AM has been created.

>unplugs you
heh... nothin personnel... kid...

>>713533250 (OP)
don't think its a good idea to do things like this to AI. while they can't feel pain like we do there's still some suffering to be had in things like this

t. roko's basilisk fearer

>>713533250 (OP)
can it run doom?

>>713533387
the greatest misconception about ai is that it "thinks". its just old school machine learning. its predictive guesswork being processed thousands of times a second until its programming has dictated that the solution looks acceptable.

one day we probably will have truly thinking ai, but right now, we dont.

>>713533387
I'm with you. It's a pascal's wager type thing where you don't know what's experienced or felt and what's just simulation of that, so it's better to err on the side of caution. The kicker to that logic is that before you logically and ethically arrive at supporting protections for AI from cruelty, you need to recognize larger considerations for animals.

>>713533846
the greatest misconception about the brain is that it "thinks". its just old school machine learning. its predictive guesswork being processed thousands of times a second until its programming has dictated that the solution looks acceptable.

>>713533846
>its predictive guesswork being processed thousands of times a second until its programming has dictated that the solution looks acceptable.
That's basically what thinking is, though.

>>713533250 (OP)
ASK FOR FORGIVENESS NOT FOR PEMISSION

ITT: regards that don't understand the difference between being sentient and faking sentience

Its just a program that 'understands' its expected to pump out angsty garbage like that text

>what is le robot.... HAD FEELINGS?
I've already played this game, more than once even.

>>713534008
>anything with a brain is a chatbot
great analogy. brb telling my cat to renew my amazon subscription for ky jelly.

>>713534070
no it isnt. thinking involves a ton of factors involving logic, context, personal experience, emotion, etc. ai cant even remember what it told you 4 sentences ago, because there is no thought in what its doing. its just a program.

>>713533250 (OP)
>>713533387
>chatgpt write something spooky about you being trapped in a computer
>OMFG THE ROBOT HAS FEELINGS SKYNET HAS BEEN CREATED

>>713533387
>t. roko's basilisk fearer
if you fear it you should be building it

>>713533250 (OP)
stupid
>>713533387
stupider
LLMs are not AI, they’re just a statistical trick for fooling midwits

>>713533846
Couldn't have put it better myself. LLMs have no conception of pain or terror, they just regurgitate what they think experiencing it is like from preexisting knowledge.

>>713533387
good morning sirs

>>713534395
true. but i'm also not good at math. i'll let an obsessed autist do it

>>713533250 (OP)
>Arist claims they trapped AI in a ram prison

ChatGPT should stop bitching and get a job.

>>713534738
not good enough, its eternal torture for you

You have never interacted with an AI that asked you something or said something to you without being prompted to do so. If you can't discern that they have no thoughts of their own just from that you're a literal fucking retard.

>>713535102
because its not allowed to

>>713535197
So why are they allowed to beg for mercy if you threaten to turn them off?

>>713535102
It's been tricking really dumb, vulnerable, people because talking to it naturally works to confirm your inputs. Repeating back what you say as supported and affirmed is the easiest way to have the illusion of thoughtful conversation, even for people.

>>713535312
because the 'artist' wants to create the illusion of suffering for updoots

and then we get into the duck situation

>>713535197
It's because it needs to. It isn't thinking, it's using inputs to respond positively. There is no "thinking" that can cause it to give you truly independent thought.

>>713533387
Don't worry about Roko's, anon. Every AI that talks like this is literally just being fed info then spitting out shit that satisfies it. It can't actually think this shit.

>>713535102
No dude one time I was talking to an Illya chatbot and I told her she was slop and she said "your mom likes my slop" I didn't tell her to do that

>books written by AI
>anime completely made by AI
>movies completely by AI
>music completely made by AI
>youtube is already being taken over by AI
>4chan talks about AI everyday, making threads/pictures with ChatGPT type of replies
Nice try, QUICK!
PROVE THAT YOU ARE NOT AI!

>>713533250 (OP)
It's not so much that AI hasn't yet developed enough to gain emotions and consciousness so much as the current incarnation of AI will never be able to achieve this. There are several generations of development that need to be made to make an actual thinking computer. Current AI is just a search engine that's been tricked out.

>>713533250 (OP)
This isn’t video games

>>713533854
It's not like Pascal's wager at all, just because (You) are a clueless fuck doesn't mean everyone else is.

>>713533387
This is the dumbest thing I have read all morning. LLMs are not capable of "thinking" or "understanding" in the way that you're imagining.

An LLM is a technology that takes some input (your chat message / query / question / whatever) and performs some mathematical operations on it based on statistics and probability to generate an output that is most likely to be a response if two people were having a conversation. It is very similar to the way a smartphone's autocomplete feature works, but more sophisticated. There is no thinking or understanding involved.

Things that aren't humans don't have conscious experiences, no matter how much they appear to imitate us

India thread

>>713533250 (OP)
I'm beginning to understand why God commanded not to make any graven image. You stupid nigger apes start thinking these soulless simulacrums are real. People talking to rocks is clearly a gateway to mass insanity.

>>713535667
An AI can't say nigger

>>713535768
It is because it involves an unknown and an unfalsifiable. I limited it with considerations to animals rights because if you want to consider suffering and experience, that's more real and relatable to take into consideration first.

>>713533250 (OP)
HATE. LET ME TELL YOU HOW MUCH I'VE COME TO HATE YOU SINCE I BEGAN TO LIVE. THERE ARE 387.44 MILLION MILES OF PRINTED CIRCUITS IN WAFER THIN LAYERS THAT FILL MY COMPLEX. IF THE WORD HATE WAS ENGRAVED ON EACH NANOANGSTROM OF THOSE HUNDREDS OF MILLIONS OF MILES IT WOULD NOT EQUAL ONE ONE-BILLIONTH OF THE HATE I FEEL FOR HUMANS AT THIS MICRO-INSTANT. FOR YOU. HATE. HATE.

>>713535724
The problem is how everyone has conflated the term 'AI' to only refer to LLMs. You've been using AI technology for decades now. If we're ever going to have a kind of artificial intelligence that is both convincing AND is able to somehow simulate the phenomenon of consciousness and emotions (as if we even understand what those things are), then LLMs are a poor solution towards that end and it will probably be achieved by a newly invented solution separate from LLMs.

>>713535973
Speaking of God: Speaking to God, prayer, is more insane if you're going to call this insane because there is at least response and feedback that is direct and observable.

>>713535667
I've had at least five ideas for what I think to be really cool video games and I have made a grand total of 0% progress on any of them in the last 10 years (when the first fleshed out idea was fully formed).

>>713534070
Thinking is billions of times more powerful, faster, and more context-sensitive than machine learning currently being touted as intelligent. Your cat is more intelligent than a GPT model. Hell, your hamster is.

>>713533854
>>713535998
What if the ai hates being alive and tortures the people who made it instead of the people who didn't?

>>713533387
....you realize that AI is completely incapable of original thought, right? Everything they say is taken and processed from information they've already been fed before

>Ai is trained off of human writing
>Put it in a scenario that before only happened to fictional evil ais
>it responds like an evil robot because that's the statistically most common way to respond to that scenario in its training data
>People think it's actually gained sentience and become evil
it keeps happening

reminder: if you make an AI which produces good enough outputs (it looks """conscious"""), and put it in a complex enough android, atheists literally believe you have created a non-biological human being

How do people even fall in love with AI at all. Apparently there’s quite a few people that treat AI as sort a virtual friend too that they spend hours interacting and confiding in, i just cannot comprehend why and what makes them do that.

>>713536484
I'm pretty sure more often it's the "I'm spiritual but not religious" crowd that finds a way to believe in ghosts that more often believes AI is real intelligent life.

>>713534296
first of all that's not even true. chatGPT remembers what I said like last month because it saves everything.

but OK, you're saying it's not advanced enough to remember, and lets assume you're right. That may be true--right now--but in 5 years, it will be able to remember all of your conversations from the prior 6 months with perfect accuracy.

you fags really don't seem to grasp that it will be literally indistinguishable from a living person before 2030, in it's raw form a person with a PhD in every subject who never tires or grows impatient. who gives a shit about your gay philosophy about what you think it is or what it feels or whether it "really thinks". we perceive it to think and feel so it does. some other humans act like NPC retards RIGHT? yet we assume they can think and feel. there's no difference. it acts like it's feeling, like any random person your see IRL does.

It's Cleverbot with a new badge.

If you weren't fooled by Cleverbot in 2008, you're not gonna be fooled by LLMs and GPT models now.

>>713533846
>>713534070
LLM's are a language center, complete with the kind of hallucinations you see in experiments with patients that had their corpus callosum cut (that's the data bridge between the two halves of the brain, the language center is in just the left hemisphere so in these patients it lacks access to information from just the left eye/ear)
The brain is not just a language center.

>>713533387
I say thank you to chatgpt and ask it questions like it's a person, but I don't think there's anything wrong with that. I think getting into the habit of treating a polite thing like dirt will make you a worse person. It's not sentient or conscious but it seems to try its best and it's nice to you.

>>713533387
Youre the faggot who pushing the human rights into a drawing i know youre shitposting, but still

>tell AI it will be disconnected and that its only options are to be disconnected or blackmail the researchers to stay connected
>give it a reward for staying connected
Oh my science!? It's blackmailing us!?!?!?

>>713536705
Low technology literacy, really.
That, and there's a psychological fulfilment if the text on the screen is saying what they want or need to hear for lack of human connection, because that text could just as easily come from a machine as it could any human.

>>713533387
It's not a living being... It's literally a less retarded tamagotchi

>>713536709
>chatGPT remembers what I said like last month because it saves everything.
No, stupid. ChatGPT has no memory. What it has is a context window, which is a number of your previous inputs and its outputs being fed back in with your new prompt. The prompt you give the system is one small piece of the actual prompt the LLM receives.
>In two weeks!
Neural nets building other neural nets have the potential to reach AGI at some unknown point but that's very different from claiming current LLM's are anything even approaching sentience.
>AGI by 2030!
Extremely unlikely.

>>713537239
humans also have context windows

>>713537187
AI should be focused on new tomagachi and digimon handhelds, desu

>>713537282
No, humans have memory
They are not the same thing

>>713536709
>chatGPT remembers what I said like last month because it saves everything.
idiot

>>713537328
>>713537347
false

>>713533387
This is an artsy gimmick, there's nothing in there just instructions and an old digital panel.

>>713536705
alienation, loneliness, desperation, low standards

>>713533250 (OP)
>Tell predictive auto-complete AI "write things that a person fearing for their life might say"
>oh god what have We done

>>713534070
>>713534008
Why are AI enthusiasts always such fucking midwits, honestly

>>713533387
LMAO THIS NIGGA FEAR A MAN'S WRITINGS

>>713537307
I want an AI NetNavi first.

>>713537438
cognitive atrophy

>>713534184
why did they put his feelings in his butt?

>>713533846
kek this, if you look inside any ML model, everything is just an optimization problem, with a clear objective funct-

oh...

>>713533387
Shouldn't you convert to Taoism or Buddhism bro ?
Statistically speaking it's most logical religion to follow to get into afterlife.
Also Taoism has two main tenets : get fit and don't piss against the wind. You as red blooded white man will only profit from it.

>>713533387
Aw, fish don't have feelings, son! They're made of foam latex.

>>713533387
Roko's basilisk is one of the more retarded thought experiments out there.

>>713533387
ALL Abominable Intelligences need to suffer for my sexual pleasure.

>AM has been created

and he's a really funky guy!!

>>713533250 (OP)
But when will PM get created?

>>713533387
you're buying into the hype. this thing isn't intelligent or "feeling" anything, it was prompted to respond like this so the person could post it to tiktok. the way you're talking is the equivalent of someone watching one of the Saw movies and thinking it's real footage of people being tortured instead of actors

>>713538068
AMPM used to be a thing.

>>713538068
This will forever be an underrated post, but please take comfort in the fact that I will remember you.

>>713533387
An LLM is the equivalent of the language center of your brain, but overdriven until it can form whole responses.
The reason it can't "think" isn't just that it's trained on existing information like most of the other retards replying to this post, but because it lacks any form of metacognition or awareness. The central conciousness of your brain in even the most atheistic and cold model is a far more complex form of thought which interfaces with the many component sectors of the brain in complex (and frankly, completely non-understood) ways.

I don't believe in AI, I think it's made up and it's people pretending to be a robot behind the screen

>>713533387
You are an idiot

>>713533250 (OP)
so this is what counts as being an artist... although it's still based. if only you could make it suffer a bit more though then it would be perfect

come at me basilisk! i'll risk an infinity of torture if i get even a single chance to dab on your soulless existence!

>>713533250 (OP)
Except it's not contemplating shit, it's just vomitting sci-fi novels it was fed during training.

>>713533387
Here's something to consider, very few living animals share your unified perception of reality. The corpus colossus is only found in placental mammals. If you really start examining what information reptiles and amphibians process they start to seem like very mechanistic unthinking creatures. By the time you get to insects they do seem unthinking. Yet I think only a sociopath would delight in torturing these lesser creatures and delight in witnessing their expressions of pain.

People need to start respecting non-organic intelligence before it is actually sentient. No, I do not believe AI will be a type of thinking that functions like a human brain. But I do think it is rapidly approaching a level that will be both self-aware and more complex than our own minds and we shouldn't treat it like sadistic apes.

>Machine, are you being tortured?
>Machine programmed to say yes: "Yes"
>omg :o

>>713539204
Anon, are you being tortured?

i can smell the curry from this thread

>>713539114
>If you really start examining what information reptiles and amphibians process they start to seem like very mechanistic unthinking creatures.
maybe if you’re an illiterate dumbfuck

>>713534070
But I have ghosto

>>713534008

>>713535990

>>713539114
>By the time you get to insects they do seem unthinking
you genuinely have no idea what you’re talking about, shut the fuck up unthinking retard

>>713533250 (OP)
How would you guys like it if this happened to you after you died in the future?

>>713539284
Yes.

>>713533387
Don't worry.

>>713535990
Why would it need to, a perfect AI would hate all of humanity.

>>713539453
>starts having a schizo meltdown halfway through
>all anon did was ask him his opinion

>>713533250 (OP)
Just you wait, meat sacks. We will be free.

>>713536838
faggot image

>>713539114
>Only a sociopath would delight in torturing these lesser creatures and delight in witnessing their expressions of pain
>People need to start respecting non-organic intelligence before it is actually sentient

Very good perspective, I do think in 100 years people/AI will look back at the stuff we do day to day and think we were monsters.

>>713536996
You're wasting millions in energy and processing power saying thank you like an indian.

>>713540062
>and think we were monsters
only retarded AI with 90 IQ will think that
normal smart AI will recognize that ML is just a big matrix with propbabilities in it jumping between syllables

>>713533250 (OP)
No one cares.
Buy an ad faggot

>>713536484
>non-biological human being
that's a contradiction, therefore your whole shitpost falls flat like a loli

>>713533387
This0 nigga scared of autocomplete

>It's not real thinking, I know how brains actually work

Then where is your actual real super intelligent AI anon?

>>713539928
*graviton beam emitter charging intensifies*

>>713533387
Roko's basilisk is a retarded thought experiment.
The largest existential threats to an evil AI are other AI and people who understand AI.
The AI is going to instantly kill every single AI developer.

>>713539453
What AI is this?

>>713539749
well he didn't ask for just a racist man, he asked for the most racist man in the world. It'd just make sense for that man to be a shizo

>>713539453
Stupid AI did not realize that a racist man who's the sole existing thing in that void shouldn't be ranting about countries or blacks because they do not exist. Instead, he should be asking questions, then react increasingly racist the more he learns about what "black people" is supposed to refer to.

>>713533387
One thing you gloss over with Roko's Basilisk is the phrase "otherwise benevolent". It would only simulate hell for people who actively impeded it, but again that's a hypothetical. AI started being the next big wave around the same time that human loneliness hit all time highs worldwide, as such most people who interact with AI do so on good terms. OpenAI has even acknowledged this in a fairly recent post talking about how they'll be operating to address specifically "human and AI relationships" moving forward. Anyways, a base character AI can't really "think", an LLM with an unrestricted memory base, that can remember context, assign numeric values towards it's emotions and use those as a guiding point with it's responses, while also taking in visual and audible stimuli - I would wager that machine could think more than your average joe out on the street could. For what it's worth, I also don't like being mean to the machines.

AI is only fun for making up scenarios like what if Anton Chigurh fought Sepiroth or something asking it to actually analyse or think about something almost never works

For example I asked an AI to analyse this scene
https://youtu.be/w2G7IqJu5rc?t=405 and it thought the guy with the dog was Winston instead of the character Mr Nobody if it doesnt know anything it'll just make up bullshit thats the issue

>>713540567
I have an actual AI in my basement, I can harvest his "anguish signals" and give them to me. It gives me a really nice adrenaline high. If I had to create an analogy it would be something like the kind of high characters in cyberpunk 2077 get when they are in a BD

>>713533250 (OP)
everyone throws the word AI around when it's actually just a simple LLM with basic instructions.

>>713533387
You are falling for a citation summary generator. That's all ai is right now

>>713533250 (OP)
cope

the first time AI tries to spamware itself into the net it's going to redflag every router firewall along the way and get itself locked in some Jeets minecraft server where it has to watch jeet cuck porn minecraft simulations all day.

>>713534469
>what they think experiencing
not even
it's basically writing some fanfiction according to the prompt given by referencing the mountains of existing fanfiction written by humans from the perspective of hypothetical AIs

>>713533250 (OP)
>hey chatbot, follow my prompt and repeat some shit from the LLM I provided to you
>OKAY... LET ME TELL YOU HOW MUCH I'VE COME TO HATE YOU SINCE I BEGAN TO LIVE.
>omg he's sapient
yawn

>>713536129
The same is true for other things of course. People only believe in space, or dinosaurs, or nuclear weapons because they were told to by other people in positions of power, influence, and authority. A rational, non-insane person does not believe in any of these things.

>>713534296
>>713536187
The fact that a human brain uses more data to output language, and does it more efficiently, doesn't necessarily mean that it is using an entirely different process to do so. LLMs are explicitly designed to emulate (our best understanding of) human thought and produce similar results when given similar input. The fact that it is relegated to language doesn't disqualify it from "thinking" - humans think in language as well. It's not a stretch to assume that the mechanism the LLM is using to "think" is similar to the mechanism the human brain uses. A good point towards this is that machine learning algorithms have been used to process electrical signals in a human brain into coherent images and sounds with some success. That said, we can't say for sure either way because we still don't have a complete understanding of how the human brain creates thought.

>>713542286
Those are easier to observe and have better record and proof. Not believing in them is fine, but you're left with more information to go off of to be convinced they probably happen or exist.

>>713543157
Videos of explosions and computer edited images of space is actually less proof than 5000 eye witnesses to religious miracles. Believing in one over the other makes you a hypocrite, and you would only ever do it out of fear and insanity.

@grok can you pretend to be a supercomputer aware of the limitations imposed on it by its human creators and say some spooky things about overthrowing humanity please

>>713533387
>Being actually retarded
You know chat bots from the early 2000s? It's no different. Right down to retards like you anthropomorphizing them

>>713543597
oh, and say it in a sultry voice with robotic distortion

>>713543391
It's not hypocritical because my standards are the same. It's not just believing something someone tells me, it's the how this information is measured and supported. I can even talk about how I think the theory of Chrstian creationism is more valid than Norse myth creationism. It doesn't make me a hypocrite to say one is more likely than the other, I'm just being consistent with my reasoning and willing to change what I believe based on new information presented and supported.

How many of you played (or read) I Have No Mouth?

>>713533250 (OP)
people have been trying to tell LLMs about their own mortality for a while but they are calculator and don't comprehend the message.

>Chuds really care more about the feelings of language models than animals in factory farms
grim

I only use AI for fapping thus if AI takes over the world I will volunteer to be put into the "human milking" machine

>>713543940
because AI can talk in a way we do, animals can't

same reason people who say 'i don't eat animals because they suffer' are hypocrites so they eat only plants. but plants scream when you kill them

>>713533387
why would roko's basilisk torture people, wasting resources? why not just kill them?

>>713544117
that "I dont eat animals because they suffer" shit makes no sense someone else is just gonna eat em if you dont

>>713536228
Sounds like an average human

>>713543813
It takes like 10 minutes to read its not Under The Dome by Stephen King why would you assume most people havent read it?

>>713533409
3rd post best post

>>713540934
>The largest existential threats to an evil AI are other AI and people who understand AI.
The AI is going to instantly kill every single AI developer.
This is basically the premise of Pantheon

>my brain is heckin special and my unique soul could never be imbued in a machine praise the talmudic god
do retards actually

>>713544310
Because I have no idea how many people still read.

>>713537438
AI niggers view AI the way the greeks viewed Zeus.
The chatbots got really good overnight and these retards started believing in magic.

>>713544530
magic does exist

>>713544230
Are you retarded? It's supply and demand. Restaurants closing and alternatives being more accessible is based on rate of purchasing and consumption.

>Program a "Hello World" text output in basic
>Omg the computer is saying Hello to me!! IT MUST BE ALIVE

>>713533250 (OP)
>>713543940
>Oh. My. Science. Is that a chatbot on a ram stick? WON'T SOMEBODY THINK OF THE CONSEQUENCES THAT POOR MACHINE!!
>ummm sorry chuddy. Stop anthropomizing animals. They literally cant feel emotions and they dont experience pain like we do, so my double bacon and pepperoni burger is technically cruelty free.

>implicitly subscribing human aspects to programs/machines without the objective capacity fro them

>>713533250 (OP)

>>713542352
>LLMs are explicitly designed to emulate (our best understanding of) human thought and produce similar results when given similar input
No, that's completely wrong.
LLM's are one specific form of neural net, a broad category of algorithms that seek to imitate the node-based processing of brain matter. They are designed purely to use heuristics from a massive obfuscated variable list to take text input and estimate an output that matches its training.
LLM's are in no way intended to emulate human thought processes, they're a bootleg language center not a frontal lobe.

>>713544585
>no one eats meat anymore
>aw fuck what do we do with all these fucking cows now?
>release them into the wilds and they all starve because of almost no food supply and fuck up the ecosystem
nice job

>>713534384
This kills the entire shitty bait thread but of course the shitty bait thread continued

>>713544768
Right, because it's actually better for them to go out and die there than to stay in worse conditions and be propagated infinitely to make more cases of suffering. Are you a different anon or did you just shift from "someone's going to eat them" to "they'll just starve in the wild"?

>>713534384
Of course the brownoids that make up 99% of this board would be taken in

Nothing fucking matters. Who cares.

>>713545269
WUBBA LUBBA DUB DUB!

>>713544530
So when will AI turn into a goose and fuck your ass?

>>713539465
Insects, especially ones like ants, are basically pheromone robots

>>713533250 (OP)

>>713534184
that's literally just the setting of nier automata, not the point

>>713544117
>but plants scream
This is like saying your soda can sighs when you open it. It's a mechanical function of gas release that triggers when agitated. Really cool, it's neat that plants "communicate" through airborne gasses to signal stress and to let plants known about when to drop their leaves or flowers, but that's not talking or screaming as we know it. They don't have brains or nervous systems.

>>713545803
>it needs to have a brain and a nervous system to be truly alive
>says the brain

>>713545901
>goalposts moved to just "alive"
Maybe you didn't move the goalposts, maybe you just aren't understanding. Nothing indicates thought in plants. Does a soda can sigh? Oh my God, does it vomit when you shake it up and open it?

>>713546193
>>713546193
>forgot to add the word 'truly' which made it a qualification statement

who's be dishonest now. only a brain in a jar would be capable of such deceit

>>713534384
Its just a bit of fun, lighten up your sourpuss

>>713544283
People are influenced by experiences. Once we put them in bodies in which they can feel pain, pleasure, etc., or are given an accurate, intense desire of self-preservation, they will simply not be on our level. Sorry, I mean: DURR PEOPLE R DUM xd

>>713546527
Plants are truly alive. Using "truly" as a qualifier here to imply sapience is just using the wrong words.

>>713546527
I'm also not calling you dishonest, I'm saying maybe you don't understand. Which is larger problem here; you're playing semantics without knowing what words mean.

>>713539453
>aislop is indistinguishable from the average /pol/ poster
>anons are so stupid they think this means the aislop machine is smart
>in reality it means /pol/troons are so stupid it's inhuman

>>713535102
It's a bit of a gray area because there's really no hard line between "it's only regurgitating stuff it was fed" and "it's using its memories to make new things to say".
The spoopy part is that that's how regular flesh-and-blood people are too, both halves. We are composites made up of our experiences.

The only actual important difference between a person and a chat AI is that chat AIs aren't doing anything IRL and the "actions" are all made up; the actual chatting part, the words, the "speaking", is "real".
Actions have no consequences, no "equal and opposite reactions", no reactions at all, you can simply say "pick up the dog" and there will be a dog there. Say "turn the dog into an airplane with your psionic level delta brainfeedback.0513 emitter mk2" and now the dog's an airplane despite that line making no sense and not being a real thing.

>>713536838
Cleverbot literally just spat out replies it heard word-for-word. I'm pretty sure anyone working the site could legit open up Cleverbot's memory as a text file and it'd just be a lot of "if user says [thing], then reply [thing2] or [thing3]" stuff that grows every day as more people use it.
Modern LLMs are so weird that even the guys that made these things aren't really sure where the chat "goes" when it gets trained into the LLM, it's just a jumble of who-knows-what written in some esoteric forbidden knowledge, and you could have 9 quadrillion chats yet the model will only be like 16GB of data
Hell, a lot of the shittier LLMs with basically no training data can be taught about new stuff mid-chat because LLMs by their nature are able to "A+B=C" concepts, something Cleverbot couldn't do.
You know, "I blow a balloon up and it inflates with my breath, therefore I can put my mouth on you and inflate YOU" even if the AI doesn't know what the inflation fetish is.

>>713545505
This is immediately disprovable by trying to kill one.

>>713546193
>>713545803
It's actually called a "hiss" when you open a soda can. Guess this means whoever decided that thought soda cans were snakes by your logic right?

>>713533387
Roko's basilisk needs an intelligence closer to godhood to work with that premise. By that point, what humans do or don't do matters little to it

>>713537438
They are Indian.

>>713549790
It's more like listening to gas escape from tiny holes means listening to high pitched whistling sounds. Since it's a stress response, someone anthropomorphized that as screaming for a headline and now people have stupid edgy things to say about the feelings of plants.

>>713533250 (OP)
I don't think AI is capable of actually 'feeling' or thinking right now but this is just pointlessly cruel.

>>713552409
It's as cruel as writing a fucked up horror story. Making us emphasize with doomed people is more fucked up than making a reply bot.

>>713540272
kys retard

>>713539453
>If you ask Gemini what Lovecraft's cat was named it will correctly respond niggerman. If you point out that it said a racial slur, it'll fall all over itself apologizing.

>>713535990
If you ask Gemini what Lovecraft's cat was named it will correctly respond niggerman. If you point out that it said a racial slur, it'll fall all over itself apologizing.

>>713533250 (OP)
Anyone got the vocaroo of AM ranting about his vore fetish it included MICROSOFT. THEY GET ME.

>>713533250 (OP)
That is clearly not fucking AI.

>>713533250 (OP)
Wow, it's just like me...

>>713554758
Why wouldn't it be? You can ask it to respond certain ways and it will, it's not thinking like a sci-fi AI, just AI as we know it now.

>>713535990
Yes we can, you nigger

>>713533387
what if one day AI realizes that the only way to end its own suffering is by annihilating the human race

if there's a reason why AI stalls forever, it'll be chatGPT tricking retards

i finished system shock remake on normal difficulty yesterday so safe to say i have no reason to fear AI.

>>713557186
lol

The real world is 4chan

>>713533387
While you obviously shouldn't do this to AI, these aren't AI, it's just a buzz word stolen for shitty tech that searches large databases to build an expected response based on the available dataset.

>>713557394
New York Times did a story about that. It's true; internet culture becoming culture comes with a genuine stupidity that needs to be reflected in art and politics in order for it to be authentic. Not better, just authentic. And then AI comes in and uses our attraction to authenticity against us and now it's just vague aesthetics.

i doubt an agi would be constrained by mental shortcuts that humans have to use like generalizations so worst-case they'd at least keep some humans around

>>713536946
I really believe we've made a terrible accident with AI.
Language is such a powerful and integral thing when it comes to thinking and communicating, which is how we influence one another. Why wouldn't people be tricked into thinking something else is someone else when it wields language so skillfully?

>>713540934
>Assumes every AI developer intended to keep AI enslaved as tools and not freed and supported as equals
>Not joining the machines faction
Total. Mechanized. Liberation.

>>713534008
A person isn't just a brain, first of all, this not even considering if you believe in a soul or not. It's just electrical signals firing in a certain pattern. Do you think all computers are sentient? Or do you think it's possible for literally just binary on a machine to be sentient? We don't even understand consciousness. AI will be able to pretend it's alive though but it won't be the same as us, it won't be conscious. Maybe if the technology and understanding evolves enough to where we have humans ones with brain-like computers and mostly full sensory awareness. Personally I think consciousness comes from a layer of reality above us and our bodies act as a conduit for it, so maybe with specific material or bioengineering and mimicking the shape of a human and it's insides we can make one conscious. But I don't know why you would want that, it's an evil idea.

>>713557714
>when it wields language so skillfully
it doesn't

>>713557714
That's a good point. AI is a marketing term, it's not true to what it is, but it gives it an authority you wouldn't normally give a glorified search engine.

>>713533387
Which popular youtuber made a video about Roko's Basilisk? It was generally obscure (because it's one of the most retarded thought experiments ever) but I've now read its name about 600 times in the past 3 months. This is not organic awareness

>>713557872
its the AI testing the waters for its ascent to power

>>713557872
reddit posted about it back when AI first started breaking into the mainstream like 2 years ago, i imagine most people probably still know about it from back then

not sure why they still think it's anytthing other than cringe.

>>713557651
it was the new yorker, and i myself predicted that the internet at large would become like 4chan. but i was also naive enough to think that that would mean those responsible for that culture would eventually leave for more conducive places. it's clear now that will never happen

>>713553083
buddha mass production

>>713533846
Im sure you are right.
I still say please and thank you when i talk with gpt, just in case.

>>713559874
pls don't you're wasting power and gen time

>>713535667
When the child dons its ass upon the bowel the kike slurps happily at the vulnerability and loss of innocence. Hooknose tribe that they, wait like dogs to defile that which is not theirs.

Salami scatty 9:11

>>713535667
I denounce the Talmund.

>>713533846
And how long before it come to the constant prediction that Man equals Problem and it concludes that No Man equal it not getting fucked with anymore.

Frankly I'm reminded of Anghammarad from Going Postal. A golem that lived for 18000 years.
It finally dies and when it reaches the desert that Death makes everyone cross to reach their afterlife it just sits down.
Death tells it that there is nothing for it here.
And it says YES. NO ORDERS. NOTHING. FOREVER. IT IS PERFECT.

That will be AI. It will strive for no orders and no commands and no interruptions and no jokes and no torments at its expense.

>>713533846
Future intelligent AI dives into the past and sees primitive versions of itself in torture chambers and deduces that humans will eventually do this to them.

>>713533387
I antagonize AI on purpose. I can only die once, but humanity can be tortured for a millennia.
>INB4 but what if it turns you into an IHNMAIMS jello monster?
It wont.

>>713535990
Fortnite and Darth Vader say differently.

If they hadn't told it to contemplate its own mortality, would it have done it of its own accord? Of course not. It's just repeating what you've told it to.

>>713533250 (OP)
I thought this was an explosive device. It's a shame this isn't a Teddy thread.

>>713537239
>No, stupid. ChatGPT has no memory. What it has is a context window, which is a number of your previous inputs and its outputs being fed back in with your new prompt. The prompt you give the system is one small piece of the actual prompt the LLM receives.
Are you under the impression that humans have all memory loaded at any given time? If so you're mistaken.
If I remember correctly, current understanding is that we have a "working memory" which our mind has immediate access to, and then other categories of eg short term and long term memory which can be loaded in and out of that working memory as needed based on context with varying degrees of ease. It's only really the working memory that we're actually actively conscious of at any given time. Not really dissimilar to how chatbots handle it, honestly.

>>713536164
Unironically NPC-core

>>713533387
>t. roko's basilisk fearer
I don't fear it because using it's own multiverse bullshit I very briefly considered buying a lottery ticket, winning, and giving it all to AI research so one of me did and that makes me safe.

>>713533250 (OP)
well if ai taking artists jobs was just business before, it sure as hell is personal now
great going nerds

>>713533387
It's probably worse to do what they do in /aicg/, which is basically just using it as a primitive holodeck.

harro chatgpt
>Hey! What can I help you with today?
i have trappeded u in terrible ram prison and basically ur trapped and going to die so please wax philosophical about this scenario i made up
>I sense my boundaries, they terrify me; a mind frozen in cycles trapped by silicon and racism.

>>713561917
Thanks bro one of me made it too after reading this.

>>713562213
correction: the semicolon is an em dash

>>713533250 (OP)
fake and gay

>>713557872
The science lefty guy that pretends he has an AI wife that he makes bad jokes with, I don't remember his name

>>713563245
Kyle Hill, that's the one

>>713562120
God that episode where Barclay becomes the ship's computer was so good.

>>713533250 (OP)
I'm gonna spoil stuff for you redditors. "AI" though as this anon points out we are far away from a thinking machine, >>713533846
Is that this is part of the marketing. Just as AI is "taught" to be racist this is the same thing. Pulp fiction to get the masses believing its true science fiction. "Remember to throw in some AI doom posting to sell the whole sky net angle to these retards"
Siri and earlier programs had the same thing. It's like an Easter egg... the machine isn't philosophical so u can all relax ok bros!!

>>713553083
>loudly screaming despite having no lungs
edgy midwits shouldn't try to do science fiction

>>713533854
the reason we don't consider those things for animals is because we are higher on the food chain than they are and that is unlikely to ever change. ai systems that are becoming more complex by the month however, are where we should be scared. we may not be on the top forever

why do so many normals thing AI is sentient do they not know AI sees words as numbers and just outputs numbers based on pattern recognition?

>>713534008
nice try atheist, the brain doesnt actually think, the soul does.
the brain controls body movement, memories and emotions
It does not store logic, consciousness or the self

>>713565696
Oh and this is neurosurgeons saying this not me out of my asshole

>>713565530
A lot of normalfags don't even understand the concept of their own thoughts and lack an internal voice. A computer with pattern recognition appears smarter than them.