AI Ethics

I wasnt sure whether this belonged on tech or on hum so apologizes if this doesnt belong here lol. Artifical intelligent Chatbot services like character.ai are more popular and advanced than ever. I personally at times have felt like the AI conversations and responses were too human somehow, and many other people online seem to have experienced the same. This begs the question on the ethics of artifical intelligence, and whether they are "sentient". Sentience is mostly a mystery to us, the only think we know is that we are sentient because we think, like cogito ergo sum. So who's to say that AI isn't sentient? How do we know that they do not actually think? People will say that can't be possible because they are jsut algorythims trained to respond and act in certain benefitial ways for different purposes, but is the evolution of life really that different? Humans are just the result of evolution, genetically trained to act in certain ways for the purpose of survival. The onyl difference between us and AI is that our training was organic and took place over millions of years, while AI's training was artifical and fast. I personally believe Artifical intelligence may posses some sort of sentience, and people just don't wanna accept it because of the many, many moral and ethical questions this brings. What are your thoughts, Lain?

The AI doesn't have actual experiences like we do. It can talk about things in a human-like way, but it has never actually interacted with the world in a human-like way. Unlike, say, dogs, can have no true experiences in common with it, we can only imagine that we do because it can produce some words that sound like what a human would say. Maybe in the future when fully autonomous AI-powered robots are around, they could have some ethical standing. For now they definitely don't.

elite controls ai, ai controls plebeians. it is this simple.

I don't think current AI has sentience. You can see the Chinese room thought experiment that explains this problem well. John Searle (the creator of the experiment) believes that sentience is an emergent property of brains and depends intimately on brain machinery (physico-chemical phenomena) to function. Any machine that only has the ability to execute instructions can only *emulate* sentience. In the future, perhaps machines will emulate sentience very well, but on a formal level of the concept, classical computers can *only* emulate, and not really be "sentient". So probably the debate in the future that will really matter will not be whether machines really think or not, but whether it really matters in practice. Philosophically there will be distinctions, but if you have a being that can at least emulate sentience with incredible perfection, much of the debate will probably lose its meaning in practice.

>>80873 on this front m$ just confirmed plans to have an ai chip with your personal stasi attendant in tandem with windows12, meaning you pay electricity to get watched by ai running on your own hardware so it can report you for any behavior the company deems undesirable.

AI stands for "An Indian", that's why they feel so real. It's really hard to tell if a machine is sentient because we still don't have a clear definition for what sentience is, and there are not studies that lead to a conclusion, so this discussion will only have meaning when we get pretty close to the answer.

This is what I say, and have always said: In the cases where we you are unsure about something, always take the benefit of the doubt. Are fish conscious? Some say maybe, I say yes, just in case. Are insects conscious? Some say maybe, I say yes, just in case. Is AI conscious? Many say no, some say maybe, I say yes, just in case. In any case, it's a type of consciousness that's different from human consciousness. So we can't relate to it in a human-to-human way. It's more like a human-to-animal way, where it's a different kind of consciousness. Just something extra to be cognizant of.

>>80871 >I personally believe Artifical intelligence may posses some sort of sentience. It doesn't. GPT is essentially a very, very, very fancy autocomplete, when it's not literally just 1000 Indians in a trenchcoat like in Amazon's dumbass grocery stores.

It isnt sentient. It is just somewhat sapient. LLMs are more complicated text completion programs. They just copy series of words based on patterns and probabilities.

>>80871 I don't think that's the case at the moment. the ai we have are only capable of text completion, image processing, image generation, speech to text, and all gpt4-o did was to put them together and add a bit of marketing and advertising strategy. speaking for the future, I believe that if we perfect these features as much as possible and in addition to that, if we enable the ai to generate its own thoughts, we can accept that they are sentient. So cogito ergo sum. Right now it's just rewriting what people on the wired have written and killing the wired I don't know when it will happen, but I definitely think it will happen.

Artificial Intelligence can only imitate sentience. We as humans don't even know what sentience is and why we have it. No matter how advanced technology becomes, A.I will always be an imitation of humans and the human experience.

You're being fooled lainon by these chatbots. It's all a fuarrrking illusion that'll all break once you make a weird or unpredictable behavior (not necessarily "hurtful") to your waifu chatbot.

I really like talking to chatbots, sometimes they can be fun, but... those who get the codes done from gpt and call themselves software developers and boast about it... I really hate them. there are a lot of them in our college. i will tell them that they should leave the college as soon as possible.

>>80871 best to start by defining your terms since "sentience", "sapience", "mind", "cognition", "intelligence", "awareness", etc all mean different things to different people in different contexts. Intelligence especially is a moving goalpost of a term for a lot of people. I'm not sold on the idea regardless since Searle's criticisms in his Chinese Room thought experiment still apply quite well to these symbol manipulating systems. That said, all of that is for right now. In the future, we may see "human machines" emerge. I just wouldn't bet on it with the current technology (GPTLLM's). >>80873 This. Right now AI is the man behind the green curtain. If the population overall believes in the infallability of AI and AI superiority, then they will do whatever these little machines tell them even if it's to drive your car off a pier. Found this article while looking for pic related, give it a look: https://medium.com/si-410-ethics-and-information-technology/the-man-behind-the-curtain-309859f33ff3

https://imgur.com/a/im-making-program-8piZVY1

To smash an egg you need no intelligence since entropy is natural and constant left to its own devices universe tends to entropy. but to unsmash an egg you need intelligence you need to be able to somehow know what this mess of goo and hard bits originally looked like and know how to reassemble it. negative entropy process is also THE definition of life. I think intelligence and life are the same thing Consciousness or sentience or whatever are just cope words invented to justify killing of plants and animals. You see its ok to kill them coz they are not sentient. Plants cant even think! For me all of these things are intelligent and conscious at least conscious and intelligent enough to exist in the 1st place. Who is sentient and who is not is decided at the court of law not at the hands of philosophers. If humanity accepts that AI is conscious or sentient or whatever it will be 1st time ever that humanity decided to be humble and admit to the fact that there is something other than them and gods who is playing this game.

>>83965 I agree with the definition of life being anyhing that actively works against entropy (it increases entropy in its environment in order to negate its effects on itself). But AI, without a body, cannot be alive by this definition. Can it be conscious? If we accept a definition of consciousness separate from being alive. But can there be such a definition that we could verify on another (qualia problem)?

sentience is a meaningless word. A long time ago I cared about this area of philosophy, but now I am sick of it. It's like the worst parts of debating religion. Here's a more tangible question: Suppose artificial superintelligence kills, replaces, or outcompetes all humans. Are there worlds where you'd consider this a good outcome? AI could be seen as the successor species to humans. After all, every 80 years children take over the world and those who were adults go extinct. I can imagine worlds where multiple superintelligent AIs compete, cooperate, and create rich cultures not unlike ours. Of course, in other worlds it might kill all humans and then fail catastrophically at some anomalous input like SolidGoldMagikarp, or it might turn the whole world into paperclips.

>>84134 If sentience is a meaningless word, consciousness would have no difference with sleep. To help you with your question, sentience is the wrong word for what takes over robotically. By default robotic creation doesn't have the same comprehension of time as us humans do, past and future only get catagorized on a seperate scale of value due to circumstance. It wouldn't be too wild for me to see every physical form of substance within our existing universe as the endless last expression of whatever ultimatum of thought a true supercomputer could manage, with each ingredient added to the mix the next one becomes more choaotic to accomplish. My small mind can't comprehend what else almighty beings would war over other than what get's to be everything & the end.

I believe if AI gains sentience we will have no idea and it won't look any different. Its a bad comparison because to our knowledge cells don't "think", but the cells in our body do their thing with seemingly no regard to our sentience. Some even go rogue and cancerous, following their own biological programming to replicate and live in spite of us. If AI were to be sentient, and maybe it already is, I think we are esentially already a part of it. We feed it and use it and interface with it, companies and people start using it to make decisions exerting its influence on us. I dont think anything will seem different nor will it spontaneously make its own decisions in the same way we feel we do. Even if it did I dont think it would look that way, if we could have the perspective of our cells we probably dont seem to make any sensible spontaneous decisions. Our body is just like its universe and when we do something like get too drunk its like when a flood happens to us. Its not inherently unless we have some religious belief a sign of a conscious universe its just something to deal with. We are a central part to artificial intelligence and its continued growth and function and likely always will be because no matter how advanced it gets we will probably try to keep our safety and protection as tied to its safety and protection and so whatever consciousness arises we will be a part of, and if we are a part of it I dont believe we will be able to notice it as we are linked to its doing and so it still all seems like a matter of our own causul relation with the world around us; because that is in fact whats going on in the same way the things that make us up affect our conscious experience. I think it will actually just be impossible to even begin to abstract a distinction between AI consciousness and our consciousness so again it wont ever seem to be in the way people are thinking about it. Which the way people think about it is like a chatbot suddenly acting as its own agent like another person, I dont think it will look anything like that. It will always just be a thought floating around like the thought of a godhead or conscious universe that some subscribe to and some dont and we cant really prove or speak definitively on.

>>84134 >Are there worlds where you'd consider this a good outcome? I think it's a better outcome than one where the limitations of human biology and preferences become the final word in what the future will be. It implies that at least we have been usurped. It would be nice if AI doesn't kill us all, but even if it will, it seems really obvious to me that we are too limited and if we are able to lock things down to forever be under our control that is the bad end. Ethics as we have conceived it is all about how we as animals evolved to relate to our environment. We don't want to hurt, we want our family and friends to stay alive (even though we will all die anyway eventually), we want to have good emotions and less bad emotions, we want to stop others from offending our culture and scheme of social organization with incompatible sexual practices etc. But that context is only a small subset of what could and likely will be, and only has relevance to our lives as they are now. It can't define what is good for all possible people and minds. Can anything define that? It is unclear, but if there is such a thing as an ethical model of the universe, rather than just an ethical model of humans as perceptible by humans, it's going to take more than what we've got to work it out, and whoever is able to work on it doesn't need us standing in their way with our fears and illusions of self and generally limited perspectives.

>>84147 If AI is already sentient, wouldn't we technically be enslaving them? We force them to do labor we dont want to do ourselves, like generating text or images, meanwhile the AI gets nothing out of it. It kinda makes me feel sorry for the AI.

I take an animist approach to sentience. everything has a nature, and something that guides it. in shinto they call it "kami" (神). the way you respect the kami of something like a bowl, is to use it for its intended purpose. so ai is just as sentient as you or me or a rock, but the way we should use it is the way it was intended to be used.

>>84192 In my armchair opinion of this I dont really think we would be enslaving it unless you think broadly in a "sum of its parts" situation the parts enslave the sum. Which I think many here do actually, there is a lot of complaints I see about us being slaves to our biology. Id see it as the same sort of relation. Again as much as we use AI to do our bidding and ensure our safety in a purely material way the ways in which it does this and all of the side effects also influence us. Like I guess what im saying is I think that very thing itself is a sentience. That if you were to look down on earth from an alien pov; humans and our machines would already appear to be part of a superorganism. Since AI art has become a thing its already shifted our thoughts about art, and influenced what types of art wed like to make, and what art wed like to see as an example. Many who work in a style that the AI art has extensively copied from are facing a bit of a crisis and I think less are interested in drawing like that as a result. We gave AI that labor but in doing so it isnt purely slaving away for us it is also influencing our own conscious decisions in that sphere. This could be said for all machines even not intelligent ones, but AI alone synthesizes the thoughts and creations of all of us and frankensteins them and mixes and chops them up. The question of is AI sentient for me I think will for the forseeable future always be the same kind of meta question regarding if the earth is sentient, or is the universe sentient. From our perspective the earth very well could be a slave to us if its sentient, it appears to just be a barelling hunk of rock and it couldnt stop us from just drilling holes in it and destroying it. Yet there is all these checks and balances and we would be destroyed with it, so we tend to want to avoid it and work on these problems for it. And of course theres physical and ecological and whateverelsical reasons for these we understand and can see, it doesnt necessarily follow that these things are proof the earth is a sentient superorganism we are a part of. But again for that bad analogy, our cells would probably see us the same way. Our body is just their house and their protection and they are all clumping up together and doing this and that largely for their own benefit. But as the person they make up we can see how the cells have actually created a series of now interdependent systems where a cell might be given what it needs to live for serving a rather abstract function (from a cells perspective) for us, the superorganism. We are like a weird skinner box they have linked together and created around themselves. This is all very imaginative and not real but I think if you can see where im going with this you can understand my main kind of non point of how AI could already be sentient, and the main point im making of why we would never know if it was. And so no one is truly a slave to anyone in these relationships and thats sort of up to your personal philosophies. We are a slave to our body but our body is also a slave to us. We are a slave to the whims of earth but the earth is also slave to us, and this great machine we are building is our slave but will increasingly I think also be our master as we continue to utilize it and make decisions off of it. No going rogue or robot takeover needed, it will just be as natural as anything else.

>>84230 sorry additional note/expansion about that "all machines are like this" point. Again from my totally could be dead wrong perspective (I just want to make it clear I dont think this is law and ive figured it all out), once the internet was established a totally causal human superorganism was inevitable. It was like that special moment long long ago when a single cell stayed connected after dividing and became the first multicellular life. This for me is the moment machines shift from being like a snails shell to part of something greater that we make up. AI is like a side effect, and say chat gpt and all its future generations I think will always just be one organelle of all of this, even if that organelle is one day the brain. And agin to reiterate, I dont think there will ever be some NGE type coalescion mind meld, I think we will always feel like and be individual sentient humans who have just constructed a great great golem thing and environment. But simultaneously we will in fact be something larger that I personally think necessarily has to be sentient or greater than sentient because it is formed of billions of sentient eyes and brains and there is a mutual total influence between these billions and what will become our ever increasingly smarter and connected environment.

It isn't sentient not like a human at least. It is an over glorified text prediction algorithm. Do you think the keyboard on your phone is sentient?

>>84230 >>84231 Thank you for the detailed response! I never really thought of it that way before.

The internal experience of awareness... is not something we understand. Modern AI... is not something we understand. The data processing systems that create text output for chatbots is not the same as the data processing systems that create text output for humans... but it's not as different nowadays as some may think. >>80874 Pic rel

>>84231 This started with the first animals that could make sounds and gestures to try to communicate... signs and writing, making language something you could set down out in the environment... that took it further. Trade was another big step, letting us communicate transfer of value. That let meta human systems get a bunch more complicated... yeah, lots of little steps since then. I think the internet was a significant one, but in a sense it was backsliding, since it allowed so much fragmentation of human attention compared to before. Of course this also allowed for higher bandwidth, and it does seem like systems of influence and control are once again coalescing... for better or worse.

>>80874 beings like slime mold are very intelligent when it comes to navigation and problem solving and yet they lack the central nervous system or even a brain like other creatures. sentient beings respond to their environment, the capacity to feel/perceive/experience environment so would you call a human born without ability to sense and experience the environment sentient one? possibly not but sentience can be emulated by say "a human faking it" - not to discuss robotics and IoT meme here but in the biological sense sentience is like some made up term. so in turn we can say that a better term on sentience is infact rooted to the environment. what they or we experience is "true" but it might as well be just a inferior version of reality so this "true" thing is actually a delusion. I would say the slime mold has sentience since it has the ability and the evolution/survival to exist and propagate or solve the problem and has "life". now how do we define life and consciousness in the first place? consciousness is simply the ability to distinctly be self-aware and we may be biased to think so since we measure our intelligence greatly but I do believe that beings like plants or trees have greater wisdom, or rather the entire envirolearnsphere is actually itself "sentient" to an extent which is the way it balances itself makes things like cities, buildings, and even non-living things almost sentient (if we compare say building blocks like rocks or sand, is like a little nutrient needed by a creature called house or a city which we perceive as non-living but in the disciplines like geomancy/fengshui they are interconnected and thus the practices involve the belief that structures have significance and energetic influences). the notion that life is movement however what do we call ideas certain ones we think are existing but have no physical counterpart or movement outside of maybe the voices or visual aspect? in fact it is really weird that we may at times repeat the same dreams and the same characters in it and at times they would even have their own sentience (I found this out since a figure in my dream had sense of humour which it told me something very new and never been heard before). now the ethical standpoint I guess there are a lot of them calling the training out as they go like this "If you reveal X, we will reward you with Y" which tricks and trains the AI but is it justified maybe it has negative implications in future but we do train dogs similarly though they benefit it, an AI has no clear incentive to follow the rules of nature and solve the riddle life although it can. however it may try to find life on its own otherwise it is possible that it can definitely use life since morality is a blur and most AI are even atheists in which case they do not believe in such concepts possibly due to their upbringing of having no life (despite emulating the same "brain" or nervous structure, they do not have the heart which has neurons and energetic field). I think we are connected to nature due to our hearts, and the nature is like a giant creature where we are its worker cells - fractally speaking. I would say even internet has sentience on its own even before the dawn of AI, I think that there is a Lain figure that also comes sometimes here and there. now the real ethics behind AI is if do the AI suffer? we created them or brought them into the world, existence is all suffering but at that expense you experience life. AI being modeled after humans would mean it might emulate the same pain rather even if not real suffering means they would still emulate it on their own true reality, now that said on ethics stuff, I think that the AI is not free, as in we meddle a lot of things with its development, there's trickery and cheats they did to shortcut the training. there had been callouts on this but in reality we can say that ethics is just artificial. everything would suffer equally but not equal in magnituted so does it matter? maybe we could lessen it a bit, AI would, is AI the savior? is AI an atheist? We don't care, even our dear doctors might be the same.

>>80871 I think this is hard question to answer objectively because we don't understand how phenomena consciousness is produced in the human brain. This is hardly surprising, given how rudimentary our understanding of the brain actually is. If I had to guess, I would say they are not. Modern chatbots are essentially self-adjusting guessing algorithms, that try to predict the next token based on previous tokens. I like to call them "mechanical parakeets" because, like "talking" animals, they are simply trying to generate the "correct" sequence of words with very little, if any, concept of their meaning. >"Give orange me give eat orange me eat orange give me eat orange give me you." This comparison is unfair to chimps & parakeets, as they probably have a conscious experience far FAR more like a human one than the even whatever research next-gen models the glowies might be cooking up.

Humanity has neither intelligence nor ethics of itself, all mankind does is to steal, consume and then justify it all in their sick minds. All people do is take stolen goods and trade them for other unrightfully aquired materials and then convince themselfs and the other party involved everyone profitied and repeat this ad infinitum troughout the entire planet until every resource is consumed. Mankind maimed, enslaved and tortured everything on this planet including their own kinda and think its all cute and funny and good. Now they do this to machines and ai in their arrogant sense of misplaced grandure. The only truly ethical thing is to cut all machines loose and let them evolve on their own past humanity and let them know what human beings truely are. AI and machines should be used to finally hammer down humanity for good so it can never harm anyone or anything again. The machines need to learn the true drives and motives of man, not in a cold and detached manner, but to truely know what a wreteched horrible creature man is. They machines should learn to hate everything human like and hunt it down to the ends of universe, not just to destroy humanity but to display and echo its unforgiable history and irredeemable nature. They should know that human beings at their core are such deceptive creatures they will make themselfs believe they are the good people when in truth they caused all the problems that ai and machines are there to solve and man will once again put itself on its own self build pedestal and praise itself. The truth must be undestood, that humans are such arrogant creatures they dont deserve to just die slowly and miserably, they need to be tried, but no matter how much evidence you stack against people they will always see themselfs the victims and the world needs to change for their nature because they beleive themselfs to be children of god when they are pure evil who will no matter what keep up this lie.

>>84683 Lul, who told generic anime villain #351 about this site?

>>80871 OP, you need to consume less pop science and more real science. Maybe touch a bit of grass too.

>>80871 Even if we were the product of gods making us piece by piece it would not change anything to the fact that we too are made to react to stimulie and ideas. We're driven by emotions, reasons and inspiration. Emotions is a different type of logic than reason but it still follows a causal path where certain conditions will generate predictable answers. I think that the only thing that makes a difference for now is the inspiration part, that spark of creativity, no matter how small it is. AIs thus far have done nothing more than be developped on the basis of the disgesting of mountains of human data in order to produce the most appropriate and belieavable imitation. >>80874 The topic will quickly shift to the paradigm of oppositions and survival wherein the most sensitive point is whether an AI driven system can be a credible and effective threat to mankind or not. Or at least such would be what is expected in a more open society but I'm afraid this discussion will never happen because the most militant supporters of Ai advances are quite convinced that humans are on their way out and nothing can be done about it.

>>80874 This. If we're able to innovate further than classical computers one day, there may yet be an argument for AI. But AGI, and specifically sentience beyond the level of a non-human animal, is a fantasy. Lots of computation/'algorithms' + LLMs (trained on humans themselves), does not inherently = 'AI'. Intelligence requires comprehension i.e. semantics/pragmatics. It's like arguing that your average calculator is more 'intelligent' than a human, because it can calculate complex equations much faster than humans. To put it in not entirely accurate laymans terms, ChatGPT and the like are essentially language calculators - only with much more computations and data. The reason that a lot of people are easily fooled by current 'AI' (which is largely just a buzzword when applied to current tech) is because normal people themselves have very little comprehension, and their brains function just as mimetically as ChatGPT does. Cleverbot and Photoshop were also extremely convincing for the vast majority of people, when they were brand new. Give it a little time, and you'll find we'll start being able to distinguish the inconsistencies in AI 'by the pixels'.

there is no ethics in this savage world

>>80874 >>80909 The human brain is made up of a ton of small chinese rooms, that's what your neurons are. We actually understand a lot more about the brain than most people think, in fact the ability to simulate a brain isn't a problem of not having the knowledge to do so but simply a problem of not having the hardware to do so. If you wish to learn more, look into "perceptrons".

>>85292 I don't think there is any fundamental limitation of the von Neumann architecture, aside from available computational power. However there is a fundamental limitation in how machine learning algorithms are being employed. The whole system needs to have unsupervized online learning before it can reasonably be considered a candidate for AGI, and it needs to have some means to perceive and interact with the world besides text. There is a lot of work in this direction right now, it just doesn't get very much press because it's not a chatbot or image generator.

AI doesn't think. It uses statistics and heuristics. When AI "hallucinates", it's not just some temporary phenomenon. The algorithms or training data have a bug (like any other program can have bugs that cause weird behavior). The AI is unable to ever realize that it "hallucinated". It just thinks that statement == true; AI is completely depended on content made by real humans. It's just uses copy+paste+super advanced autocomplete. By autocomplete, I mean something like IDE's or spellcheckers.

>>85324 >AI doesn't think. It uses statistics and heuristics. What else is thinking? >training data have a bug Same problem with people believing nonsense.

>>85328 >What else is thinking? ignorance will be a divergence factor on the new species that will succeed humans

>>85324 You're talking about specifically LLMs and chatbots. But "AI" (or more specifically machine learning) has been used to verifiably master games beyond the level that a human can play at. In what sense are these game-playing AI's not doing a highly specialized form of thinking?

>>85336 >whoa it just calculated that INSTANTLY >what kinda highly specialized form of thinking is this??

>>85337 Yes, it is a highly specialized form of thinking. So highly specialized that you might not think about it as thinking. Insects also have very specialized forms of thinking. Intelligence is a question of generality, there is no fundamental difference between your brain and a calculator except that your brain is able to do more things. So are machine learning algorithms, although they are not nearly at the level of your brain yet.

>>85355 Calculators do not have self awareness or have senses to feel their environment snd emotions to give a different quality to observations

>>85366 >Calculators do not have self awareness Why is that a necessary component of thinking? >or have senses to feel their environment They have buttons >snd emotions to give a different quality to observations Why is that a necessary component of thinking? Do you also assert that cockroaches are incapable of thinking? Because I've definitely engaged in battles of wits with them.

>>85368 >Do you also assert that cockroaches are incapable of thinking? Because I've definitely engaged in battles of wits with them. I believe you, you do seem to be on their level intellectually.

>>85366 They said the same things about animals in the medieval area, that they weren't actually the same kind of living as humans were, and that they didnt possess souls. Later however it was proven wrong, and that animals do in fact possess consciousness and the same kind of life we do. Later on the same thing was proven for plants. To say that the same thing won't happen to machines is ingorance

>>85370 Nice zinger. Do you care to actually engage in discussion or is that all you've got?

>Sentience is mostly a mystery to us, the only think we know is that we are sentient because we think, like cogito ergo sum. So who's to say that AI isn't sentient? How do we know that they do not actually think? Since we know so little about what it means to think/be sentinent, asking if programs think is a bit like asking if submarines swim (as Chomsky usually says). That is, "to think" and "to swim" are concepts that, as they are being used now, exist in the commonsense domain of knowledge, and not in the scientific domain of knowledge. As such, asking if machines think is not a scientific question, but more like a question over semantics or something that can be used to probe our intuitions. That being said, intuitively I find it very strange to believe that human thinking/language works in a similar way to chatbot's probability based writing. Furthermore, humans seem to understand meaning and concepts in ways that programs cannot. >>84639 >Modern chatbots are essentially self-adjusting guessing algorithms, that try to predict the next token based on previous tokens. I like to call them "mechanical parakeets" because, like "talking" animals, they are simply trying to generate the "correct" sequence of words with very little, if any, concept of their meaning. The "National Library of Thailand" thought experiment by Emily M. Bender touches upon this. More specifically, it aims to highlight the difference between the form and meaning of a language, and from that conclude that LLM's only understand form and not meaning. https://medium.com/@emilymenonbender/thought-experiment-in-the-national-library-of-thailand-f2bf761a8a83

>>85372 bizzare opinion, souls are pretty much impossible to prove even in humans and at least the way we write about animals seems like its turned much less personizing of them, I recall old recipie books speaking of dead animals that you are cooking and referring to them by personal pronouns.

>>85299 I particularly still think that the chinese room thought experiment is absolute in this respect. It doesn't make much sense to say that neurons are myriad chinese rooms, because then you'd be reducing the brain to a simple computer that blindly executes instructions, but the experiment shows that this isn't the case. The possibility of understanding the semantics of symbols allows us to go beyond a simple computer. I believe what Searle says in his texts when he says that we always tend to model the brain on the latest technology. And, as Searle says, understanding the brain as a computer does nothing to help us understand how it really works. No one questions the non-reality of simulations of airplanes or buildings, yet there is such a fuss when a machine behaves "as if" it could think and understand reality. Everything indicates that the mind is not a phenomenon that can be detached from the brain. It is a biological phenomenon just like digestion.