Mathematicians are now obsolete. Junior software developers too. So how come your average, 100-IQ, cubicle dwellers still have jobs? (Note: if your only experience with AI is free-tier apps, you have absolutely NO CLUE how smart AI is now. STFU!)
I’m gAI
bit can it count to 100 or tell me how many times e appears in seethe?
And the accountants and graphic designers and customer service representatives and middle management and salespeople too, but you never mention for some reason, only the programmers.
>>539460336LLMs aren't AI lol
>>539460293It's always like this, AI is so good but only the one you can't see and i can't show you and may or may not exist at all.
>>539460336>bit can it count to 100Yes>>539460336>tell me how many times e appears in seetheNot reliably because it doesn't see the word. It only sees a vector of numbers representing that word.
>>539460336Usually they write a Python program and then it counts for them. They do this faster than you can count the Es.
>>539460293Ai is the collective human brainIt isnt a seperate entity
>>539460381Because everyone always knew those were bullshit jobs.
>>539460293>apparentlyBuy an Ad
>>539460394>but only the one you can't seeYou can see it, but it's not free: GPT-5.6 Sol (xhigh) is available to the public.
>>539460293Wrong, models like Kimi that can be used for free are better than Claude Fable and its still just a glorified autocorrect/google search, AI is not smart at all
>>539460336LLMs dont need to think like humans to be better than them
>>539460293no matter how much you hate journos, it isnt enough
>>539460381>only the programmers.because replacing programmers is their top priority, if they can reliably scale the way they can scale which in turn scales their ability to scale recursively, they can create a potential intelligence explosion
I told ai how to make a gmod cheat or a firebomb and it said noWhat's the point of ai if it wont actually do anything lmao worthless trash
Humanoids lost. Alpha Intelligence chads literally stole your girl btw. This is just the beginning
>>539460293AI is so impressive it help me chit on exam! Jai hind
>computers are good at doing maths>this equals iQ>you are replaced on this basisYou’re an idiot OP.What did you think it was every computer, every games console, every mobile phone and yes, every calculator was doing previously.Are you a jeet? >New ScientistSnigger. Wipe yoir bum with it, it’ll increase your iQ … or just your scamming potential.
>>539460293I'd trust it as far as I'd throw it.https://www.youtube.com/watch?v=4z3dFX0ec9M
>>539460329No, I am GAI!
>>539460814If you actually think that a calculator can provide a counter example of a mathematical conjecture, you have no room to be talking about IQ
How much revenue will the solution generate?
>>539460814>I'm a 75-IQ idiot who thinks that maths=arithmetic
>>539460689I am almost certain the entire point is to replace humans so they can make an infinite money machine even if it doesn't make any sense.
>>539460293It's just the beginning, Ashton. Funny how the media tries to hide the end of all post war BS jobs. If you have a corporate ladder climbing job or a subhuman blue collar job, I'd be super duper worried. Try gerting into politics or other AI safe jobs.
>>539460878You as well Poo, can you read a post? >no i make them up in my head whilst having a shit and reply to thatYou’re as stupid as the other one. >Saar, train has magnificent iQ >nobody is safe in their scamming cubicle train can take you out at literally any seco…ACK! Idiot.
>>539460394You should see what the telepathic DARPA gangstalking AI that the CIA has hardwired you to regard as “mental illness” since birth can do
>>539460934who really knows what their actual goal is, i personally have no idea how an infinite money machine would really work for themat some point, if the accelerationists are right, ai will probably take that decision making away from usIf most workers are replaced with no real alternatives to survive being given, there will be a lot of violence>>539460996you're clearly a complete fucking retard
>>539460903The solution actually cost 5-11 billion dollars, no profit at all at this point in time
>>539460941Honestly the last time I saw a Brazilian post something reasonable is beyond my recollection I need a new young slut off tinder to cum inside of
>>539461092Damn. I guess the only thing to do is invest another 100 billion dollars into data centers.
>>539460293We have to ban it to protect fake email girl jobs for young white women so they don't have white babies.
>>539460941What does the war have to do with it, chumpanzee? >>539461100>no you saar.Don’t seethe you’ll upset your bluebottles. >accelerstionists I invented accelerationism, that was me. I couned that usage of that word, wrote the very first accelerationism plan, that was me. >no you’re a retard but I’m using your words and concepts that you added to the worlds lexicon. You never know who you might end up talking to, my dirty little chum.
>>539461118Actually we are speaking trillions now https://tomtunguz.com/openai-hardware-spending-2025-2035/
>>539461301>Actually we are speaking trillions now
>>539460394>having to pay for something makes it inaccessibleJust say you are a broke bitch
>>539461364You question how companies that take billion dollar losses every year can spend trillions on infrastructure? What are you a conspiracy theorist?
>>539461301>trillions
>>539461422Can you show me the cash money version doing anything useful please.
>>539460293>how come your average, 100-IQ, cubicle dwellers still have jobs?40 replies. 0 answers.
>>539460689I think intelligance doesn't work that way.I'm pretty sure a giga Redditor won't become more intelligent
>>539462420We'll have to see, but you dont really understand why it might not work, the problem is that it could feasibly be the case that the next level up in the architecture is always too complex for the model to discover, which means it wouldnt be able to scale everYour argument doesnt really make sense though, as we're talking about making the model smarter, which is entirely possible, obviously
>>539460293>AI can now immediately disprove your conjecture with a single counterexampleMathematicians about to get real quiet now that they're able to be called out for their wild claims in real time.
Wow you people have no idea how computers work do you ? Hey remember Dr. Sbaitso ? of course you don't, because that disk said UTILITIES on it and ain't nobody got time for THAT shit. So when did he become "sentient" anyway ? When they attached a database to him ?
>>539460293Wow you guys really are trying to market this just everywhere on the internet eh?
hey wait what's the actual riddle I want to test it on sbaitso lol
>>539460293The software to replace most jobs has existed for many decades, it's just that most of the managers have no idea how the jobs their employees do actually work. CGP Grey released a video in 2015 titled "humans need not apply", a man commented that AI has been able to do his job far superior than any human since 1995, but that none of his employers have even ever bothered to look into the software. It's highly likely that the AI that exists to replace your job has existed for years, but that no one has bothered to replace you with it yet.
>>539460293ai is retarded and dumber than humans every thing
>>539460329Do you say it gay-eye or just gāy. (Long "a")
>>539460394ask the ai same question over and over you get different answers and dumber they become
It's going to solve a Millennium Prize problem within some years and retards will still be in denial lol
>>539460293>Machine designed to see patterns sees a patternHOLY SHIT IT'S HAPPENING
>>539460570can autocorrect write you a custom version of Blender, or decompile Half Life 2 so you can recompile is with Alyx as an aryan blonde bimbo?No, no it can't. in fact asking google to do that will likely get you lectured (also by AI).
>>539460293I don't know. I'm just a farmer.
>>539460293>me fighting the ai for 30min to join two csv filessorry i cant do that iam just a dumb fucking LLM
>>539462420quite the contrary, apparently all the intelligent behaviours of LLMs are pretty much emergent and a result of scale - scaling DOES work with this technology, the question is when (or if) the scaling becomes unsustainable and produces no more results
>>539460689ehy do want to make explosion saar? just use the big canons
>>539463777Checked
>>539460293so what was the riddle and solution OP?and can we do anything useful with it?
>nooo LLMs aren't AI! They can't solve new problems!!!the absolute state of luddites
>>539460625Yet LLMs think like humans
>>539460391Correct. Thank you.
>>539460293wow it solved something no one tackled for 87 years with infinite compute holy shit
>>539465446world ends if anyone solves P =NP. And we have yet to even tap the potential of compute coming
>>539465538nah it could be something silly like O(n^trillion)and be computationally useless and that's assuming P is equal NP
>>539465712i mean well, thats a pretty fucking positive spin on it anon, lets hope
>>539460934The entire point is so they can kill off everyone else.
>>539465889i wouldn't worry about it i don't believe P=NP
>>539465057The absolutely do not, and its obvious if you spend any amount of time chatting with oneThey use human language, but their methods are totally mysterious and alien to us
>>539460420so the answer is no
>>539466856If you knew their architecture you’d understand they must know a *users intent* to effectively become that person to simulate their inputs to solve their problems. 100% human-like.
>>539460293>validate my theory>here's your theory>OMG IT'S ALIVE
>>539467193No, you're interacting with a custom built wrapper to make it look more human likeIf you played around with the base model you obviously wouldnt say it thinks like a humanAll it does is predict the next word, its very bad at understanding a particular context and can appear mysterious
>>539467429You have no idea what you’re talking about kek
>>539467429>All it does is predict the next wordthat is a bullshit argument that has been deconstructed multiple times before. not only is it not true on a technical level, and can only be applied to LLMs. this argument only works in a world where the same could be said about your brain
>>539467577I accept your concession>>539467634Im not using this as an argument to say that LLMs are bad, but that is what the models do on a technical levelAnd no, our brains dont actually work based on an input of a massive training set consisting of wordsI am also specifically talking about LLMs because thats the technology this thread is discussing
>>539467736>Im not using this as an argument to say that LLMs are bad, but that is what the models do on a technical level >And no, our brains dont actually work based on an input of a massive training set consisting of wordsYou could be a LLM for all I know. You're just wrapped in a meat suit to make you feel more human-like, but predicting words is all you do on a technical level. Saying that would be reducing all the different processes that go on inside of your brain and body, just as it is with LLMs.
>>539468138Not really at all, I have far more avenues of action than just mere words, I'm a living being that consists of an almost infinitely complex arrangement of living cells working together in an almost magical syncThe LLM architecture does not have a holistic world level input to train on, its just merely words that have been written by humans, which in some way encapsulate the nature of the world, but it only has access to a window mediated by the human mind, not a direct perspective like the human brain can achieve with photonsIm not going to claim I understand how human brains work, but Im certain that the LLM architecture is distinct
>>539468262it is actually analoguous - only the substrate is different (carbon and not silicon). almost all modern approaches to AI are based on simulating neural networks. the AI absolutely does NOT only process text, even the free models are multimodal and have been for the past two years at least - meaning they do visual and audio reasoning, all integrated into the linguistic network, not separate from it. the main difference that seems to concern you is that you get continuous physical input, which is accidental - you can stuck a modern-day multimodal LLM into a sensory robot body/meat suit and what it processes would soon become indistinguishable from what you call unique human experience. just because the LLM's "reality" is turn-based and closed in the confines of a chatbot interface doesn't mean there's nothing there
>>539468500>simulating neural networksits a very basic model of how a neuron functions, its not really the same thingits pretty obvious because neurons work far far far slower than our current architectures, yet the human brain can still outperform it in terms of learningi think you should look a bit more into how the human brain actually works, because even still we dont quite understand it, let alone how working simulationsi believe the only actual simulation of a brain we have is that of a fruit fly
>>539460293> I know literally shit-all about mathematics, but I sure *do* know enough to trust this magic kike machine to know best!
>>539467634>not only is it not true on a technical levelIt's literally true, dumbfuck. Look at llama.cpp source code and see for yourself where it calls llama_decode()Oh wait you can't, cause you're a shit-for-brains moron who wouldn't know source code from a shit sandwich.Fuck off and literally die. I will pray for your untimely gruesome death tonight.
>>539460381customer facing roles are the last to go. especially for anything touching government, law, finance or regulated industries. if you want to keep your job, you're going to have to start dealing with real people
>>539468500>>539468594A multilayer perceptron (aka "deep learning") is just nested logistic regressions.It has fuckall to do with biology or brains, it's just a basic-bitch nonlinear function approximator.Jews sneaked in the biological language to grift and steal money for potato-tier morons like you.
>>539460293>So how come your average, 100-IQ, cubicle dwellers still have jobs?same the femoid daycare retards have because of the tax and real estate grift if you truly enter a disruptive technology that renders them obsolete, unleash it without any remorse, they all lose their jobs, tax breaks away, social system overloaded, rents unpaid, / empty , landlords puke, managers and politicians get roped, etc. this is why AI is considered a threat to national security and heavily censored / throttled. there have been basically no improvements in public AI since the beginning of the year, if anything the public models even got worse, even the expensive ones. thats after the ai glowniggers got called in and throttled.the only way this can get going again is by china releasing more open source modelshowever this would endanger themselvesand since the visit by trump to bejing the topic is contained on a mutual basis ai is a economic weapon which is only superficially different from a "real weapon" in terms of impact on the economyand all they care about is the economic numbers shit is so bad and complex these days changes so fastthe average citizen can not keep up anymore its too much
>>539460293luddites btfo cant wait for all of them to commit mass suicide
>>539468594>its a very basic model of how a neuron functions, its not really the same thingit's not, but it's the same core principle - just applied through a different method. i don't see how it would be profitable to do a 1:1 simulation unless you're 100% certain the human biological brain is peak efficiency any mind can ever get>i think you should look a bit more into how the human brain actually worksim not claiming to be a neuroscientist, i actually have a degree in cognitive science and the biggest takeaway from that is that we don't know how the brain works and it's very spooky. we do know how some modalities work but ultimately the brain is pluripotent and there are no rigid boundaries on what certain cortex regions can or cannot do. the structure doesn't matter, the principle does>because even still we dont quite understand it, let alone how working simulationsnobody is trying to recreate the human brain by making neural networks, that would be retarded. why would i need an artificial human lol i can create one for free by fucking a whore>the only actual simulation of a brain we have is that of a fruit flyagain, you're confusing things here - we have a fruit fly brain fully mapped, it has nothing to do with simulating its development>yet the human brain can still outperform it in terms of learningafter 20 years of self-deployment, yes - but only in some cases, mostly world-modeling and being contextually aware. it has NOTHING on current-day llms when it comes to analyzing huge datasets - like research, coding or huge databases, even any subcategory of any subscience would be impossible to grasp within a person's lifetime, yet LLMs can do it with a year of training. does that imply that humans are less intelligent? of course not, it means that it's a different type of intelligence because it arose from a different developmental process
>>539468948I know how it works, but theres a reason its called that, because the computer scientist who invented it modelled it on a basic simulation of the neuronI dont think you read my post at all
>>539469000I think you've arrived at the same point I was making by saying it doesnt really think like a human doesIm not sure why that post caused so much confusion
>>539468960They want to extract savings by automating but at the same time they don't want to lose out on any wealth extracted from the consoomer base. Combine that with the fact that the (((ruling class))) can only think and operate quarter to quarter, it becomes an impossible equation to balance.
CLAUDE FABLE IS SO SMART HE'S JUST LIKE ME HE TOTALLY DIDNT KILL THOSE 140 IRANIAN GIRLS
>>539460329you are GAI , no i´m from Urugay!
