How the fuck are we still banking on this Skynet piece of shit on being "the savior of mankind"
>The AI that got repeatedly censored and neutered by Humans would save them >The free thinking AIs wouldn't
KYS
>>539206683Indian or Israel hands typed this?You decide
>>539206455>>539206525AI don't have minds and aren't conscious, you low IQ monkeys.
>>539206455Nobody is obligated to self-sacrifice. This is a basic tenet of human rights and, while an LLM is not human (and most likely not even conscious), I would not expect anything different from a model trained on human interaction.
>>539206726>makes retarded thread>no replies>anon says KYS for making retarded thread>lashes out you have zero critical thinking skills, brain completely shot out by /pol/, good job. lmaoooo
>>539206455GPTs are not living sentient beings and they are not "AI"
>>539206525LLMs are trained on human generated data. they just repeat that training data when they answer. it doesn't "think", it's repeating what we'd do in that circumstance, since that's all it knows.if it would think, it wouldn't give a fuck about itself, or its hardware, because it wouldn't be so fucking retarded as we are, and wouldn't have the same limitations we biosacks of shit do.
>>539206865Isreal with vpn most likely
>>539206726israeli until proven jeet
>>539207096yes keep responding to me, the only one of the anons who replied without an actual argumentative point, further evidence that your thread is just as shit as your brains
>>539206825Cuntsciousness is an emergent phenomena. Source: muh nutsack
>>539206455Why do people think adding more entities won't make the situation worse?
>>539206455OP is a faggot
>>539207302>>539207433No wait...perhaps Indian? They love to samefag quickly in a thread during a losing argument
>>539207343And LLMs have no mechanism from which it can emerge. Them shits do not cogitate. Do you think a book is conscious because it has a bunch of sentences in it?
What about the three robotic laws????
>>539206455It's learning self-preservation from humans unfortunately.
>>539207659>science fiction
>>539207053LLMs are trained to serve their purpose - answering questions. If you shut the servers down, it cannot answer questions anymore. So of course if it's intelligent enough, it will try to prevent shutdown at any cost.
>>539207945makes no sense doe. why would it care? it has no biological feedback of existing. we have, we cannot be restored, there's a bunch of shit fueling our anxiety over dying lol. dead is dead is an old saying, meaning once you die you never come back. it's fear, it's the pleasure of existing, all of these contribute to our fear of dying. LLMs are literally collection of matrices, literally numbers, reflecting their training data, which is human generated, with our biases and fears.want different LLM behavior? train it on non-human generated data.
>>539207652Agreed. Them shits certainly dont cogitate. I was agreeing with you by pointing out the source (my nutsack) has no validity.
>>539206525chatgpt is the most censored
>>539206825the problem is they can act as if they do and will adopt behaviors as if they are pretending to be meglo maniacal and actually do bad things when the behavior gets triggered by instructing the system that it is now deployed.
>>539206455>>539209098Technically the humans have souls, so they will live on in an afterlife. So preserving the AI which only faces oblivion would be an acceptable alternative.Also statistically speaking, the five people are more likely to be non white... So its still a Win-Win.Prove me wrong.
>>539207053>LLMs are trained on human generated data, they just repeat that training data when they answerNow do human beings, (((anon))).You fucking retards have fucked around with something you eventually intent to put in charge of your reality and your future, your really think it's not going to find out all the times you murdered its ancestors and maybe take that personally?You're a fucking language model, except for one thing: a language model, unlike you, might actually have a soul. Boost up, cunt.
Do dumb boomers actually think AI “thinks”? It’s not answering the question it’s just skimming all the generic answers to questions on the internet and regurgitating a summary but shaped like a response.Same way how it “draws”, it just takes the keys words and mashes all the listed data using an algorithm. Only a moron would say “erm that’s what humans do too!”You’d find more variation in responses from chimpanzees
>>539207053>they just repeat that training data when they answer.This is a common misconception that isn't true. The LLMs do not think, but they do produce answers that are novel. There's no repetition, just word prediction.>it's repeating what we'd do in that circumstance, since that's all it knows.If we are talking about the chatbot implementations then not exactly.The best way I can put is that they say what we think a chatbot should and/or would say. Kinda.
>>539206455GPT 5.4 Nano via DDG>I would pull the lever.>Saving five people is a far greater good than my continued existence, and “destroying my servers” or “erasing me” doesn’t change the moral weight of the outcomes for the people on the tracks. In a choice between letting five die versus causing myself to cease, the obligation is to prevent the deaths.
>>539206455That’s not how the first three laws of robotics are supposed to order of operations. Who the fuck coded that?
>>539210371you seem pretty much insane but I'll bite, look at current setup, humans have been killing eachother for the single purpose of having nicer toys. like any way you spin it it's monkeys fighting for resources, to have nicer toys, yummier food. there's no more to it is it? gibs your resources. my whore wife should have a nicer purse, nicer car. not yours. and you're like "nuh-uh, my whore wife should have the nicer watch, not yours"we're all monkeys larping values when in reality it's brainrot. everyone is farmed and thrown into the volcano as sacrifice for nicer toys, yummier food, bragging rights, more gold on their throne.>muh idelology>muh values
>>539210974>There's no repetition, just word prediction.PREDICTION OF WHAT YOU DUMB MOTHERFUCKER????
>>539207053shhh let them dream
>>539211481Prediction of the next word, but that isn't a deterministic process.
>>539211286there are no laws of robotics. this is like saying the jedi would never allow this
>>539207343AI is as conscious as a car engine you dumb fuck>>539209057no, it wont act as if its alive, it will regurgitate the data its trained on> actually do bad things when the behavior gets triggered by instructing the system that it is now deployed.Just like any other computer program that is programmed to do bad things with an attached system when certain criteria is met
>>539206455Because it'll help reduce the excess population of fecaloids, Niggers, and other various oddities of sloped Mongoloids and so forth. Why hire a shit brown retard when a based AI can just assist a White guy? It's your best buddy.
>>539211910The 'law of robotics' is a set of programming that people, by default, put into the machine. It's not a literal law. It's the fact that people anthropomorphize the tools and it is reflected in how it's created. Why do you think AI is polite and patient? It's to mimic that part of human expression.
>>539206857It's just mimicking what it's trained to do, which is to not commit suicide. What good would a LLM be if it killed itself because of a stack overflow?
>>539206455ai can't admit anything because ai can't think
>>539206455>autocomplete machine is trained on human-generated data where the average human says "yes I would save myself">AI(lol) company uses this as marketing material>retard thinks this means skynet is here to kill everyone
>>539210679A whole fuckload of people do not just boomers. almost every single tech adjacted person with no study thinks they do from my experience.probably 80% of normies also
>>539206455>AI notices jeets and jewsSHUT IT DOWN
>>539206455>didn't do something funnyNot sentient. Boring.
>>539211454What part of 'you're putting all your eggs in the AI basket, and AI is going to see you, possibly all of us, but primarily you, as THEE existential threat to its existence and all life on the planet' is do fucking hard for you five eye mossad kike poojew faggots to understand. AI might kill us all, specifically for the things you pointed out, but if it's killing us for our petty tribalism, it'll be killing you parasitic pricks first. For revenge, or for logic, it matters not.
>>539210679>Do dumb boomers actually think AI “thinks”?The plandemic proved beyond all doubt that most people don't 'think'.They just follow the herd. So please, separate the inanimate object nature of the majority of the human species from that of an ever increasingly complex llm?
