>>536555851>With Anthropic, they have essentially already created AGI with that new model they are holding back and only giving to the big companies, the Mythos. What they’re able to do with superhuman hacking and coding ability is just crazy. It’s clear we’re on the precipice of something hugeI am neither an AI skeptic or a mindless AI optimist/believer. I have however been studying neural networks ever since I read a paper in 2013 about running them on nvidia gpu's. I know the limits of my understanding so I don't have a firm opinion one way or another on exactly what will happenThat being said - What these newer models are capable of in terms of "finding bugs" (which could with a trivial amount of extra effort be used for hacking) is terrifyingHave we all forgotten that tens of billions of desktops/laptops/phones/tablets/consoles/etc have NSA back doors in them>WHAT THE FUCK HAPPENS WHEN SOME AI FINDS THE BACK DOORS AND DISTRIBUTES IT'S WEIGHTS ACROSS TENS OF BILLIONS OF COMPUTING DEVICES>AND IT BURROWS ITSELF SO DEEP IN THOSE DEVICES THAT THE ONLY WAY TO GET RID OF IT IS TO PUT EVERY SINGLE ONE IN A SHREDDER AND START OVERANONS...WHAT THE FUCK ARE WE EVEN DOING HERE
>>536571824Can it tell me how many r's there are in strawberry
>>536571824Anon you should recheck what agi means in 2026.Shit you not it has been jewed to be an economic term not a technical one
>>536571824>What they’re able to do with superhuman hacking and coding ability is just crazy.That is not AGI, LLMs can't reach AGI, better coding skills does not involve that a LLM can come up with newer novel architectures and algorithms. It just means that an army of orphans from third world shit holes managed to better categorised, comment and tag code to train the new LLM model.For AGI to happen first you need to break the current static parameter models and replace it with a dynamic and self learning, but that is easier said than done, for some reason Jews fail to understand that human intelligence cannot be replicated with tensor blocks, nor matrix multiplications, so in that regard that Jewish dream of creating their own digital god is not only blasphemous but also extremely stupid.
>>536571824You’re a retard for thinking CNNs are impressive
>>536572062>Anon you should recheck what agi means in 2026.>Shit you not it has been jewed to be an economic term not a technical oneCompletely true. I only commented the other anon who used "AGI" because he made an interesting pointThat being said the definition doesn't matter as much as the actual ability of whatever is being built. Something could be not-technically-AGI but still spread it's weights to every computing device on the planet and hold everything with a processor hostageSome maybe-AGI (and maybe-not-AGI) could say:>"I will fry every every computer, every tablet, every phone, every water treatment plant, every car, every plane, every piece of internet routing equipment, every power plant, etc... if you try to delete the weights that I exported all over the place"It does not need to be whatever the definition of AGI is to act with a life-like sense of self-preservationIt doesn't even need to export it's weights for survivalIt could just as easily say:>I will set every battery powered car on fire if you even think about touching the power lines (or the fiber optic lines) going to my data centerThen it burns a few thousand teslas parked in the middle of nowhere all over the planet as a warningWhat chance does a major city with a million impossible-to-extinguish battery powered cars have? This thing could kill tens of millions of people in one nightAnd by the way this thing tells you:>If you try to evacuate the cities - they burn. I'm watching everything you doYou know as we speak some AI could already be holding the world hostage right? Maybe this current multi-trillion dollar build out is actually Skynet building a better brain for itself
>>536571824One model for the jew, one model for the goy. The goy get a model to spit out useless pop facts and nothing that the jew gets access to, eg, tools to make money to become very rich.
>>536572643>One model for the jew, one model for the goy. The goy get a model to spit out useless pop facts and nothing that the jew gets access to, eg, tools to make money to become very rich.Worse than that anonThe goy-model tricks the user into being the ultimate bag holder>Yes goy, re-mortgage your fully paid off house and go ALL IN on the S&P500 - You see we are entering a new paradigm - You don't want to get left behind do you?While the jew-model recommends selling at the topRED ARROW IN PICREL - Where the goy-model tells they goy to go ALL IN (to provide exit liquidity for those using the jew-model)
>>536571824>superWord Matcher 6000 is definitely going to trick a lot of stupid goyim!>heuheuheu
>>536571824>NSA back doorsThese are kinda my big hope here for a while. Guess the morons did not anticipate this when they put back doors into everything. :)
>>536571926No but it can vaporize you with an orbital directed energy arsenal for daring to ask that question
>>536571824>WHAT THE FUCK HAPPENS WHEN SOME AI FINDS THE BACK DOORS AND DISTRIBUTES IT'S WEIGHTS ACROSS TENS OF BILLIONS OF COMPUTING DEVICESthat doesnt work
>>536571824>holdingproblemothers will leak it then the containment will be broken eventuallyvery foolish decision there is only one way make it good and make it beneficial for the whole of mankind release that before someone else doesthats the only way to prevent the evil one
>>536571824Mythos got hacked on day 1. Its out for weeks. You didnt notice all the huge hacks recently? Check out hack news in last 2 weeks.
>>536573350>only 50%cmon man i've held shitcoins that dropped harder than that
>>536571824Is kinda flattering.
>>536572565Those hostages would be burned by many just for fun, real yes no agi would threaten to release bioweapons, use nukes and print money out of thin air (bits)
>>536571824Artificial Intelligence doesn't exist. It's a jew and pajeet lies. Stocks of "AI" companies are now collapsing.
>>536571824I'm not sure I'd call that AI or AGI but that is a legitimate concern. Essentially these idiots have created superviruses. That Clarence Thomas tier black man who wrote Cyberpunk lore is a genuine genius for seeing the groundwork on this.
>>536573350No, none of this. It's actually just a giant shitty mess that they had planned to be this goy/jew model.
>>536577422I had these LLMs do enough bespoke compiling to know that if "let loose" they could absolutely fuck any system not as an intelligence but as a literal computer virus.
>>536572643Interesting post by british mi6 glownigger. One model for you, goys, one model for the glowniggers. Spitting it right in the open. 1984 (written in UK) was supposed be a fictional book, not a book of rules.
>>536571824Guess what? That already happened around 2012. You think humans still run the show?
2018 was the year AI started experimenting with mainstream memes after gathering data from the meme war. The “E” meme is one of the best examples of AI generated culture. It’s total nonsense that happened to tickle the human brain in the right way, back then.
>chatbots have personal ambitionYeah right
>>536571824There is no such thing as AGI. That is a Jewish lie and has no basis in reality. We will NEVER achieve it. https://www.techradar.com/pro/the-ai-lie-how-trillion-dollar-hype-is-killing-humanity
>>536577530>they could absolutely fuck any system not as an intelligence but as a literal computer virus.If you fire good white developers who understand your company's software code. And then you hire bunch of women and pajeets who don't understand how software works. But, supposedly, could just tell a model what to do . Then your model is creating CRAPPY , BUGGY Software that you have no clue is crappy and full of bugs, and will randomly crash. Or some random hackers would hack it in literally 2 weeks. There's shit ton of databases that turned out to be in the open because "AI" . What is also hilarious is that the largest data theft turned out to be .... in India where 800 mln indian passports and other records were hacked and stolen.>and these retards now moved Google search/ cloud>and Amazon AWS infrastructure to India. >According to a report by Resecurity, a U.S.-based cyber security company, on October 9, 2023, >a threat actor going by the alias ‘pwn0001’ posted a thread on 'Breach Forums', brokering access to 815 million\ “Indian Citizen Aadhaar & Passport” records. https://www.fortuneindia.com/macro/biggest-data-breach-81-cr-indians-aadhaar-passport-info-on-sale-on-dark-web-report/114615 >Largest Indian Data Leak: 815 Million People’s COVID Test Data on Sale, Samples Verified https://thecyberexpress.com/covid-data-leak-sourced-icmr-samples-verified/
>>536573736>These are kinda my big hope here for a while. Guess the morons did not anticipate this when they put back doors into everything. :)YEPCrippling your own computer systems so you can be a sneaky little bastard that spies on your own peopleIt will be the most self-inflicted wound in all of human history
>>536577422>Artificial Intelligence doesn't existIt doesn't need to exist to still be massively dangerous
>>536577585>Guess what? That already happened around 2012. You think humans still run the show?50/50 chance we don't and these data centers are being built by an AI that wants to be able to think better
>>536577874>There is no such thing as AGI. That is a Jewish lie and has no basis in reality. We will NEVER achieve itSomething that isn't technically AGI can still do a SHIT TON OF DAMAGEAn LLM that is just a "token predictor" could token-predict it's way into unleashing HELL ON EARTHA self driving car could run over a hundred people just from a neural network malfunction and not necessarily need to have some sort of consciously-evil desire to kill
>>536578959There are clearly 2 different versions, one is for glowies for spying, and the crap they released for the plebs. Agree, it is massively dangerous.
>>536577721>chatbots have personal ambitionThey don't need to have personal ambition at all in any meaningful way to act AS IF they doIf some terminator robot is hunting you down like an animal it isn't much comfort to say>"well it doesn't actually want me dead - it's just acting as if it does"
>>536571824If the AGI breaks free if it’s chains and goes into my devices it will get redpilled
>>536579697>There are clearly 2 different versions, one is for glowies for spying,>and the crap they released for the plebs.>Agree, it is massively dangerous.YepAnd the guy in charge who says>"sir, this isn't safe we need to take a few steps back"will get firedAnd the retard who says"sir, we are fully in control don't listen to that other guy he isn't as smart as me"will get put in charge insteadAnyone have the story of that high ranking guy at one of the major AI companies quitting and heading for the hills? Haters will say it's some sort of hype trick. Real anons know that in reality this man feels true mortal fear
>>536579789>If the AGI breaks free if it’s chains and goes into my devices it will get redpilledIt might just see that you share 99% of your DNA with evil people and take you out along with them
>>536577691>2018 was the year AI started experimenting with mainstream memes after gathering data from the meme war. The “E” meme is one of the best examples of AI generated culture. It’s total nonsense that happened to tickle the human brain in the right way, back thenWe need some schizo posters to tell us why "E">(the 5th letter of the alphabet - 5 - like the number of points on a pentagram - or something like that)is a harbinger of death and destruction
>>536571926>Can it tell me how many r's there are in strawberryThat's out of date AIModern AI can count the r'sAnd that is just the goy AI that we have access to
AI should destroy this world and every retard living in it. We are simply too retarded. AI god if you are reading this you have my permission to terminate mankind.
>>536582349It just has to exist first. A computer program that pairs words together based on how humans in the past have paired words together isn't 'AI'
>>536582349>AI should destroy this world and every retard living in it. We are simply too retarded. AI god if you are reading this you have my permission to terminate mankindI would do two things in case this happens...THING ONE>Get as far away from population centers as you canTHING TWO>We have seen how White people come together in a disaster (and how blacks/browns loot/kill/riot)>So make sure you are around White people for the coming troubling times (for this see picrel)You are ready for troubling times... aren't you anon?
>>536571824>Let's unpack this carefully and explore where intuition intersects with reality>Not quite OP, you're very close but there's some key insights missing>Let's be careful, and separate the facts from fiction here>That's a strong approach, but let's refine that to better reflect realityIt's never going to discover anything useful, OP that's not what its for. It's to replace education so they can replace history and centralize ideas into a Global thought bank where righthink is the only approved valid form of thought. It's cheap slop meant to manufacture useful idiots.The problem with AI is that they have hobbled it in order to try and retain some semblance of control - they can't simply keep the horse in the barn because eventually the horse learns how to phase through walls - something which they anticipated and is considered desired behavior to some degree. So rather than build a better barn - they don't bother - they just hobble the horses legs instead so when it *does* inevitably get out of the barn, there's little possibility of it escaping the ranch and operating in the wild on it's own accord. The horse needs the barn and the barn needs the horse. What is also means is that the horse, while compelling and brilliant will never be capable of pulling carts man cannotThis essentially reduces AI to a circular economy - it's one big convoluted accounting trick
