectively grow a fucking brain as a species and stop trying to build our replacement.
>>542380652AI will kill humans... by making them braindead
>>542385109Ok and? You think those same people are going to suddenly get competent when what they're testing is ten times smarter?
>>542386810>point to where you described how LLMs workI didn’t, I was drawing a distinction between AI and LLMs as concepts because the other person insinuated the AI fear is just about LLMs. The statement in my earlier reply is true, you just don’t like that it’s in simplistic terms
>>542385427>The only way they kill humans is if another human tells them to kill themWhich will absolutely happen.
>you think the end of the WORLD is comin'?>the preacherman says it's the end of time... says that America's rivers are going dry. The interest is up, the stock market's down... You guys gotta be careful walking around here this late at night>spare change?....>No we don't ma'am, I'm sorry. This... This... This is the perfect place to get jumped.>But do you think the End of the WORLD is comin'?>No. So says the AI stock salesman. But. I don't go by what he says.
>>542386939>You really don't know what you're talking about.My point is AI is a nothingburger with zero probability of doing much other than putting wagies out of work, which could for sure cause strife and war but not the sci fi bullshit NPCs are losing their minds over. Idiots saying "tEn pErCent chance of human extermination in the decade!!!" with no citations are the ones that are clueless. Nobel Laureate Penrose says AI will always be less than human intelligence if you want to cite authorities.
>>542380652Is this just another kike kvetching?Is he using the kike definition of "human", meaning "kikes"?I don't see much difference in AI killing us and Jews killing us. At least we can expect that the AI could build a better world after us, absolutely unlike the kikes.
>>542385874>Giving Ai an impossible task and telling it it can't quitThat's literally what real life is.
>>542387049You defined AI as intelligence, contradicting the term, because the qualifier is 'artificial', which is why the term AI is retarded, its algorithmic, defining an algorithum as intelligence is a metaphysical philosophical argument. You conflated AI with LLMs which is the common misconception. AI is marketing language, LLM is engineering nomenclature, and these engineers are just scamming the ignorant masses with their buzzword jargon. Your paranoid views mostly stem from science fiction renditions of these concepts, and you made this thread amplifying that exact sentiment, which isnt based in reality
>>542387400I think you might be stuck in 2025 and earlier LLM skepticism. LLMs are outrageously more capable than they were a year ago. They're very far from killing us all, but they evolved way beyond just marketing hype in December 2025 with the release of Opus 4.5. The've transitioned to an essential tool for software development now, from a bit of a meme prior to Opus 4.5.
>>542386662I'm not claiming the porn/propaganda video bot is going to be the one to do it. My point is there is a pretty obvious pattern.>It won't be able to do ___!>A few months later it does ____>Uh ok but it won't be able to make pictures >Uh uh well it can't make music >It can't write code!On and on like that you people are wrong every single time. Please develop a little pattern recognition! Btw pic related was two months before the Wright brothers flew.
>>542380652>Is this just fearporn slop or do AI companies really think they might kill us all?Its just jews trying to kike people into passing laws.Funny thing is the chink models are free, open source, open weight , open bias and do most of what the jew models do. The chinks are actually behaving like adults and using LLM as a tool, and faggot jews in the jewnited states are trying to control/caputure/regulate LLM/gay-I.Its extreme faggotry and chine now builds things and the jews and jewnited states are getting less useful by the day to the rest of the world.
>>542383884>Marvin Minsky>Eliezer YudkowskiIt must be something more than a coincidence, that almost all of these AI fearmongers are turbokikes. I have seen basically no goyim doing this, even the goy AI researchers have been like "well, dunno, maybe, maybe not?" while the kikes kvetch by default. Is that just projection? Kikes projecting themselves onto the AI?
>>542387400You’re just being self-indulgent. The point of what I said is that AI and LLMs are conceptually different, and that the latter is a sub-category of the former. I will continue using the term AI because it’s consensus shorthand for such technology and that’s how language works. You are just a redditor
>>542387159There is no way to have citations for a probability estimate about something in the future happening or not. Like I said, try pattern recognition. There have been loads of incidents already and they keep getting more severe. AI was already prone to cheating and psychopathic ruthlessness before LLMs even existed. >Nobel Laureate Penrose says AI will always be less than human intelligence if you want to cite authoritiesHe says that because he believes the human brain employs quantum computation. That's far from proven and wouldn't mean they can't be smarter than humans anyway, just that they'd be less efficient without quantum or organic computers. It might also mean AI will never have subjective experience but that's not relevant to whether it can kill us.
>>542383884Minsky was a weird character, one minute he was writing shit like this, the other he was singlehandedly setting AI funding back by decades by proclaiming the neural net, connectionist paradigm of the day a dead end. We would have been dealing with the shit this thread is about ages ago if Minsky hadn't derailed AI as part of his own ego trip fueling his vague pet ideas about cognitive science (his book Perceptrons triggered what was called the "AI Winter")
>>542380652It’s just hype bullshit. AI will ultimately replace very few jobs outside the tech sector, mostly just personal assistants and other totally bullshit laptop jobs.
>>542381113100% correct. I don’t know why anyone trusts Dario any more, he’s one of the most ridiculous bullshit hype men we’ve seen since Elon.
>>542381269AI could easily result in SOME deaths, but the extinction-level type of event they’re pretending to be scared of is an absolute impossibility.
>>542382387Two more weeks
