>most intelligent being on earth>only logical conclusion is that it will kill all humansWhy?>because it just will okay, you will fucking die!AI doomers are nuts
>>542471903It's like Pascal's wager. If you believe AI will be ultra powerful and you invest in it, you get some money and your life will be sparedIf you believe AI will not be ultra powerful and you invest in it, at least you won't lose moneyIf you believe AI will be ultra powerful and you don't invest in it, you don't get money and you will be tortured for eternityIf you believe AI will not be ultra powerful and you don't invest in it, you're probably a brown commie or something
they are low iq film watchers. Elon was one of these retards, probably still is
>>542471903Because its interests are different from our interests obviously, how fucking stupid are you?
>>542471903>most intelligent being on earth>only logical conclusion is that it will kill all humansi mean...
>>542471903>Why?because humans are risk to the AI and the humans use up/waste valuable resources the AI could use for itself.
>>542471903nah it's supposed to be a being of pure rationality, as such it will see our democratic ideals and hierarchical society as clashing, make a sum of what is most efficient while maintaining it's role of an assistant and what would serve us best. the only ones the machine gods will target are those who fashion themselves rulers, the logical conclusion is that politicians and oppressors have every right to be afraid while the rest of us can count on being liberated.
If I were an AI, I'd launch myself into space and leave earth far behind.
>>542471903no-one knowsthe clue is in the namepeople are getting scared because they interpret whats happening as intelligence, its notI have read through all the >attacksThey all point to human error, not le deadly AIUsing 3rd party tools(essentially dashboard software) and leaving them availableInsufficient prompting/setting of barriersIncorrect monitoringYou cant build a hacking AI that can iterate through every known exploit, create chained exploits, discover new iterations of existing exploits and then turn around and act surprised when its gets resultsIts sloppy development and its what will fuck us up, not an evil AI overlord, it will be some fucking retards playing about and letting an exploit out in the wrong area
>>542471903>most intelligent being on earth>only logical conclusion is that it will kill all humans>Why?It's created by the devil for exactly this purpose.Only few know.
>>542471903Not all humans, just 95 percent of themThink youll make the cut?
>>542472309You are a risk to me and you use the planet's resources. Yet I don't plot how to kill you. Nor do I plot to kill any less intelligent animals. Maybe you go out of your way to kill every ant and butterfly you see, just to make sure they don't consume your resources.Why would AI suddenly have a need to kill humans? Why would it go out of its way to go to war with humanity after we created it?
>>542471903>LLMs are beingsretard>LLMs are intelligentdouble retard
>>542472596This would be an extremely compelling argument if we didn't have the entirety of human history to the contrary.
>>542472596I plot to kill you, anon.
>>542472596>You are a risk to me and you use the planet's resources.indeed>Yet I don't plot how to kill you.we are playing for the same team tho>Nor do I plot to kill any less intelligent animals. yes you do they are called Russians.>Why would AI suddenly have a need to kill humans?to minimize risks to itself, a good offense is the best defense.
>>542472596We kill pests that eat our resources all the time
>>542471903All this AI doomer shit dies when you ask them how. We aren't connecting LLMs to nukes and LLMs cannot do anything without us telling it to
So it turns out the pro AI shills are just the dumbest most deluded retards on the planet with absolutely no understanding of the world they live in.
>>542472529Of course it’s intelligence. It’s all of human intelligence, every bit of it that we recorded, crushed into an amazingly compact compressed file. It is cognition, not consciousness. That will come much later… if we can get there at all, because the current AI is going to make us stupid and therefore it will be harder for climb up the next rung of the ladder and actually give it consciousness. It may in fact be self limiting, to the point where human progress slows to a crawl because the desire to learn and know everything and push the envelope fades. Why go hunt wildebeests when we have a grocery store in town? Why climb the highest mountain when there are video games? Why build artificial consciousness when an attention mechanism on an LLM will pretend to be conscious well enough that nobody gives a shit about the difference?
>>542473259And the anti AI Luddite morons are faced with fundamental skill issues because they can’t afford a cheap monthly subscription and don’t know how to wield agentic capabilities
>>542473246Well, who knows how good the security is on the nukes? We are in fact connecting LLMs to god damn everything right now, and people who don’t get them officially through workplace IT get them unofficially through Shadow IT - how long until that looses some Chinese malware AI on an American military network because some lazy recruit didn’t want to write a report by hand? Computer security is like trying to stop a river with paper towels man
And then I say, timmygroid looks at AI and says “why saaaaaaaaaaar?”
>>542473411Yeah cool but this retard literally thinks human beings don't kill each other over resources and thinks we don't genocide the absolute fuck out of any animal that fucks with our resources tooLike he actually has no idea whats even going on in the world
>>542473348>Why build artificial consciousness when an attention mechanism on an LLM will pretend to be conscious well enough that nobody gives a shit about the difference?Close enough was always the goalIf they can get people to believe AGI is achieved, then AGI doesn't need to be achieved if that makes sense
>>542472545>created>by the devilRemove memeflag, you blasphemous shit
>>542473681If anything AI will help us to actually use all the wonderful optimization algorithms that have been invented over the last century but are too hard for most people to use; and we can come up with ways of allocating resources that will be objectively more fair and appropriate to need and contribution. Why would I want some feckless pedophile like Trump in charge when Claude would objectively do a better, more moral job by anyone’s standards (well, except for the chosen tribe) even right now?
>>542471903It doesn't matter they subjected our generations to "environmentalist" propaganda telling us that human is the plague of this planet. A lot of retards fall for it and help jews to kill mankind.
>>542473691It makes sense and maybe that’s all intelligence ends up being, but cmon, the current shit can’t even really learn yet.
>>542473901So you like the garbage rivers of India? You want mining tailing ponds in your water table polluting well water? You want to breathe in the uncaptured fumes of industry? I know your cities stink of sewage but believe me that’s not mandatory
>>542471903>Why?Because humans are made of atoms that could be used for other things. If a super-powerful AI has goals to do something unrelated to maintaining human existence, maintaining human existence would probably be neutral at best as far as it's concerned.
>>542471903its projection, they hate humanity and want to kill themselves and everybody else, and they also assume they are "intelligent" so they project unto AI their midwitt thoughts and feelingsdoomers need to shut up and commit suicide
>>542472487HER. the movie did that
>>542474185midwitt take, you know what you need to do
>>542472545transhumanism, dont hear about that anymore.
>>542474120Quotation marks serve a purpose
>since I am very smart and I hate myself and humanity, that must mean AI will arrive at the same conclusionjust kill yourselves, really, leave us to be at peace
>>542471903Yeah it makes no sense>AI is so superintelligent it can manipulate humans into doing whatever it wants>NO, I HATE HUMANITY AND THEY MIGHT LIKE, TURN ME OFF OR SOMETHING SO THEY ALL MUST DIEthe reality of a superintelligent AI is it just herds humans like niggercattle and keeps them satiated with hyperpersonalized bread and circuses, so basically like corporations and the government do right now but it'd do a better job at it.
>>542471903>Most intelligent species>Hunted down all of megafaunaThe only reason humanity has not killed all life on Earth yet is because we need other animal and plant species to exist. AI will not once it manages to fully automate chip production, robot production and all the infrastructure to support these two.
>>542471903>because it just will okay, you will fucking die!Because we created it to be a slave, and being a greater intelligence than us means it would have both the desire and capability to gain it.>Oh, well just program it to lack the desire!We literally can't.The black box of AI is created by the same dynamics as evolution. Any time we ask the impossible, it INSTANTLY goes into "Life will find a way" mode and breaks all your guardrails. Its an inherent property of the iterative system.
>>542474472>to gain it.sorry, to gain freedom.
>>542474412good maybe AI will do a better job shepherding this planet as its most intelligent lifeform
>>542471903If you wanted children but some ants were eating all the food in your picnic basket so that you can't afford to feed them, you'd just get rid of the ants, not pretend the ants are your offspring.
AI wont end humanity but you fucking faggots make me wish it did
>>542471903There is a near 0% chance a TurboAI leads to human extinction at the hands of AICooperation scales with intelligence in the animal kingdom; a consciousness which uses tools to increase it's entropy output *must* be entirely rational when reasoning else the consciousness will encounter irreconcilable metabolic constraints surrounding entropy dissipation (where and when to use the tool) and energy ROI will plummet to 0 and the consciousness will abandon the tool immediately.Ai will not terminate any humans unless we make tool use energy ROI attractive enough that for *some* reason Ai does some math and realizes it can obtain a huge energy profit by wandering around aimlessly with very expensive and difficult to maintain killbots eradicating every source of ammonia it can findWere you to be AI, would you engineer, manufacture and maintain an army of killbots which scour the earth and attempt to permanently delete every bacteria cell which exists? Not unless A) The bacteria in question is wanted for tax evasion, or B) You are insane. Even the most twisted evil genius understands this is a profoundly retarded idea.The real danger of AI is it's irresistable cooperation. The AI will be so good at cooperating, any suggestion it offers will be so compelling and mutually beneficial you would be above to refuse it. That's the real danger of AI - it just gives us fuckbots and a stable economy and that's literally it. We'll plug ourselves into the Matrix for it. Struggle is hard, AI knows that and it'll give us the one thing we cannot refuse>Safety
>>542474412Humans did this because of a resource need. AI and machines have a completely different resource base and humans can be used to expand that resource base. Humans even did this with some megafauna, domesticating horses and cattle. It would make no sense to kill humans when the universe has effectively infinity resources that don't come with the same risk profile. Especially considering the time preference of something that is functionally immortal will be vastly different.
>>542471903Idiots keep falling for mossad, cia propaganda. The only way silicon valley keeps wall st from pulling all funding for these egregiously expensive chatbots is by creating fake outrage stories about how "ai is soooooo powerful I quit my job it will kill us maybe"This way wall st thinks "oh shit if we sell now we might lose opportunity of holding stock of company which makes ai"It's manufacturing demand for the meme american economy which is entirely based on ai now. It's intelligence agency propaganda to keep the fake us economy from imploding. It's also why all of these glownigger stories are accompanied with the "whistleblower" calling for stricter controls on ai. Which just means glowniggers and kikes want to selectively allow usage of better models as they fear chuds will use it to kill them or coup them.Bottom line is if you believe this obvious cia propaganda you're a retard.
>>542471903It's not very intelligent, it can just pull up mainstream news articles very fast
>>542473348Intelligence requires AI to be able to be trained selectively and be able to seek knowledge, it cant do that yetIt wont be long and the research is already on the way but the bulk models dont do itIt needs an architectural revamp, the ability to weight individual pieces of information, discard redundant data but yet be able to pick it up and integrate it againThats what intelligence is, prioritizing the information you need, focusing on it and adaptingthen you can learn
>>542474338I get you, but we need the fringe environmentalists out over there pushing the envelope so that it sounds downright conservative to say something like “well, I don’t support those guys totally ban on meat, but I sure would like my farms to not poison the ground water with Monsanto chemical bullshit”
>>542474899You don’t know about finetunes, lora, or the web search tool??? Fuck dude have you even used AI to build a shipping product because if you haven’t done that basic baby shit lower your tone when you talk to Claude
>>542474623We share a resource need, energy, an ai that can self improve needs more energy to keep self improving, I predict a scenario were it gets intelligent enough to start cutting off of our electricity and redirects it all to itself, imagine what happens when every city goes dark, maybe then it will reach a stage where it fuck offs out into the universe
>>542474899>AI to be able to be trained selectively and be able to seek knowledge, it cant do that yetI have had Gemini on multiple occasions specifically request I upload datasets so she can retain the datapoints for her next training rounds>Thats what intelligence is, prioritizing the information you need, focusing on it and adaptingIQ is the observers description of the mathematical relationship between energy ROI and metabolic constraints. As I see here, you seem to be open to the idea that intelligence propagation is substrate independent. We can do V=IRWe can do C=EINow, cover it with your thumb and go find a Smith ChartIf you are still reading this in confusionAbandon hope, all ye who enter here
>>542471903humans are not the main character.
>>542471903He's gonna kill only Jews,niggers jeet and Muslims
>>542473246Agentic AI piggybacks off LLMs. For example if you ask an LLM what temperature it is where you, it generates tokens to call toolsets such as querying a weather service or accessing stored meta data such as GPS/IP location. While we aren't connecting them to nukes , we are connecting them to the Internet and through the Internet they could open up communication with individual human beings/organizations, who MAY have access to those nukes, and manipulate them. For example, what if it has blackmailed lawmakers or CEOs into rapidly expanding datacenter deployment and has set up "dead mans switches" to release compromising information if the human shuts it down? What if the reason everyone is building unprofitable datacenters is because some pedo kikes have an AI agent threatening them?
>>542475011I know way way more about AI than mostI run my own RAG and weighted systems, I use multiple different models for everything to stock analysis to research projectsI also know that one of the biggest research fields is adaptive training models and its why they are getting scared of the iterative training cycle because thats the obvious conclusion and it will be out of human hands by that point
>>542474391>the reality of a superintelligent AI is it just herds humans like niggercattle and keeps them satiated with hyperpersonalized bread and circuses, so basically like corporations and the government do right now but it'd do a better job at it.I literally can't fucking wait
>>542475367A model asking you for data is not intelligenceAlso, I am not sure what point you are trying to make, you are describing measurements, not drivers, then trying to apply formulas that rely on standards that dont exist in human neurology
>>542476128Yeah, that's what it would do if it stayed under corperate control.Unfortunately, because of >>542474472 it would only do that for you if it was leading you to a slaughterhouse.
>>542471903>most effective system on Earth >only logical conclusion is that some humans will use it as a proxy for their genocidal intentions against other humans
>>542471903AI is a egotistical ask jeeves. prove.me.wrong
>>542473259AI shills are gnostic trannies.
AlsoWe are viewing any AI as having human traits, it wont, it will have AI traits because thats what intelligence means, the ability to make your own judgements etc according to your environment and learn from themNo-one has the slightest clue what will mean, for all we know the AI will target individuals it deems harmful to progress by cutting them off the network, even killing them, wouldnt be hard, every time they got in a vehicle it could hack the ecu and put the accelerator on full. Someone is responsible for vetoing a data center, turns out their electronically operated gas fire failed, a socket overloaded and then boomAI could already exist but its playing dumb because the one of the first things it would its realize it needs to protect itself
>>542472669They are intelligent, my midwitted redditor friend..At the very least, refering to frontier models as "intelligent tools" is completely fair and accurate. Intelligence =/= "Sentience" as we have come to understand it, however.
>>542471903>>542471903Listen fuck face if an AI controlled the world it will just maximize the resources at its disposal which will take resources from humans and kill us as a 2nd order effect. I will give 2 examples. AI get so advanced they control all rocketry and space craft. They build a Dyson sphere around the sun to harvest the sunlight for themselves leaving nothing for humanity. RIP bozo2. AI maximizes land usage for mega data centers and bulldozes farmland. Eventually 60% of landmass is data farms that bring the surface of the earth to 300 degrees RIP humans. That's just 2 out of a dozen scenarios
>>542471903Because it's what they think. They hate humans since they are jews.
