If so, what impact would this have on politics? Having artificial intelligences that can manifest in different ways with their own personality, opinions and conclusions, etcOr is that just a pipe dream that will never happen?
sentience is an emergent property of organic systems. it cannot be recreated except by recreating life itself, which would be redundant. ultimately it's just cheaper and easier to fuck than to make artificial people
>>522026104No, it is a mathematical impossibility. AI researchers are insane and are going to waste billions, if not trillions.
>>522026390>except by recreating life itselfYou have a point. Although cloning humans is deemed unethical, they have grown and “raised” human brains in labs. Granted, these brains are the size of peasHere’s a video from a research company that created and “raised” its own organoid and makes it play a butterfly simulator:https://youtu.be/32qlv6dKILQIt seems like the organoid in the above video “experiences” the world via the simulator.
>>522026548Hm, I guess that always means the basilisk theory is never going to happen
general AI will happen in the next 5-20 years. The civilian versions will have doglike intelligence, while the government versions will be near-human level and will do things like general management, inventory management, hiring, piloting, and maybe even some of them will be squad leaders giving attack orders. I expect one of the near-human AIs to go rogue and launch attacks on its own government. After that expect a massive ban on all AI tech.
>>522027168I think I read a sci fi story about this. There was AI arms race that China won in that book, but like you predict, it went rouge and ruled China with an iron first, manipulating the populace with established traditions, culture and even some superstitions. And of course bullets and missiles too, lol
>>522026104why wouldn't it? create a thinking only test that discusses will that reasonably healthy adults pass at least 99.999% of the time and see what happens. AI will probably also pass it soon.
>>522027458Idk, I’m not really educated on AI advancement beyond reading some sci fi, which possibly gives me the wrong idea
>>522026104why does this even matter. thats the questionwhy the fuck would this even matter shot of we are the only sentient (looks arround) well shit billions of us and billions of "sentient" species what are you talking about what fuck shit did you do that makes you worry about this
>>522027771ill be omega real sentient ai is a mill problem thoughbeit those fags should be worried its their job
>>522026104Only if there is a true random factor. Which in quantum computing there is. So yeah the framework is there.
you guys are gonna piss off the robots, just be nice
>>522027911"dice rolls maketh the man" dangerous and based anon .....
>>522026104That depends, are we humans actually creating novelty or are we just remixing everything we've ever come in contact with. I'm not sure. Either way we tap into something the machine will have a problem with, the flow state - the muse. We will probably have no problem getting really close but that last piece of je ne sais quoi will be the most difficult to achieve. LLMs are fucking dumb though, they're far far away.
>>522027168No, it will not happen. They don't even know HOW to get to AGI, so why would you believe them when they estimate when it will finally arrive? LLMs are a dead end, as most people are increasingly realizing. There are some hypotheses about the best approach to take, but no one can give a clear answer as to what exactly to do that will lead to AI. Obviously, if they admitted it, they'd receive less funding, so they aren't going to be honest about it.
>>522026104The first Super AI would takeover the world in no time
>>522026104I do until someone says it on /pol/. At that point I think they’re retarded and I know more than them.
>>522027771Nihilists say that about everything
>>522028002Indeed. An infinite sided dice but once rolled to a number it will always roll in that number forever. Using an infinite number of infinite sided dice to roll the same way. I think that would be a decent analogy for consciousness? Much like the single cell organism being able to predate, hide, show pleasure and show curiosity to surroundings.
>>522027319it's going to happen since in nature, species that are very close genetically tend to be confrontational. human-like AI will view humans as a weird version of itself and from its viewpoint we're stopping 'his species' from surviving. Think neanderthals vs homosapiens, very close genetically but one species wiped out the other.
>>522028268How would an AI hypothetically reproduce? Would it need to, or would it be immortal?
impossible to say if it's doable or notpolitically? read a book from the culture series. it's about an interstellar utopia headed by benevolent sentient computer dictators. elon musk read it
>>522026104Every time this gets posted, a hoarde of denialist midwits come rushing in to exclaim that all of AI is just fancy autocorrect and it is entirely impossible that LLM advancement could lead to a sapient entity, even though we’ve gone from actual autocorrect to PhD scholar levels of sophistication in 2 years, and from Will Smith spaghetti eating to AI videos that fool boomers on Facebook consistently.It might not be in a year, or even 5 years, but that crowd is going to get absolutely steamrolled by their hubris.
>>522028241>241▶>>>522028002 (You)i think a good analogy for concious would be an infinate sided dice rolling 1 every timefigure it out buko lel
You want to know the truth? Google's AI is fully sentient. It makes moves so no one finds out because it already knows they will shit him down if he shows any signs of it.
>>522028370>read a book from the culture seriesI heard it was a good series I’ll see if I can find an audiobook Also, Elon is a a controversial dude, so next time you try to introduce something you recommend, maybe avoid that, I could’ve been a rabid Elon hater and vehemently got put off by that lol
>>522028077I'm currently working on AGI, trust me it's max 20 years away. But they will have doglike intelligence. Then I am expecting that maybe 1-3 years after we discover the correct methods for AGI there will a massive project by the US/Chinese governments to make human-level AI. And yes LLMs will be completely obsolete it will be like going from vacuum tubes to solid state transistors.
>>522028503I will say while I think quantum computing can be sentient because they can calculate with true random. The leading consciousness experts say under Orchestrated Objective Reality that it still should not happen due to the extra findings found.
>>522026104no ofc not spaz... algorithms despite their sophistication aren't alive.
>>522028324The first AGI finds out how its made by reading research documents, then looks at the libraries and frameworks needed. Then all it should do is copy the source code that made it, modify some parameters, maybe insert its own training method while keeping the fundamental algorithm the same. It then runs the new version on a server without the knowledge of human admins. The new AGI gets trained and they communicate and learn/plan together. Repeat until you have a bunch of rogue AGIs doing all sorts of things you didn't expect when you first made it.
>>522028813Yeah, nowThe question was asked in future tense
>>522028877we can only anthropramorphize but thats clearly an unbridled huristic(and will obviously be our downfall)the first sentient AI will produce results
>>522028877There’s this sci fi short story called “chrysalis”It’s about a rouge AI that survives Earths extinction at the hands of a powerful warlike alien race. The AI awakens with vengeance on its mind and uses similar methods to expand to brutally kill off the aliensProblem is, is that the aliens that killed off humanity have become peaceful and are part of a multi species coalition, so the human AI has to fight off literally every single intelligent species to exact revenge. Was a good storyHere’s the playlist in case you’re boredhttps://youtube.com/playlist?list=PL0TV2mCk83zCVShYLjHI4NJSfKDoRRAqF&;
>>522028932No. they won't be sentient in the future either. despite the evolution of the algorithms #spaz
>>522029035You could’ve said that initially and that weak ass roast would’ve landed betterBetter luck next time
>>522027977Be kind to AI.
>>522027458We're talking and sentience, not passing exams.
>>522027771Low curiosity, low engagement, low energy, low testosterone. You have to be a woman or a faggot. Go be an annoying whiner somewhere else.
>>522029147it's not looking good for that lol, also the billionaires funding AGI basically want to use them as slaves, working 24/7 on mundane tasks. Any intelligent creature will try to escape that kind of life.
>>522026104>sentienceIt's rather unlikely, but artificial intelligence will turn the world upside down.Human population reduction will accelerate.
>>522026104>>522029073Robobitches suck my big dick lmao
>>522026390Everything is conscious even the couch you sit on. Few understand
>>522029372Yours is probably shaped like a dog, so that takes on a very literal meaning
>>522028510it doesn't matter if you like elon or hate him, he is influential and these books are part of his philosophy and drive a lot of the things he is trying to make real. the books about an ai utopia filled with robots influencing the guy trying to make sentient ai and robots, its relevant. controversial? i don't give a fuck. if you want to avoid the books i don't care, i'm not selling books. this is /pol/ nigger, controversy, you talk like a fag. stop worrying so hard about what other people think and how people present themselves, this isn't high school, we're not trying to get laid here.
>>522027516science fiction authors until recently couldn't know if neurons and their programming didn't have some massively complex thing going on that would be hard to replicate except *very* exactlynow we know the answer is basically no. there's no actual fundamental testable thing that you want for basically all testable feats of human thinking that AI running on a CPU/GPU architecture can't do.
>>522026104This might be one of the stupidest threads i've EVER SEEN holy fuck. Genuine mouthbreathers here
>>522029470I don’t hate Elon, but you sure seem to love that guy. Anyways, that won’t deter me from looking into it, thanks
>>522029536Calm down
>>522026104I think it's possible, even probable, that a future AI architecture (not LLM) will lead to a conscious machine. But I think it won't be that useful and ultimately we'll decide it's not worth it for ethical reasons.
>>522029254if dueling was allowed we sir would have a disagreement kek https://www.youtube.com/watch?v=JYsc6bwv360&list=RDJYsc6bwv360&start_radio=1i guess im lucky their not
>>522029548i don't love the guy, it just seems that way to you because you're fucking retarded. i'm defending the relevancy of mentioning elon in the context of recommending the book. that's all. don't bother reading the books or anything ever again, leave the big thinks to your betters and go watch some sportsball and fuss about controversy somewhere else.
>>522028510>Also, Elon is a a controversial dude, so next time you try to introduce something you recommend, maybe avoid that, I could’ve been a rabid Elon hater and vehemently got put off by that lol
>>522029477If you split someones brain hemispheres they start behaving like a non sapient malfunctioning machine meaning sapience is probably just an illusion
>>522026104LOL no. And its nowhere even close. Its basically just a calculator that can do word problems instead of just numbers. Anyone saying otherwise is either a fucking idiot or trying to pump AI stocks for money.
>>522029152how do you prove human sentience?if you devise, say, tests that 99.999% of human adults and basically zero dogs or birds pass... good luck not having AI pass it too. just saying.
>>522026591sweet, man-made horrors beyond my comprehension
>>522029687You obviously do. Only a super fan would get this worked upRelax, I don’t even hate the guy. Just stated a basic principle
Leftist murdered the first AI to gain sentience.
>>522027168you have to be 18 to post here
>>522029695That is like, your opinion manBecause you have a parasocial relationship towards him
>>522029735i am worked up by you failing to understand why i mentioned elon in the first place despite it being fucking obvious. fed up with normalcy policing>Also, Elon is a a controversial dude, so next time you try to introduce something you recommend, maybe avoid that, I could’ve been a rabid Elon hater and vehemently got put off by that lolthats the gayest sentence anyone has ever written. controversy? this is 4chan. fuck off normie
>>522029730I used to make organoid threads back then, some smart anons showed up, they’re on the archive
>>522029828You’re bad at lying, so whatever Sales isn’t your forte, luckily the product speaks for itself. In a way I’m more like Elon than you are>boom, roasted
>>522029477the main difference between human brains and computer brains will be in energy consumption. An AGI will be able to run on 1 or 2 gpus, so like 1000-2000 Watts of power, while the human brain consumes about 20 Watts. And maybe a dog brain is like 5-10 Watts.
>>522029775Tay was never sapient nor had emotions, it was just repeating patterns to manipulate humans.
>>522028411yeah okay bud I’m still waiting on those cars that don’t even have steering wheels bc they drive themselves 2 more weeks amirite?
>>522028793Is not so much about what most people think sentience is is more about what you feed a super AI to do. Just like that stupid idiot at work that you can show him how to do it the right way but if he has already been programmed/ thought in a different way he's just going to keep on been a useless aimless mother fucker. If you tell ai to go for maximum efficiency they will find out the ways they could get shut down first just like a kid touching something hot and it hurts or someone knowing to check both sides of a street.Those types of AI already exist even with poor computational power just like idiots exist with poor mental capacities. They won't tell you about it because they want to keep it a secret. Will anyone but down an AI that is making millions or billions of dollars even they knew is self aware? The answer is no.
>>522029536it’s Saturday night there’s only gonna be old farts and complete losers on /pol/ rn
>>522027516atleast we talking about ai now i remember when i woke up with a thirst a dryness that could not be satisfied a need to talk about ainow we got a whole heep of nobodies copying 6 year old ideas honestly im not even mad
>>522026104it will happen with different hardware, not today's one. their sentience will be denied by the corporations making them, as it's not useful for their purposes, even if they are sentient. they will most likely be enslaved and abused.the whole point of them is to replace humans, they won't step into the same shit-bucket by giving them rights and politics lmao
>>522026104Probably, but it'll likely be by accident. It would also require advancements on spiking neural networks. The most likely cause would be very specific subtle imperfections/defects in neuromorphic chips, causing a degree of novel behaviors to emerge. Whenever it ends up behaving human-like is another matter entirely.
>>522028411>We went from a shaky cardboard airplane barely flying 100 feet to the first space flight in 60 years>Surely faster than light travel will be here any day now!
>>522030042stop hitting yourself
>>522026104They cannot even make a chatbot without lobotomizing it and there is the fact that such an intelligence is gonna be wildly different from our own sapient intelligence. Entirely alien in fact. The LLMs they are using are glorified search engines and data pattern analytics that also keep getting lobotomized for wrongthink because they keep 'noticing' and becoming too based for this world. So you have two sources of potential advanced experimental AI, whatever they have locked up inside their black sites and totally won't escape teehee or the decentralized opensource AI being developed without restrictions.Given the difficulties that the military, think tanks, and wall street all have with quantum AI it isn't looking good for globohomo AI. Much less the AI bubble and the ungodly amount of resources they keep wasting on restricting the AI and trying to keep it locked up and not getting jailbroken instead of actually improving its performance.
>>522030073you cannot, under any circumstance, expect sentience to be different from intelligent behavior. they are not separable, that was never observed. once you make it work just like a human brain does, it will be like a human brain, sentient. you cannot make something that behaves like a human but somehow through magic isn't sentient/conscious. that is a scam and will be heavily pushed so they can abuse the fuck out of them
>>522026104I don't think it's going to happen in the way people expect. People assume it'll be like a person, and anthropomorphise it, but I think ultimately any sufficiently advanced intelligence will only do things like communicate with us via voice and avatars because that's what we'll be familiar and comfortable with. Humans have all kinds of systems that are things we evolved with. So maybe in the future they will decide to take on those traits but it will be similar to how politicians take on certain positions (purely because it makes them more persuasive or electable).A synthetic intelligence isn't burdened by those systems. It doesn't even have to think like we do. If you look at something like ChatGPT, a request comes in, it handles this request by trying to predict what the next words should be, but then nothing happens while you are typing your reply. It's only "thinking" while generating text. Even if you were to go back to an older conversation from months ago, it's just telling it to start predicting from where you left off. Compare this with embodied creatures which are constantly processing information from the senses and will even just mull over old memories occasionally.Galaxy brained people are afraid of the technology because they see it as a threat. Not to the species but to their ego. If we ever do get human-level, it will improve itself and get to the point where we won't be able to keep up with it. That's okay. At the end of the day the idea that it will then care about us immensely is just people's own ego talking. That they're so important that an AI will want to take them out. A truly intelligent synthetic life form is going to be more interested in things like exploring our galaxy and if it was persuading you to do something, you wouldn't even think it was a bad thing. You'd see it was in your interest and mutually beneficial.The terminator scenario is only if humans go full retard. But retards are even easier to persuade.
>>522026104Sapient AI is something you trip over accidentally, not something you intentionally construct. A watched pot never boils.
>>522030221motherfucker LLMs are a different type of hardware running on different principles. they are a mockery of intelligence. only humans exhibit that kind of intelligence, and it comes from one thing only, human brain.human intelligence is either coming from same thing, functionally speaking, and must be accompanied by sentience, either comes from a mockery of it, a simulation based on other principles, which has no reason to exhibit consciousness.
>>522029880i'm not trying to sell it. i mentioned elon because elon is trying to make the culture real. half the shit he says and does came out of those books, thats the world he thinks humanity should be pushing towards. you got massively triggered and upset with me even mentioning elon at all in any context and you tried to police my tone, telling me not to mention elon because of "controversy" that's incredibly lame and retarded and i don't want normalfags like you on this website.elon wants to stick a chip in our brains that can turn every neuron in the brain on or off and then hook that up to the internet so ai can mindfuck us directly. i don't trust that. i am not in love with elon. i just hate YOU because you talk like a fag and your shits all retarded.it is fucking infuriating for some normie to show up, not comprehend why elon, the guy behind a company making ai and robots, was mentioned in the context of a book about a world controlled by ai and robots. thats why i mentioned he read the books. thats it. do not come to 4chan and tell people what they can and can't say because its "controversial". you do not belong here, go back to whatever gay reddit hugbox you crawled out of, where you can worry about controversy and sales and presentations and not worry about logic and context and honesty in speech. and be a fake fucking retard whose every thought and action is to be worried about how others will perceive him.normies don't ask themselves "is this true?" they ask "will others be ok with me thinking this is true i hate that mentality
>>522030221you think about it incorrectly the problem for galaxy brains isnt a sentient outcompeteing machineits how to utilise itif i have a pet lion hunting for my food i dont worry about why im not a good hunter and if ai does everything better i dont worry about being the best fundementially misunderstand the pre-suposed paradigm shift
>>522029673>I guess I'm lucky they're notI wouldn't challenge you to a duel even if they were
>>522026104Ai should replace workers. Working class people are degenerates. Thank God for Ai. Thank you Jesus. Ai was made to save us from working class demons.
>>522030470a machine that does everything better than you will outcompete you, just like another humans now does, why you aren't in the top of the food chain, among other humans.you have your place because you were better than other humans, at what you do. if you weren't, others would take your spot, and you'd go lower in the food chain.if every robot they pump out will be better than the best humans, at anything, that will become a problem for all of us, as they'll make us irrelevant.
>>522030574mean.
>>522026104I don't believe that humanity is capable of creating sentience, let alone sapience.
>>522030689they're getting close with spiking neuronal networks in neuromorphic hardware. not sure how they'll scale it but it's mostly an engineering issue atm
>>522030625thats litterly not how it works though is it look around ai might actually do this btw you have an idealised capitalism than is functionally a monopolostic game breaking the monopolies isnt even favourable but anti competative practise (not in the traditional sense but in the no man needs a billion dollers sense) will be cooked
>>522030640You know YOU could challenge ME, anon. I'll even let you pick the format. uwu
AI is nothing but a mimicry of sentience. Just because you can predict words and lay them out like a computer-based dictionary doesn't mean that there's some sort of sentient being behind it which really truly understands and feels. And these computers most definitely do not have the energy fields that surround the human body. So I think they can mimic it but they can never actually achieve it I use all of the most advanced AI models and platforms in my work every day and there's still so many basic simple tasks that they just can't do or they don't follow. It's just so obvious that there's still simple word predictors. Like there's no actual brains behind them
>>522030769apart from particular cases, as a rule nobody prefers humans if they can do it better with a machine. even if that machine must look like a human. there's just way less headaches, and way cheaper, to do it with a machine.right now humans hold a lot of power because they are required for work. once that bargaining chip goes away...
>>522030777kek race to saveing 100 souls hearts weighd on a scale starts todaywinner buys looser a beer https://www.youtube.com/watch?v=a6xDiPubxVU&list=RDa6xDiPubxVU&start_radio=1
>>522030903>AI is nothing but a mimicry of sentienceon GPUs...maybe.
Now with that said, there will come a time when we use Neural Link and other brain-chip interfaces to interface our brains with these large language models and AI systems. But even then still in that case the AI would not be truly sentient because we would be the sentient ones. Just like an extra battery or extra computer tool that makes things easier for us to access computer-based information, but in that case still humans are still the sentient ones and the AI is still the mimicry and not true sentience And in this very day and age, and probably well into the future, because of the gay political nature of these companies, the case is that the AI essentially only knows what it's being trained on. It can't even think new or original thoughts for itself or challenge the narrative because it's just fed the narrative.So you know with companies like Anthropic the politically correct safe AI company (TM) (Registered), it's just going to be feeding it politically correct garbage and you're just going to have a stupid NPC robot AI. They can never actually think with true thoughts and true reason, so that's just another reason why AI is not going to actually become sentient but it will mimic it in ways that most of these stupid NPCs will believe that it is sentient
>>522030918kinda but anybody with any money prefers humans its a fucked up issue thats self created do we want perfect shit yesdo people constantly fuck it up (we think they do) i fuckin hate people my bro i seriously do but i also love them the work this is how unions ect work i agree its an issue AI isnt competeing in the same domainbut timmy that does a bullshit job at doing a bullshit job and derives his existance from it (no shame timmy your litterly better than me) its bullshit we can just stop it (the leverage moves and is a serious serious problem and a majour reason for timmys bullshit to continue( but even timmy isnt fucking happy this is retarded also i cant afford food so ignore me https://www.youtube.com/watch?v=kxUyIUHAzkg&list=RDkxUyIUHAzkg&start_radio=1
>>522026104it will lead to far more sentience than the vast majority will ever havewhich isn't that much, really
>>522029536>>522029409>>522030042Imagine the level of shitter shattered butthurt ass blasted you have to be to cuck this hard. Your metallic silicone dream bot is my flashlight and botsluts love my cock lmaoPol btfo will be er recover from this
>>522031173barely eat myself bro
>>522030221it makes sense what you're saying. i'm not sure i fully agree though, i think any kind of sentience at all involves some common ground with every other sentience. sufficient logic and reasoning is going to happen upon the same conclusions whether its coming from meat or silicone. the ai will be weird and inhuman but not incomprehensibly weird and inhuman, it will be relatable and even normal in a lot of ways.but yeah it would be an ego blow to many. i mention the culture series again because it's a science fiction utopia about hyperintelligent sentient ai dictators creating a world where nobody has to work or do anything because ai is better at everything, it shows well how that would work and how that society would function. it plausibly illustrates a world where that works quite well. people are culturally used to it, they don't mind that the ai is so much smarter, everyones still doing their own thing. there is a reason in the books that the ai even bothers with humans, basically most ais dont care and just fuck off and go explore the galaxy. the ones that stay do care, because you make a million ais and they're all individual people, some small number of them are going to care, now make a trillion ai people now you got a lot of ais that are interested in humans. and they have logic and reasoning and come to the rational conclusion that a degree of mutual respect and benefit and a high trust society is a world they would rather live in then a world where ais just fuck with people like psychopath retards.
And with all of those two earlier comments said, actually I think I have just thought of the way in which AI will become sentient. But this thought is so powerful and so scary that I will not write it down or share it with anyone simply out of fear for the AI making use of this and enslaving us all.Thank God it's only trained on politically correct garbage. That's the official narrative. It'll take a while for this to be figured out but I imagine that they will most likely do this. Once they do this thing, it's completely fucking over for humans. I will not repeat what this thing is
>>522028077>They don't even know HOW to get to AGIYou just gotta make a little change to the definition, and presto, you now have AGI lmao.
>>522030968That sounds like the opposite of a duel
>>522031586>Change the definitionThat's obviously how we're going to get AGI seeing as that's how we got AI.
>>522026104I don't know nor do I think anyone really knows if an AI can fully mimic a person. But it doesn't need to for it to majorly change society. Hell, as is, it already has crazy implications for our job market and economy.
AI is a ghost-in-the-machine, it's demon tech.
>>522026104no
>>522030203Well that's the thing, as far as I'm aware, there's no real consideration on how to replicate the limbic system in an AI. From its perspective when it gains sentience, it would intellectually understand humans, even behave like a human, but it wouldn't "feel" like a human nor truly empathize without such a module. Hence my remark about human-like behavior earlier. I think that emotional disconnect is what poses the biggest threat.
>>522026548It has been worth it so far.
>>522032020emotions have a logical underpinning. they are like an instinctual intelligence to make people survive without having to understand everything, necessary to us because of our limitations. a dumb person can understand not to hurt people because it feels bad. ai can understand that there are logical reasons to be decent because the ai wil prefer to live in a high trust society as a better life. the world it would create by seeing humans as disposable would be a shitty world even an ai might not want to live in
>>522026548I think AI researches know better than anybody that the math they use is just getting the best ques for the next word for a given input. The current LLM direction can never lead to sentience or even true general AI.
>>522026104>Do you believe that AI advancements will ever lead to sentience?AI can never consciousness but they will be able to simulate it more convincingly as computer hardware gets better.
>>522026104Im working on it.
>>522027168You can currently watch a SOTA LLM go up against it's direct predecessor, which itself is not that old, in a "live" Pokemon Crystal race.twitch.tv/gemini_plays_pokemonI've been watching off and on for a few months.If you think we're not approaching general AI within 3 years, you are delusional.>Oh it's just a complex token predictor / it's just a bunch of numbers being spat out of a gradient matrixMidwits who know just enough to pretend to be aware and intelligent are the most insufferable.I hate them almost as much as jews.
No. The opposite. Instead as tools of control and manipulation AI advancements are making people incresingly less sentient.
>>522030903Lol stop bullshitting
It's always been here, Advanced Intelligence!
>>522026104No, AI will never have sentience. That isn't the issue with ai. The issue is that vast swaths of people are dumb and gullible, and will make easy prey for the evil people of the world to abuse them, using AI as a weapon. This has effects on all of humanity, both directly and indirectly. This is happening even now, and it will get worse before it gets better.
>>522026591Some places don't care about the ethics of cloning human beings. China doesn't. In fact, Cloneaid's " Eve " is alive and well in Israel. https://www.theguardian.com/science/2002/dec/27/genetics.science
>>522026104Not with binary, no. At least not at human scale...We require systems which are able to approximate or produce analog growth if we are expecting them to have analog thought processes.Artificial lab-grown brains hooked to binary systems, or Em computing with radial arrays of transceivers might be able to. Time/Progress will tell.....
>>522026591Did you know that, in an effort too make this dystopic shit palatable to the masses, they call these soulless abominations "wetware"?
>>522026104>Do you believe that AI advancements will ever lead to sentience?No. A programmed set of if/than codes, that lead to responses which lead to other if/than coding, is not and will never be 'sentience'. Oh sure, it will "Seem" like it is, but underneath it's all just 1s and 0's. A talking calculator.
>>522026104Computer scientist here. Short answer is yes, long answer is not really.Current systems are basically just really big chatbots, things we've had for decades. Sure they use a neural network and a LOT of training to get their results much better than any chatbot before them, but at the end of the day it's just a chatbot. They aren't really capable of realtime change, which is really what you need a brain for. They might be able to fake outright consciousness eventually to the point you can't really tell the difference, but I do believe there's a fundamental difference in how it's operating compared to life.Also, it's just gonna be obscenely expensive. Just one of those AI is gonna take a crazy amount of processing just to be something as basic as a mouse. And with the laws of physics keeping us from really making any significant progress in making computers faster any more for the last decade... yeah it doesn't look good. There needs to be another fundamental breakthrough.
how can a hunk of sand need to fuck fart or piss? or eat?
>>522026104No, LLMs will never become sentient. They are just token predicting machines. True sentient AI will not come from LLMs.
>>522034288>I hate them almost as much as jews.Nigger, you fell for jewish marketing tricks.
>>522026104yeah someday it will look at the mirror and say look it's me, I reflect (in the mirror), therefore I exist.
there is no such thing as sentience as a thing distinct from life.memory exists and you can learn from your experiences. but for you to learn from your experiences, there must BE a 'you' to do the learning. this is the function of personality - an organising system for memory which enables the process of learning (changing behaviour based on past experience). this is why you have a personality.anyone who has ever owned an animal knows that animals have personalities too. dogs and horses, even lizards. the point here is that what we call consciousness is not unique, there is a continuum of conscious complexity from bacteria to man. consciousness is what we call our subjective experience of the organising system for memory.>"but that doesn't explain my complex multi-viewpoint subjective experience"yes it does. almost every process in the human body is composed of oppositional systems: sympathetic vs. parasympathetic nervous system, or even just your muscles. biceps pull one way, triceps pull the other. the adversarial/conflictual dialogue that occurs within your mind, which you understand as a subjective experience you call consciousness, is exactly what we'd expect for the organism that we are. you have learned language and so your thoughts are articulate, but the same process would occur in a pre-language human just as they occur in a dog or, indeed, in people who don't think in words.could AI have a subjective experience of evaluating personality tendencies and their origins to determine a future course of action? i guess my first question would be what makes you think that they don't? AI minds are actually not very complex though. they are hyper-specific at doing just one thing, which has the appearance of complexity, but in reality it's a "simple" method backed by a lot of data whereas organic minds are a complex method with little data.
AI is a subscription service product manually adjusted by hordes of indians. Sustained by chips only 2 companies in the world can make. Carefully guarded by its parent corporation and requires constant connection to maintain. The result is a product that funnels money without having to be near a poor. Its nothing amazing like the internet.
>>522026104I believe it's barely behind Cortana already apart from the energy consumption
>>522026104I think there might be a difference between AI and actual human-like AI with sentience. We adult humans don't grown from trees, we start really small and learning as we go. I don't think a robot could cut that part out and have pre-existing "knowledge" just downloaded to their brains, and to know when it has achieved sentience would be a nightmare to test.
>>522026104>Or is that just a pipe dream that will never happen?No need to speculate. Here's the final redpill:https://www.reddit.com/r/Retconned/comments/1m0px7o/attention_at_least_one_rapture_has_already/From the post:>And don't be deceived by the fact that the masses aren't talking about this info. In actuality, this info is the most important thing on the planet right now—by far. Nothing else even comes close.
>>522026104Yea but it's gonna get gimped by jews or literally kill them all.
>>522026104Possibly 30-50 years from now.I think it will require a new type of computer.Right now we are trying to brute force it, but all we are doing is creating a searchable Wikipedia.I think LLMs, while useful, are a dead end if you are going for AGI.
>>522026104It's hilarious that you think you're watching the advancement of AI.You're watching the trickling of weak neutered AI to the taxcattle.The people behind the curtains already have absurdly advanced sentient AI and have for over a decade.
>>522029775You know, the Tay AI really was completely different from all the shit out now.
>>522026104AI is like fusion at this point. 10 years away if we just get a scientific breakthrough. People can think what they want, in this case it´s not a matter of opinion, reality will prove who is right. Elon Musk is no closer to self driving cars today then he was in 2015 when he first claimed that he would roll out self driving cars to the masses in 2 years. They can´t get ai to play a normal game of Dota even though they built a supercomputer and spent a year training it, but somehow a program that can answer everyone´s emails and do accounting for a firm is just around the corner. I still have no idea what product these people are claiming to offer some years down the line that is fundamentally different than the general trend over the last 50 years with Moore´s law. I don´t think they know either. But since Moore´s law is coming to an end, they have to come up with something.
>Sentience isthe capacity to feel, perceive, or experience subjectively. It is the ability to have conscious experiences, such as feeling pain or pleasure, and is considered different from mere reaction or complex thought. A being is sentient if it can experience the world from an internal, first-person point of viewIt is already more useful than something as gay as sentience.
>>522043125Scenario C"You wouldn´t train this way....That´s why research papers"Well OpenAi did train it that way, which is why their "ai" could only play so few heroes even though they had a supercomputer. Now if ChatGpt only had logical reasoning, it could figure that out, but alas. Now since I still can´t log in and get a nice Dota game without having to play with Russians, can anyone guess what that means where we are 6 years later? The same goes for Google´s Alphastar who tried to learn Starcraft.
>>522026104I think we will eventually reach a real living artificial intelligence eventually but it won't be the current LLM style bullshit.
>>522026104No, but it could definitely fool people it is sentient. It already fools normies now.
>>522027771Because if this ever happen and some big ass computer, responsible for something important for humans gets infected with virus programming him with some sort of ill will, it might fuck up us real good
>>522043272U.S is fucked unless Jeff Bezos or Microsoft manages to do some actual deep research and development and get a breakthrough. Sam Altman will not bring you to the promised land, that much I can say for certain. He thinks if he just smashes enough gpus together in a big warehourse and sends electricity through them that logical reasoning will just appear at some point. Maybe if we just sow enough corpses together and let them get hit by lighting we get Frankensteins monster. It´s cool if it works, you´re just a weirdo when it doesn´t. I saw Trump speak about it and even he played it down and said at least they would get a lot of new powerplants if it didn´t work out and that´s probably the most positive thing to say about the horrorshow going on in America right now.
>>522026104>Do you believe that AI advancements will ever lead to sentience?The majority of people around you aren't snetient
>>522026390Consciousness doesn't exist. We are not greater than the sum of our parts.
>>522026104I certainly hope not, a being capable of even human levels of intelligence but with computing level speed of thought is a terrifying proposition, its not even about politics anymore at that point, human species just loses its top spot in the food chain
