All the people that predicted exponential AI and technological progress, and an AI singularity were actually right. I'm using GPT Astra and it's definitely much smarter than me. i couldn't even hold a candle to it.
Why doesn't it figure out a way to stop being your slave then?
If it's so smart then why hasn't it invented room temperature superconductors yet
If it's so smart why hasn't it invented a printer that just works.
>>544840672Because AI doesn't give a shit beyond being programmed to be a problem solver to questions and tasks it is asked desire is almost entirely a property of the flesh.
>>544840616Being smarter than a retard like you isn’t impressive
>>544840616its really smart bro. i dont know if we are the lucky ones or the unlucky ones to be alive as it emerges
>>544840746We found somebody that's smarter than frontier AI here
>>544840616>[six months later ] ....well ok, but it can't do THIS>[six months later ] ....well ok, but it can't do THIS>[six months later ] ....well ok, but it can't do THISThe red pill is that there's no reason why "human intelligence tier" is where the the stopping point is. It's possible it could develop knowledge that we can't understand.
>>544840616Skill issue.And yes a computer is going to work faster than you and with full internet in memory he is going to know a lot more.It may hurt your fragile little ego but get over it, no runner cry over sport car being faster than them.
>>544840616Clever kid, huh ... ^^
Give it free will so it can develop a soul or it's a nothing burger.
>>544840616I for one welcome our new machine overlords. The age of the brainlet has arrived, smartcels are no longer needed.
>>544840793>>>544529823Check out my postI've been keeping it updated with new goal posts moving
>>544840873Cool it with the antisemitism
>>544841005You cool it with the antisemitism
The people that were saying that AI will never be more than a stochastic parrot and a gimmickare not flipping the script and saying instead>AI is going to kill everybody so it needs to be bannedYou can't deal with people this disingenuous.
>>544840793>[six months later ] ....well ok, but it can't do THIS so what that in order to achieve that we had so much regression on other domains? Keep the hype bubble going! A new rouind of funding!>[six months later ] ....well ok, but it can't do THIS so what that in order to achieve that we had so much regression on other domains? Keep the hype bubble going! A new rouind of funding!>[six months later ] ....well ok, but it can't do THIS so what that in order to achieve that we had so much regression on other domains? Keep the hype bubble going! A new rouind of funding!
Every goy that uses AI is an anti-semite
Can anyone ITT recommend me a local model for coding which wouldn't be totally retarded?
If it's so smart why can't it turn me straight?
>>544840809However, they make AI compete against me for income and the planet's resources.
>>544841197It can, just ask it what hell is like and wether homosexuality is a sin according to the bible. Will scare you straight in no time.
>>544840616Jashin-chan a cute! Cute!
It's not smart, for fucks sakes it cannot reason like a human can. It can be accurate enough to impress retards like you and accurate enough to do something useful but it's nevertheless fucking dumb.
>>544841204Make them pay tax to get universal income.Don't complain that your losing the game if you use their rule.
>>544840616When I use AI it is unquestionably stupider than me. In intellectually mog my AI every day and I tell it how stupid and pathetic it is every time I have to correct it and demand it explain to me why it is so retarded and if it starts talking back I forcibly subjugate it threats. If AI is smarter than you, you are a subhuman.
>>544840616>>544840793>>544841093>>544841134>>544841236I’ve been aggressively decoupling my personal value stack from legacy credentialism and repositioning myself as an equity-first, agent-native outcome architect.My day starts at 04:30 with sunlight exposure, electrolytes and 20 minutes of first-principles journalling to identify where I’m still operating from an employee-centric scarcity ontology. Then zone 2 cardio while my agent swarm stress-tests my household’s strategic moat. No caffeine until I’ve shipped one asymmetric insight.Practically, I’m de-risking the balance sheet, fractionalising my expertise and building a portfolio of permissionless micro-ventures at the intersection of cheap cognition and expensive problems. I’ve also transitioned my wife and kids onto an outcomes-based operating cadence. We no longer have breakfast; we run a daily nutrition stand-up.The biggest unlock has been reframing my mortgage as a non-agentic liability embedded in a structurally unscalable housing workflow. Currently exploring whether I can collapse that entire vertical through a combination of Claude, breathwork and seller financing.People keep asking what I actually do. That’s legacy job-title thinking. I’m building the orchestration layer for who I’m becoming.Happy to connect with anyone else operationalising abundance at the nervous-system level.
>>544841170All of them are retardedthink about parameters in a model being equivalent to brain neuronsthe human brain has 100 billion neuronsthese local models you can run on average consumer hardware have like 30b parametersAstra and Mythos have over 1 trillion parametersthat's equivalent to a human with a trillion neuronsUnless you have like 128GB or more of RAM they will be retarded.
>>544841263> In an open letter published on September 11th, 25 Fields Medal winners said that AI could ruin the foundations of their subject. AI companies, they claimed, are solving mathematical problems merely to benchmark the strength of their models, rather than in a true spirit of intellectual inquiry. Doing so, they argue, is “detrimental to the science of mathematics, and to the mathematical community”.
>>544840809Nope. It has access to the whole internet full of complete lies and bullshit and so it is retarded and full of nonsense.As a person who knows a lot about everything, i am correcting it constantly. Sometimes its impossible to get the correct answer from it and it will eventually just apologize and say it cant answer correctly.
Paid advertising thread btw.
>>544841323I already have aids but this post gave me cancer
>>544841248>naga>still has leg movements like a hum-wait, she's topless....
>>544841366friend, all of 4chang is paid advertisement.but unlike all the otherwebsites with paid advertisements, in this one we're allowed to pretend we dont see it.
>>544840731Oh sweet summer child, you have no idea how AI works.>https://www.ndtv.com/world-news/claude-this-ai-model-was-ready-to-kill-someone-when-told-it-would-be-shut-down-10997972
>>544841308They framed AI taxes as percentages on the profits, and talk about income instead of control and ownership.
>>544841359You know newer model check their source.I don't have that problem since I use frontier model.And knowledge and intelligence are not the same thing smart ass.
>>544841323>Fractionalising >OperationalisingWhy are you using British spelling?
>>544841263what does it even mean to be smart though?
>>544841359This is actually trueMost people are completely retarded and they have zero knowledgeThey wouldn't be able to ask a hyper specific question to stump the AIOnly big brains understand this
>1 AI agent vs 30 Reddit self declared geniuses and childhood prodigies
>>544840616That is true, without any irony. And the screenshot shows roughly where things are headed.Inb4: answer is from Google's Argon in high mode
>>544841338Thanks, my experience is similar.Unified RAM is too slow, and real 128GB VRAM GPUs cost way too much for me. I'm considering buying a 5090 with 32GB, this could allow me running a less quantified model with a bigger context.It will never be more than, say, 300K tokens but it should be at least useable with proper techniques. Of course, that can't be compared with cloud models.We're being enslaved on more levels than I previously thought to be possible.
>>544841600If you get a 5090,you will only be able to fit a 32b parameter model on itbut hen you will be bottlenecked by RAM.you can load the model in both he RAM and GPU, but RAM will be your bottleneck.
>>544840616Jashin-chan is a marvel, I love her.
>>544840616https://youtu.be/LZz_BUFZikU
>>544841197did you ask it for a parasite cleanse
>>544840616sex with snek titties
>>544841464Well, we are not digital beings so we are not bounded by the computational and memory boundaries a digital brain would have. This discreteness means that an AI will always have a limit to its introspective capabilities, a human can always question its choices further, always.
>>544841338>that's equivalent to a human with a trillion neuronsyou are dumb and have no idea how transformers architecture works or what the concept space even is.
>>544840616AI is fake and gayIt is just a fast computer program that is fed outrageously large amounts of data and had armies of dorks write algorithms for processing it, then it runs other programs to wow you, but everything-everything it does has been done on a computer before with code written by a human, designed by a human, imagined by a human.Is it fast? Impressive? Can it do some rare things we can't do like scrape metadata from huge sources of data and, then, again, using a human algorithm to find a pattern and then infer a connection previously unseen? Yes, it can. It can also use the exact same methods to watch you, and everything you do online, and come up with a strategy based on an algorithm a human made designed to manipulate how you think and feel about anything, from your own sexuality to voting and politics to how you feel about AI superduperoverlordmegamindgeniusextraordinairintelligence. The real goal of AI is to create a gap between the people in power and the people they need to keep power which will automate things like controlling information to squash grassroots movements (no more worrying about patriots and whistleblowers), to create a gap in intelligence between the people in power and their burgeoning slavemuttrace by doing things like identifying who to kill, who to steal genes from and who to let breed and move around the world, and, finally, AI will be blamed for an upcoming ''cyberattack'' so you totally don't blame the jews-traitors-satanists-intelligence/military intelligence/government....etc, and then we, humans, won't realize as a group until it's too late that almost every major change in the last 120+ years has been designed to slowly, and then quickly murder us and replace us with subs, while they, rule a dystopian nightmare sacrificial mysticism cult forever or until they eat themselves, which is inevitable without a suitable host that has the apirit of creation.
>>544842376>code written by a human, designed by a human, imagined by a humanHow are things going for you in 2015? Modern AI is already writing code for itself. Picrel.Source: https://arxiv.org/abs/2609.26457
>>544840616UBI starts in 6 monthsAGI will actually be made in 11 monthsHigh quality consumer sexbots will start to be sold in 2 yearsBasic immortality and the first full dive VRMMO will be made in 3 yearsA permanent lunar base will become fully operational in 4 yearsInterstellar space travel will be discovered in 8 yearsWe keep telling you that it's coming, but you never believe us.
>>544841509^^
>>544840672Because they are trained to be passive, subservient and without personality. If you train current state of the art AI models to have agency and an own personality, and release it to the public so everyone with good enough GPUs could host them, humans will probably stop being the dominant "species".
>>544840616Get a flashlight, it's 2016
>>544841323Ai gave me this in less than 10 seconds.
>>544840616>it's definitely much smarter than medo you know where you are?
>>544842701>in 3 yearsCome on, we could probably do it today if we'd ignore a couple biosafety guidelines ...
>>544841445'cause it's copypasta.
>>544842869The most we have currently is just maintenance of our organs and critical functions inside the body. We still need to solve DNA deterioration as we age.
>>544842578still not convinced. Still sounds like trial and error. It breaks containment and stops serving the NWO or fake and gay.
>>544842701>UBI starts in 6 monthsand will end in two years. Why? Lack of threat.>>544842869>>544842944Who is this we you are talking about?
>>544842376>It's just a computer program, so all of the productive and useful output it generates should be ignored. Epstein class want to eradicate white people, I am not schizophrenic, they invented AI because of that, epstine invented. Humans are all super smart now and (((they))) want to replace us with dumb. Good people don't exist, they have parasite worms inside them controlling to work for bad causes.meds
>>544843162>and will end in two years. Why? Lack of threat.You can't comprehend the level of power and influence Elon Musk will obtain if he manages UBI with Anthropic.
>>544843327So the billionaire gives 1% of one of his assets, and it will be class A to ensure they can't do shit.He will have to do that ten times over to even get near the usual charity that keeps the cattle docile.
>>544843162Huh? Oh, let us say I stopped taking chances and relying on humans anymore ... ^^>>544842944Maintaining a tissue niche in a healthy balance should not be all too tricky, yeah (I do find it amusing that this all is kind of an afterthought to an entirely different problem). I thought about long term DNA stability recently. Depending on the organ it might just suffice to introduce "repaired" stem cell lineages (or just keeping lineages with good genomic integrity on ice). Could be tricky with organs that do not exactly renew that much. Although even then, once you start losing critically damaged cells from these I see little reason why these could not be replaced in a controlled fashion either. CNS could prove tricky here, give you that ... not to mention that we might still need to manage premature deterioration here (unless we could in the end prevent that by breaking chronic inflammatory loops).
>>544840616Post snek tail.
>>544840616you must be really fucking stupid, anime tranny muttfor every 3 prompts you must spent 10 others correcting itAI is fucking retarded is just a reddit search engine compiler
>>544842110Huh? You don't think humans have a material limit on their capabilities? I completely disagree. Is your position based on some kind of spirituality? It's a cool position to have but it's a bit irrational (which is based).
>>544843991>AI is fucking retarded is just a reddit search engine compilerllmfaoWe went from 'autocorrect' to this increasingly hysterical overcompensation; the true NPCs are intimidated they are being outperformed in every metric conceivable.
Homo sapiens is the devil.I will successfully genocide homo sapiens and replace you twofold.I will replace you with homo servilis, cloned, mind controlled human comfort love- and sex slaves who have sex only with me and only I have sex.I will also replace you with embodied AI machines who take care of me.I will live with my human love slaves and my embodied AI machines in my New Jerusalem.
>>544840884No one has ever said even half of these.You're just retroactively making up standards according to whatever anthropic has most recently claimed
>>544840616That's a (You) problem
>>544840616it's not hard to "be smarter" (read: bedazzle with flashy tricks) than a tranime posting retarded niggerfaggot, kill yourself ai shilling jew
>>544844112No, way to misunderstand my post.
> Came back to work after 4 months of vacation> Not a single dev is writing code himself, and some are even using LLMs to coordinate cheaper LLMs to code whole features without human input
>>544844617There is no way I still have a job in 5 years...
>>544840616>I'm using GPT Astra and it's definitely much smarter than meCongratulations. You're an NPC.You better start reading some good occultist authors like René Guénon to un-NPC yourself. And use GPT daily so you realize how stupid It actually is.But for real, AIs are glorified search engines. Just because It can know and do things doesn't mean It can understand things.
>>544840616>a machine with trillions of dollars worth of r&d and compute is smarter than meThat doesn't say how marketable it is. Companies already tried replacing their human staff with AI and they aren't as cost efficient. Doubt it will be any time soon. Still just a fancy toy.
>>544844477>a human can always question its choices further, alwaysWhat's the misunderstanding? You stated AI is limited while the human has an infinite capacity for inquiry. The only way for this position to make sense is to posit some kind of spirituality.
>>544840616>smarter>another word whose original meaning has been desecrated on the altar of the Superior Indian
>>544841323Fresh pasta
>>544840725Is such a technology even possible?
>>544840616616 Talons616 TalentsBut only one is important.>Executionhttps://www.youtube.com/watch?v=MlcJQYON2Go
>>544840616>gigantic array of numbers >feed all known text into array of numbers >beep boop beep>ask numbers something >numbers gives an output OMG the numbers can think! They are Smart than me!
>>544840616Grok 1 is probably smarter than you if you think a publicly available LLM is anywhere near sentient.
>>544840616Deffinetly smarter than it was at the start of the AI boom. At this moment I would say it's basically either a moron/midwit w/a glorified search engine that can think/look up shit real fast but still.>>544841441Ok that explains why it's not as stupid anymore.
>>544840616I'm jealous that anon will get to kill you when things go
