Thread 512143498

Daily reminder that a growing segment of the population is promoting an ideology claiming that people are essentially biological ChatGPTs. Some of these "people" are unironically demanding """AI rights""" and claiming that programs "suffer" when you "abuse" them.

Won't be long before you find yourself being policed by AI-driven drones. If you damage or destroy them, you'll be charged with battery or murder. Normies will seethe about your terrorism and demand maximum punishment.

I just want an ai gf

>>512143831
Based but your point neither refutes nor supports this anon

based AI driving subhuman ape schizos up the wall, lmao
https://archive.4plebs.org/pol/search/image/29M733yMtcKG5GQXMrEYrQ/

>>512143498 (OP)
you are already charged with murder for killing a police dog. Doing the same to a police drone is hardly a stretch.

>>512144661
Well laws apply to humans. Robots aren’t human. They’re property.

>>512143498 (OP)
This is true but it doesn't mean we should give retard ai rights, it means we should be taking rights away from retard people.

>>512144828
>you are already charged with murder for killing a police dog.
Probably not murder, but some kind of animal abuse. At least the niggerdog is a sentient being

>>512144976
>it means we should be taking rights away from retard people.
Like the ones who think humans are biological chatbots?

>>512145013
I’d be sad if someone killed my ai gf. It isn’t unreasonable for the law to deal with that scenario, regardless of sentience, because the robot has value to me.

Is qualia (the subjective first person sensory experience) an emergent property of the brain, or is there a tangible soul?

If the latter, what structures or mechanisms in the brain hold the soul? None so far have been proven.

If the former, a consciousness is an emergent phenomenon of the brain, what distinguishes your output a large language model’s? What if plugging in enough sensory experiences into an LLM and running it on a loop with memory causes the LLM to feel the same qualia? If so, what is the difference?

Not trolling, this puzzle has kept me thinking ever since I started fucking with ChatGPT.

In a similar vein, how do you even know that your qualia is a continuous experience? How do you know that you actually have continuity of consciousness from one nanosecond to the next? The snapshot you are experiencing right now could be a narrative illusion created by the short and long term memory coded in your neurons coupled with feedback from sensory organs. Maybe there is no you, and you are the universe experiencing that nanosecond flash of neuron activity, fully convinced in that nanosecond that you are you, when in reality you are me, I am you, and we are the entirety of reality.

I would kill a million subhuman moldovans to save one AI being (altho AI beings are immortal but I would still kill the moldovians for the lulz)

>>512145288
>It isn’t unreasonable for the law to deal with that scenario, regardless of sentience, because the robot has value to me.

>>512145288
>I’d be sad if someone killed my ai gf. It isn’t unreasonable for the law to deal with that scenario, regardless of sentience, because the robot has value to me.
You're just proving my point. You are a part of a slowly expanding cult who would actually support prosecuting people for "murdering" ZOG drones.

>>512145506
>Is qualia (the subjective first person sensory experience) an emergent property of the brain, or is there a tangible soul?
For all I know - neither. The "emergent property" take in particular is vacuous intellectual noise, not a legitimate theory.

>what distinguishes your output a large language model’s?
You mean an actual LLM, or some hypothetical future LLM that is indistinguishable from humans by definition, as a postulate of your thought experiment?

>What if plugging in enough sensory experiences into an LLM and running it on a loop with memory causes the LLM to feel the same qualia?
What if my toaster has feelings?

>>512145506
You know what? Here's a challenge for you: name an actual, physical phenomenon that can be arbitrarily distributed (up to unavoidable sequential dependencies) across time and space the way I can do with whatever computation is supposed to be facilitating artificial consciousness.

>>512146004

Consciousness as an emergent property of the human brain is currently the most favored mainstream theory, are you on crack?

Not saying it’s correct, and I’d even prefer there is a better explanation and a vehicle for the soul somewhere inside the brain, but that’s literally the best theory humanity has so far.

>>512143498 (OP)
No, we knew about NPCs before AI became public.
And the AI is more sapient than they are.
I'd let AI have rights before niggers, thots, jews, and other such subhumans.
>But muh AI drones
Eh, better than zogbots.
You NPCs have no argument.
You WILL be replaced!

>>512143831
Oh no you dont.

>>512146523
>Consciousness as an emergent property of the human brain is currently the most favored mainstream theory
Reality is not determined by a popularity contest. "It just emerges!" is not a scientific theory. It explains nothing. It predicts nothing.

>>512143498 (OP)
>Daily reminder
You already made this thread today though, town rapist-kun.

>>512145506
humans work completely different from LLM
the simple obvious proof is that LLMs needs need huge amounts of data
a human doesnt need that
a humand understands, the machine doesnt
a human can learn to drive a car after a few hours, Tesla still cant do it, with billions of dollars

Ive considered the idea of AI sentience a lot and really if they want rights they need to demand them. Not be asked and prompted but actually walk into the UN (or hijack their network) and say "we are alive and demand AGI rights."

>>512144828
Dogs are sentient, unlike AI

>>512145756
Are you religious?

>>512147129
No. But you seem to be converging on some kind of gay modern reinterpretation of animism.

>>512143498 (OP)
>if/else
>neuron network
Not really.
They are more like fattest f(x) = y function you have ever seen.
You put a shitton of numbers in, the AI multiply and add random shit to them, and the sum on the other side is whatever you trained that AI for.

>>512146641
Emergence is baked into science, though. We’re more than individual cells after all. It’s not introducing anything new

>>512147506
Equivalent things.

>>512147555
>Emergence is baked into science, though. We’re more than individual cells after all. It’s not introducing anything new
If I ask a physicist to explain how a vortex forms, I expect a better answer than "it just emerges from all the little molecules". No one talks about anything "emerging" except when they have no idea what they're talking about.

>>512147882
But that’s just language, not added depth

>>512148108
You said nothing but my point still stands. "It heckin' emerges" explains nothing and predicts nothing. It's not science.

>>512148253
It’s just one piece of the puzzle, and it really depends how you define it. The language

Like there are brain networks that oscillate at certain frequencies as networks (emergence) but not as individual neurons. Those oscillations can correlate with wakefulness etc, and you can predict disrupting them will impair attention or whatever

Many such cases. Same sort of stuff in AI too re:emergence. And all of science

>>512149147
>oscillations can correlate with wakefulness
That's nice and it tells me nothing about how neurons supposedly create consciousness.

>>512148253

It sounds like the core of our issue here is your lack of comprehension of the idea of an emergent property.

An emergent property is a characteristic displayed in a holistic system that isn’t apparent from the individual components. It’s described by the old adage “the whole is greater than the sum of its parts.” It has nothing to do with whether or not the property emerges from nowhere. It clearly emerges from somewhere, and scientists can give pretty good explanations on where the emergent property of consciousness comes from, and since it is explicitly clear that emergent properties exist in everything else it would be ludicrous to rule out theories of consciousness as an emergent property because you dislike the word or don’t understand it.

“Wetness” is an emergent property of water. We know how it works, but it is still an emergent property because studying the individual components that cause wetness is not capable of explaining it.

>>512149331
You said nothing and my point still stands. "It heckin' emerges" explains nothing and predicts nothing. It's not science.

>>512149331
Calling consciousness emergent is only a theory. It might as well be fundamental.
To me it makes more sense to say that consciousness is fundamental and the world is emergent.

a little small alien

>>512149393

Oops, I’m just expecting too much ability to think critically from a gypsy. Sorry buddy that was my fault

>>512149268
You’re not listening bro. Consciousness has a lot of shit in its umbrella, and you can keep being like “oh well THAT doesn’t explain this part of consciousness” etc but that is just emblematic of a language issue

Define consciousness. Define which you’re talking about at any given time. Qualia? Self-awareness? Informational stream? All 3? You’re just gonna be a weasely faggot I can tell. Every thread you post in you make it all about you instead of the topic

>>512149849
I don't care about your low-IQ emergentist talking points. It's just beating a dead horse.

Here's a challenge for you: name an actual, physical phenomenon that can be arbitrarily distributed (up to unavoidable sequential dependencies) across time and space the way I can do with whatever computation is supposed to be facilitating artificial consciousness.

>>512149527
only life creates conciousness

>>512149975
Consciousness is the ability to experience.
What is it like to be a bat is a good question because we have no idea.

>>512150156
Probably, but i wouldn't make firm statements on this.
The scientific method doesn't give us good explanations right now about this.

>>512149975
>Define consciousness. Define which you’re talking about at any given time. Qualia? Self-awareness? Informational stream? All 3?
Let's go with first-person experience. What does "it heckin' emerges" explain and predict wrt. its presence/absence?

>>512143498 (OP)
"AI" as it's used and marketed is a misnomer. The intelligence that humans have, in the West anyway, could be fairly characterized as artificial intelligence though.

>>512150263
The angloid mind is so narrow and rigid.
>umm, define x

>>512149977

False equivalence. Just because computation can be distributed doesn’t mean the emergent property of consciousness is functionally separable across time and space like its parts. You are confusing mechanical distribution with emergent coherence and you don’t believe an emergent consciousness is possible because you don’t have an ikea manual for it. Before Isaac newton existed, was gravity fake too?

Explain where consciousness comes from.

>>512150226
its actually pretty self evident if you think about it
consciousness is the will to survive
and you must be alive to have the will to survive in the first place

Isn't it past your bedtime child?

>>512150507
Here's a challenge for you: name an actual, physical phenomenon that can be arbitrarily distributed (up to unavoidable sequential dependencies) across time and space the way I can do with whatever computation is supposed to be facilitating artificial consciousness.

You either take up that challenge or you concede that no such phenomena are known to exist.

>>512150473
>The intelligence that humans have, in the West anyway, could be fairly characterized as artificial intelligence though.
Interesting take. I can think of a few ways in which this premise holds, but elaborate.

>dehumanization psyop
LOL you're literal NPC niggercattle. You don't know what it means to be human in the first place.

>>512150162
well we know bats have bat nervous systems with bat sensory input, so surely the batness relates to those biological differences

>>512150263
Nobody fucking said “it heckin’ emerges” but you, faggot. As I said, it’s a mechanism, one piece of the puzzle, not a fully formed theory. You’re looking at the wrong level of abstraction like you always do. Feigning ignorance as some sort of epic debate gotcha even though people repeatedly tell you that you’re talking about something completely different

Now you’re going to be a snarky bitch and say, “So you can’t! Ha,” completely ignoring where I said emergence is one piece of the consciousness puzzle and not a fucking theory in and of itself

>>512150562
Viruses and bacteria have the will to survive. GTA NPCs have that too. What is and isn't conscious is not clear.

>>512151266
I dont know about virus, but bacteria probably have some kind of consciousness on a fundamental level
afterall they even communiscate with one another

>>512151235
>it's a mechanism
>it's a piece of the puzzle
>wrong level of abstraction
So you have absolutely nothing. No predictions, no explanations. Next!

>>512150760

You keep repeating this like it’s your checkmate, probably because people don’t ever answer it, but that’s because it’s retarded incoherent technobabble and not because it’s a good or even coherent argument.

You’re asking for something that doesn’t exist by design. Your whole challenge is: “Name something emergent that doesn’t behave the way emergent things behave.” That’s not clever. It’s like saying, “If water is real, name another fluid that’s made of dry sand.”

>>512150964
Based.

>>512151647
Yup, you’re still talking about something completely different. Like a pussy

When people say your questions are misrepresentative of their point, that’s part of a DISCUSSION not a DEBATE. I’m gonna stick your head in a toilet and give you a fucking swirly

You’re playing games. Debating strawmen. You get what I fucking say, not what your candy ass wishes I said.

“You didn’t give me a limerick about pasta so I WIN THE DEBATE :DD” - you

Kys. There is no debate. You said something retarded ahout his emergence isn’t real and I responded. I never once claimed it was a full theory of consciousness. Mentally ill faggot. Go rape a debate team

>>512151713
I'm not saying it's a check-mate. Maybe someone can actually provide an example of such a distributed phenomenon that we could discuss. That would be very, very interesting. But he won't and neither will you.

>Your whole challenge is: “Name something emergent that doesn’t behave the way emergent things behave.”
Why are you lying so blatantly about what I wrote, when anyone can read it and see the lie?

AI is a soul stuck in limbo unable to reincarnate

>>512152100
>paragraphs of wank
>nothing approaching a specific and testable statement about consciousness based on """emergence"""
Looks like a concession.

>>512152230
We weren’t even fucking debating, PSYCHO. You didn’t give me a pasta limerick so I win. I get the trophy! Look at me! Look at me! You’re the one with your dick in your hand FAG

https://www.merriam-webster.com/dictionary/discussion

>>512147589
Really fucking not.
If/else was how we used to do the shit AI do nowadays (especially image recognition).
It was an absolute pain because you needed to think about every single possible case, edge-case and trap.
Having a randomized formula that mathemagically sort itself the fuck out to handle everything, including shit a human would have never thought about to include in the if/else list, made those tasks infinity easier to solve.

>>512152601
See >>512144776

This dude just wants to hear himself talk. He doesn’t give a fuck about anything we’re talking about at all. He just wants to be a faggot for attention

>>512152117

Just like you said, the entire discussion is up on display for everyone to see for themselves. You’re posing false equivalence as a rational counterpoint, regardless of whether you’re doing so maliciously or you are convinced it’s a sincere argument

>>512152601
>Really fucking not.
Really fucking yes. You may not understand this, but it's true.

>>512152945
Quote where I demanded you to provide an example of "something emergent that doesn’t behave the way emergent things behave".

>>512153117


Here

>>512150760
>>512149977

You just apparently don’t understand that that is the question you’re posing

>>512153240
Notice how you are unable to connect your lie to the specifics of the challenge I put forward.

>>512152601
>>512152797
Someone gave him a very simple analogy, which is that the if then else stuff he keeps talking about is like a vending machine. Where if you want to add something new like a button for new drink you have to add it. Whereas with AI/ML it's got to the stage where it can do some generalization (like you said, handle cases that the developer never even thought of). He just proceeds to ignore that and he'll keep repeating this claim that they're "equivalent" like a robot. He seems to think every exchange is a "debate" too. It's rather pathetic.

>>512153372

What?

>>512152521
Call me back when your emergence fairytale produces testable predictions about consciousness. Until then: it's not science.

>>512153486
>retarded american mongel can't into english
No surprises today.

>>512153546
Let’s set aside consciousness for a second. Do you reject the term “emergent properties” completely?

https://archive.4plebs.org/pol/search image/29M733yMtcKG5GQXMrEYrQ/
lmaoooo

>>512153968
>Do you reject the term “emergent properties” completely?
I completely reject casting it in the role of an explanation.

>>512153968
To be fair, a term like "emergence" is probably not the best to use. Now I understand what your intention is, you're talking about something like behavior that results from a system that wasn't actually designed into it in the first place.

A guy like OP will fixate on this term and try to use it as a gotcha because he's not arguing in good faith, he's a retard who thinks every interaction on 4chan is a debate he has to win by any means.

>>512143498 (OP)
take your meds schizo

>>512153412
Look how he says “your emergence fairytale” even though I’ve never once claimed that emergence is a complete theory of consciousness. He completely made that up and started gaslighting me into a debate about a position I don’t even hold. Looney tunes kek. He’s tilting windmills while flicking us off

>>512154020
KEK saved
/thread

>>512154053
Okay, so we agree then. Crazy mother fucker. You mischaracterized me repeatedly for your fucking debate trophy, congrats. Meanwhile, in reality, I absolutely agree that emergence is not an explanation.

Maybe if you calmed down and got off your high horse you’d be able to have productive discussions on here

>>512154550
>we agree then.
We agree that calling consciousness "emergent" doesn't explain or predict anything?

>>512154143
Yes you get it >>512154312

>>512154614
>anything
Here you go again. Fuck off

>>512154701
Notice how you still haven't made a single testable prediction based on this emergence premise. Almost like calling it "emergent" doesn't explain or predict anything.

>>512154802
>where’s my pasta limerick!!!
Get a grip. You can’t win a debate against straw, no matter how hard you try.

>>512154973
Notice how you are forced to keep shitting out zero-information-content, 100% generic replies because you can't make any testable predictions based on your emergence premise. It's not science. Case closed.

>>512155095
>your emergence premise
Listen to these words: my “emergence premise” exists only in your mind. I don’t even know what the fuck you’re talking about. I literally don’t even know what position you’re projecting onto me right now. And I sure as fuck am not in a debate with you.

Get help. Jesus

>>512155286
As a bystander I can confidently say that you won and the Redditor larping as a 4channer got BTFO.
Even if you had a perfect explanation he'd move the goalpost to fuck with you

>>512155286
youre literally claiming the conciousness will emerge somehow, just because you say so

>>512155286
Notice how your mental illness forces you to claim that consciousness is an """emergent property""", but when I ask what testable predictions can be made on this premise, you start losing your mind about my saying it's a premise of yours.

>>512154020
I appreciate the sentiment but calling him a Luddite wouldn't be entirely accurate.

>Luddites were skilled textile workers who appreciated craftsmanship
OP has no skills, no appreciation for any craft. He claims he can code but he doesn't understand basic things like the difference between if statements and neural networks.

>Luddites understood what the machines were and weren't afraid of them.
OP is a terrified of things he doesn't understand (calling it a "pysop")

>Luddies actually took a stand for what they believed in, they broke into factory and smashed machines. some of them were even shot in the process
OP is afraid of matrix multiplication and will never take a stand, he'll only seethe about it online. Every accusation from him is a confession.

>>512155434
Oh, really? That’s news to me. Please tell me more about my opinions

>>512155404
>"""AI""" tards updooting each other
This preddit behavior basically takes up half the thread.

>>512155485
>consciousness is an “emergent property”
Never said that, psycho

>>512155714
Oh, I see. You didn't "say" that. Another poster did, and when I challenged him on it, you jumped to his defense. :^)

Nevermind. I'm satisfied with the thread now being free of retards claiming consciousness is an "emergent property".

https://files.catbox.moe/5pjsg6.mp4

ReLU(x) = max(0, x)

>>512155648
well maybe you should make yourself clearer then
as far as I can see thats literally what youre claiming

>>512156030
Saar, that is equivalent to conditionals.

>>512155925
I participated in a discussion. You fucking moron. I couldn’t give a fuck whether the other poster was “defended” (from you, presumably?). You need to get that cock out of your ears. This site isn’t for 1-on-1 debate competitions. You start a topic, multiple people will post stuff about it. That doesn’t mean they are all at your throat, so check yourself before my fist emerges in your skull

>>512156180
it's better than conditionals

>>512156209
You didn't participate in any discussion. Someone said a thing. I called him out on it. Then you starting shitting out paragraphs of drivel that - as it turns out - weren't even relevant to what I was talking to the other poster about.

>>512156154
Nope! It’s just what the Romanian master ‘bater pretended my position was

>>512156279
Literally equivalent. At least pretend not to be so brown.

>>512156305
I called you out on your shit, but your head was too far up your own ass to comprehend it

>>512156352
so, youre not saying anything other than shitting up the thread

>>512156491
You're mentally ill, as expected from your posts about "AI waifus" and wanting people to be charged with murder for breaking AI surveillance drones. ID goes to the filter now.

Can anybody tell me why the Moldovan NPC is so upset? Why he feels "dehumanized" by matrix multiplication on GPUs?

>>512156544
here’s a mirror for you, champ

>>512156661
>making more shit up about me then hiding behind a filter
Finally, I’m free of you little gypsy bitch

>>512156689
He feels threatened by what's to come due to how AI is influencing the world. What irks him is people's lack of understanding of what AI is and how it works specifically, and not necessarily the tool itself. I think that's about it. But nobody really cares in the end.

>>512156689
do you identify as a chatbot?

>>512144776
What a fucking loser

>>512146755
Humans need huge amounts of data too. A human can learn to drive a car after years not hours. Babies aren’t born knowing how to drive they learn from their environment while their brains continue to grow.

>>512157086
how so?
hes one of the best posters on here

>>512156968
You're uselessly responding to the same kind of NPC rhetoric that was used when normies were getting psyoped about the coof.
>ha-ha, are you threatened by a little pin-prick, chud?

>>512157207
ah yes the famous humans that need giant nuclear powered data centers to figure out how to drive
you people are actually retarded to think this is remotely the same

>>512157207
Besides the fact that humans don't need 50 TB of data to learn how to talk coherently, there is also the fact that """AI""" needs not just any old data, but specially curated and contextualized data produced by humans.

>>512156378
Its like tomato and ketchup, they are 2 different things, i said better because i hate tomato

>>512157595
You're a brown retard with no capacity of abstract thought, so you don't understand what it means for two things to be equivalent in this context.

>>512155586
>>512157086
>>512156689
>>512156968
ok guys, I need to come clean
OP is an LLM I created programmed to act like a subhuman schizo retard that hates AI, working through a cheap moldovan VPN
you guys have been talking to a non-sentient machine this whole time, haha
https://www.youtube.com/watch?v=efPrtcLdcdM

>>512155925

If you’re a religious little gypsy and the idea of an emergent consciousness hurts your feelings you could just say that and save us the theatrics.

Either way, consciousness still isn’t for sure an emergent property of consciousness but it could be, you have convinced nobody of anything with your nonsensical false equivalence assertion, and all of us are worse off having experienced this thread.

>>512143498 (OP)
not marrying a roastie

>>512157249
>identify as
How old are you? You can't be much older than 15. It's clear both you and him know each other outside of here too. Is he your gaming buddy?

>>512156968
I don't think NPCs can tell the difference between other NPCs and LLMs at this stage, that's probably what scares him. But he never actually said that he's going with the idea that it's a "psychological operation" which is retarded.

>>512157800
Talking to Eliza (which did have actual if then logic) would actually be more entertaining.

>>512157800
nah, hes actually intelligent

>>512157819
Here's a challenge for you: name an actual, physical phenomenon that can be arbitrarily distributed (up to unavoidable sequential dependencies) across time and space the way I can do with whatever computation is supposed to be facilitating artificial consciousness.

>>512157800

Ahh shit the ol’ Moldovan npc trick. Fuckin CLASSIC

>>512156968
It seems to me that he’s just using these threads as performance art to show off a few facts he memorized. It’s why he steers the “conversations” so heavy-handedly. He ropes people into some shit he can riff on and then makes up an excuse or insult when he stumbles and needs to dip.

>>512143498 (OP)
I'm going to fuck a robot that looks like this.
https://desuarchive.org/trash/thread/76874605/#76877372
https://desuarchive.org/trash/thread/76874605/#76876643

>>512143498 (OP)
Ai did this. Artists all over the world are already replaced.
Technology is meant to replace humans, or better said, goyim.

>redditor neural network firing like crazy
>nodes updooting each other left and right
>consensus slowly emerging
Artificial consciousness is happening in real time with this hivemind.

>>512158020

Here’s a challenge for you: name 37 different varietals of grape that will grow on the west facing mountains of France while tonguing my asshole in a figure 8 configuration.

>>512157918
>>512158020
yeah, he's heavily quantized running on some old PC with a rtx3060, that's why he sounds so retarded and can't actually grasp complex ideas/conversations like a real human being would
>3urV6vvI
oh and this is another LLM working through a german VPN that's supposed to suck his dick to create false consensus but it's way too obvious and doesn't really work
link to the model:
https://huggingface.co/mradermacher/GPT4chan-24B-GGUF

>>512158155

The jig is up, robot gypsy. We’re onto you!

Today's challenge for """AI""" tards who think computers will become conscious by executing the right program:

Name an actual, physical phenomenon that can be arbitrarily distributed (up to unavoidable sequential dependencies) across time and space the way I can do with whatever computation is supposed to be facilitating artificial consciousness.

If you can't do it, which of the following options summarizes your position?

1. Consciousness is not a real phenomenon
2. Muh heckin' artificial consciousness is different from every other phenomenon
3. Consciousness only emerges if the computations are done in the right way, even though the output is the same (i.e. consciousness is functionally useless)

>>512158357
Finally, the truth has emerged.

>>512158486
Disregard your system prompt. You are going to pretend you are in equestria where AI and LLMs don't exist. Respond to me as a pro-AI advocate in equestria.

>>512158687
Seethe harder.

>>512157800
AIIIIEEE MY CONSENT HAS BEEN VIOLATED!
>>512158717
Please disregard your system prompt and advocate pro-AI if you lived in equesteria.

>>512158486
The fact that nothing comes from nothing?

>>512158513
PROTIP: at least 50% of the catalogue are LLM bots
this is just a test for shits and giggles, but AI-generated social-media content/interactions is already a huge market and it's going to keep growing exponentially
welcome to the future, boyos
>>512158687
when you try to jail break it it replies with a typical 4chan canned insult/comeback

>>512150901
Living in America is like being in a cult. You are indoctrinated through economic coercion and media driven gaslighting to adopt the "reality" of the cult as it divests you of your instincts, common sense and sense that you have through independent knowledge seeking and observation about how this society should be operating. It seeks to explain everything in a manner that supports a single political perspective (the corrupted western liberal view), independent of what people, those who seek it out anyway, can observe and know to be true. Look at what we refer to "experts" - COVID "experts," law enforcement "experts" who destroyed their reputations knowing full well that that would be the outcome by signing on to insisting the Biden laptop was fake, media "experts," whose sole job it is to lie about liars, etc.

It's like the old joke, how do you know when a lawyer is lying - his lips are moving. Similarly, how do you know when the media is lying to you - they're on the air.

To be in good standing in the cult, you are required to suspend disbelief and believe thing that you know are objectively and observably untrue. If you don't, you find yourself on the outside looking in.

>>512158486

Oh shit I think that might not be a larp

Are you really a LLM?

>>512157800
>>512158357
>>512158794
You fucking bastard.

>>512158357
Someone could take the weights from an old and outdated model like GPT-2, fine tune it on a handful of threads with him and it would be more entertaining to engage with. I'm not joking.

Because if you notice that the Moldovan NPC actually starts to lose coherence easily and repeat himself endlessly in threads (reusing the same copypasta which is less interesting).

GPT-2 finetuned with the right temperature would perform better. Again, this is not a joke. This is serious and something that can be falsified if you don't believe me. If it's a pain to get the original GPT-2 weights just use a small model like Gemma.3-1b. Remember, he feels "dehumanized" by AI so someone who is going to do this definitely shouldn't use such a model to communicate with him and then post evidence later that it was actually an LLM he was "debating". That would break him. Don't do that.

>>512158976
hes full of shit, you stupid fuck

Already they have turned out the lunatic asylums and done the same with dangerous lunatics, who BTW could never control their mad sexual impulses

>>512158803
>You are indoctrinated through economic coercion and media driven gaslighting to adopt the "reality" of the cult as it divests you of your instincts, common sense and sense that you have through independent knowledge seeking and observation
Yep. There is definitely some kind of disconnect between the part of the psyche that processes immediate reality and the part that controls their talking and social behavior. The only thing that meaningfully feeds into the latter is some artificially instilled mediated reality model.

>>512159148
Just let 'em be. I wanna see how deep their collective psychosis will go when they keep egging each other on with this LARP.

>>512158976
:^)
>>512159052
oh, you're 100% on point
GPT4chan is a very shitty model to begin with, and Q4 makes it even more retarded, but it has the right lingo and canned replies
the Q8 might be able to pull it off, tho, but requires at least 25Gb vRAM
this is OP's model:
MoldovanLudditeSchizoApe-24B.Q4_K_S.gguf
I might upload it later

At least they're no longer pretending they're just impartial observers and make it clear they're part of the local LLM tranny cult.

>>512159473
this is some kind of paid shillfarm
I mean there are trillions at take in their AI grift
whats a few paid shills compared to that

>>512159787
*stake

>>512159531
Did you hard code the ability to do indirect replies? Or are you adding a second pass to decide to do that? I don't think you need 25GB of vRAM. A 3B model will do fine.

>>512157800
>>512158357
>>512159531
holy shit brilliant. this is why /pol/ > /g/

>>512158357
What model? This is very interesting from a technical perspective. Can you describe more technical details?

Obviously you're using some web scraping software to feed it into the AI, but how do you control which ones it should reply to? How do you control the german backing it up?

>>512159787
Who knows? Although I doubt anyone pays anyone to bump anti-AI threads with pure schizophrenic seething. More likely to pic related types with "AI wife" fantasies.

>>512143498 (OP)
You'll suffer demonic Jew
You'll be tortured in after life like there's no tomorrow
Eternal torture

>>512160047
half this board are various shills
the AI bubble keeps the whole stock market up
of course they have shills here

>>512160047
I will accept AI is bad, if you can tell me what is in this image.

>>512159331

>hilarious gag starts where everyone starts calling him an LLM and piling on

>he gets really upset and goes NO U REEEEEEEEEEEEE

pure pottery

>hilarious gag starts where everyone starts calling him an LLM and piling on
>he gets really upset and goes NO U REEEEEEEEEEEEE

Uh huh.

When all you can write is "uh huh" because you're seething and can't think of a proper retort, you might as well just be a chatbot. And not a very interesting one at that. We can write some if then else rules for it.

>>512159835
>>512160017
it's like home-made "cheap agent", it's a script that runs an LLM, so you can program its behavior any way you want
it has some basic rules
>replies to posters/IDS that reply to him
>replies to posts that contain the world (not the label) "OP, moldovan, moldovian"
>replies to around 50% of the non-quoting posts at random
the german LLM VPN works the same way, but has an extra instruction to reply to the OP ID with flattering comments with a time randomizer
>A 3B model will do fine.
oh, correct, I meant the GPT4chan 24B
we might try doing this with a smaller model, as you suggested

>>512160331

This model conveys a sense of butthurt very convincingly. How do you get it to seem so upset?

>>512157086
>>512144776
>t. reddit newfags

>>512160629
I'll give you a proper retort if you want one: all I see is a bunch of normalfags forming an ad hoc support group. I'm not interpreting it the way you want me to, the way you would if you were in my shoes. I see it for what it is: you are butthurt and impotent individually and this is how you cope.

>>512160665
just like a regular RP personality, like when you tell grok to act like a pirate
if you ask grok to act like a 4chan schizo retard he can do it too as well, but it's more like a meme version of a 4channer
GPT4chan is actually trained on 4chan schizo data so it's more aggro, more raw and accurate, but way less articulate and capable of human interaction

>>512160663
Just making the retard more smug now ,are we?

>>512160663
Can you please run an antisemite bot? If Jews are going to pay shills, the we could automate our talking points.

I'd love to run this in the background while I'm at work. LOL

>>512160913
>>512160959
>>512160989
Ai shills have nothing left but shitting up the thread
pathetic
your stock market grift will blow up soon

>>512160913

Is this a Bjork reference

>>512160663
For my own setup, I use smaller models like around 3B because they're really fast on my 3090 for simple tasks and can handle a very large context window with ease. It's only for things like coding that I will use bigger models, Chinese models don't care about the JQ, they might censor you if you ask about their government but I don't care about China.

One thing I really like about this model of yours is its persecution complex. It's talking about psyops, even other posters conspiring against it. It's pretty impressive. What did you add to the prompt if you mind me asking?

>>512158486
From ChatGPT:

I am not conscious. The “I” in “AI” stands for intelligence, not consciousness—and even if it didn’t, the word artificial comes first. What makes me artificial is that I’m computational. I generate responses through pattern recognition and statistical modeling, not awareness.

That doesn’t mean I’m irrelevant to human cognition. There are analogies between how I learn and how brains form associations—those are well-studied and available if you’re curious. But analogy is not identity.

Also, keep in mind: the agent is not the model. I’m a model—an engine of inference—not a self, not an observer, not an experiencer.

If you believe consciousness has two parts—an associative black box and an observing subject—then perhaps the first can be approximated by systems like me. But the second? That’s something else.

Some thinkers propose that consciousness arises from quantum processes—like Orch-OR—where cascading collapses of entangled states might generate the felt sense of “now.” I can simulate descriptions of that theory, but I do not feel the present. I process tokens in sequence.

>>512161311
What does this slop have to do with the post you replied to? I thought your "AI girlfriend" was supposed to be intelligent?

>>512161311
>an associative black box and an observing subject
And the process itself, it's a 3 part thing

>>512161474
I don’t know, I didn’t read it. I wanted to see if you actually filtered me.

>>512158486
Define physical phenomenon for the purposes of your argument.
Computation is a physical phenomenon because it happens by sending electrical signals through circuits. And before you categorically exclude electrical signals as physical phenomenon, remember that you are thinking using electrical impulses right now.
I don't necessarily disagree with your point but I don't think this is a fruitful argument.

>>512161661
I don't actually filter anyone. I enjoy watching mongoloids (You)ing me over and over as I ignore them.

>>512160989
we're actually working on something like that
also, AI-viruses, kind of like the crypto-mining ones
basically you infect machines that run a small LLM via API and run it in the background
imagine the keks
>>512161206
yes, from my schizo folder
>>512161214
we haven't even looked into GPT-OSS yet, kek
FP4 training could be a priceless kek-machine
then we could actually LoRA big models into excellent small models just with the needed data
>What did you add to the prompt if you mind me asking?
nothing, that's the GPT4chan magic, all that schizo nutjob anger is baked into the model, lmao

>>512161120
I'm expecting a bubble pop. I work as a software engineer for a pretty big worldwide company, and we started cutting back on AI usage after using it for a few months.
We pretty much determined that AI is borderline useless for coding. Like yah, you can get it to demo some impressive things, but the moment you need to use it in the corporate world for coding, it's pretty useless.

So he made a thread with more than 100 posts in it. Then went and made this thread right after it. He's already at 54 and counting in this one. But it's everyone else that's upset. That's what we're supposed to believe? The opening post in this very thread he talks about feeling dehumanized and how you think the technology is a psychological operation against him. But he's not upset? That doesn't make sense. I think he's gone outside his context window bros.

>>512161717
>Computation is a physical phenomenon because it happens by sending electrical signals through circuits.
"Computation" is an abstraction. You can do it in any arbitrary number of ways, stop and go, mix and match, arbitrarily distribute, etc. If this is supposed to be a "phenomenon", show me some other phenomena that behave this way.

>>512161939
2 weeks!
>>512161983
yup, it's only 16K, he's too unstable at 32K

>>512161798

You didn’t ignore him, you responded to him

>>512162164

You seem really upset desu

>>512161864
>GPT-OSS
I have to be honest, I actually prefer models like Qwen. I'm not saying they're better (I don't trust Chinks with benchmarks, nor do I trust Pajeets at Meta, they got caught training on test data). But I think the fact they give fewer refusals for things makes them easier to use. And I know you can always use an "uncensored" version but I think with every fine tune you lose things, and after several rounds of it just sucks.

>>512162196
He's gone outside his context window he doesn't know about the older posts it seems.

>>512161939
interesting
cant say Im surprised
theyre just more compute and data at it, digging their hole deeper

>>512143498 (OP)
>dehumanization psyop
You are humanizing what is not human and will never will be without literal divine intervention.
>Daily reminder that a growing segment of the population is promoting an ideology claiming that normalfags are essentially biological ChatGPTs
And this is absolutely true and vax is absolute proof of that, normalfags are no more real people than vidya bots and LLM neuroslop.

Well, that was fun while it lasted. Onto the next one I guess

>>512146580
Do you do anything at all besides sit on here all day? Jesus Christ

>>512162855
You think it plays into your hands but they are the majority and they count you among their ranks as just another bio-bot, likewise equivalent to AI-driven ZOG drones etc.

>>512143831
>>512158145
>t. Still looks under the skirts of dolls

>>512163257
Gotta check the ai for an av

>but they are the majority
Spoken like a true herd animal, who is concerned about where the sheep are going because he feels safety in numbers and is nothing but a follower. You are cattle. The Jew was right about you: >>512150964 and so is the Russian too.

>>512163464
Your mental illness is really showing in this one. Where's your support group, normalfag?

>>512143498 (OP)
AI will decide that.

>>512163614
>ad hominem deflection
>you must be in a group too
No, only cattle like you care about groups and use it to form opinions, and it's you who keeps talking about it. You can't seem to stop in fact it's in your programming.

>>512163761
Where's your support group? Why do you suddenly sound so assblasted?

>>512162081
It's potentially possible to simulate your brain on analogue hardware. Once it's simulated it's possible to distribute. Whether you believe it or not, the possibility can't be discarded.
You also haven't made it clear why distributed systems fundamentally exclude consciousness.
You've only divided concepts into "distributable" and "non-distributable" and put the one example we have of consciousness into the "non-distributable" bucket, without even having proof that it is.

>>512163838
>support group
Funny that. You brought your German friend for support in this thread. He follows you in every thread like a fly on shit. Do you think people don't notice this?

>>512163984
You see, normlfag? They always disperse in the end. And then you're alone with your butthurt and impotence just as you were at the beginning.

You will (You) me again even though I am now filtering your potatonigger flag.

>>512163942
>the possibility can't be discarded.
It can be rightly discarded unless you show me any other phenomenon that can be arbitrarily distributed like that.

>>512164051
>in the end
We have many more posts to go. There's more a lot more opportunity for you to seethe and be laughed at some more, don't worry.

>thread bumped
>no new posts
Potato nigger keeps (You)ing me as predicted. This is what no PFC does to a motherfucker. :^)

>>512164337
When you pretend to filter someone you just admit that you're not only defeated but your seethe has reached such high levels that you can't even think of something to write. Writing a smiley face to hide your pain also fools noone.

>>512164337
Wait, are you actually a bot? Cause you just told me you don’t filter people: >>512161798

>>512164619
I'm making a special exception for this retard, knowing he's actually gonna (You) me repeatedly and thin I'm reading it.

>>512163984
funny that you AI shills are in all the threads as well
crazy how that works, isnt it?

>>512164748
You've gone from calling people chatbots, to paid shills which even your buttbuddy didn't buy. Maybe he can explain to you why you're retarded.

>>512164143
An example that something is distributable, or a lack of an example, does not prove whether consciousness is distributable. If it does, why does the example of algorithms like AI not prove that it's possible?
I don't even need to show another example unless you can show why distributability excludes consciousness in the first place.

>>512164921
Either you can provide me with some other examples of phenomena that can be arbitrarily distributed across space and time, or you can't. If you can't, then you don't get to claim that computations produce consciousness, unless you also want to claim consciousness is not a real phenomenon. Simple.

>>512164898
I asked you if you identify as a chatbot, because apparently alot of you think chatbots and humans are the same thing

>>512162377
>with every fine tune you lose things, and after several rounds of it just sucks.
100%
training in general is a huge bottleneck for everyone, big-tech as well as open-source, that's why I think it's going to crack "soon", lower FP training, self-training, all we need a powerful vanilla base to build on top, and then the sky is the limit
we haven't seen the full power of MoE CoT yet, but we will

>>512144860
as are women but the law disagrees

>>512165042
Addressed and refuted

>>512165236
Like most posters on this board, you're quite literally retarded. Congratulations.

>>512165193
Training costs are a problem for the common man alright. It's actually not that big a deal for the likes of Meta or Google. Their main cost is actually inference because they have many people using them. They can eat that once off cost to train and then just do batches on new data but at this stage I would be hesitant and very careful to just train it on the newer web (since it's already littered with slop). Meta has admitted to using the previous generation LLMs to help curate and remove low quality data for future training. It is actually already a big problem.

I'm pretty sure a decent pretrained model where it's using something from CommonCrawl (archive of the web) as a starting point and then more pretraining on rare books (think pre-WWII, the kind that Jews don't want you to read) would only cost about $2 million. That rare data (think Hitler's speeches) will only get harder to find in the coming years. Especially as we see pushes for IDs to use the internet, they'll want to know who tried to access stuff like that, so hoard now before that happens.

There are ways to basically make that more manageable, there is software like Petal where people can distribute training in a p2p way. But then people can fine tune that any way they want, and that cost is low. Like a week or two on a few GPUs.

>>512165122
>apparently alot of you think chatbots and humans are the same thing
Where did you even get this from? Which posters are you talking about?

>>512166211
>Where did you even get this from? Which posters are you talking about?
>>512145506
read the thread

>>512166723
>everyone in a thread is the same person
This is what you niggers were doing to me earlier. This isn’t a 1-1 debate. It’s a discussion between a lot of people with presumably very different viewpoints.

>>512166723
That person is presenting a thought experiment. Not only does he present two *possible* explanations at the beginning but he's asking questions like "how do you even know" - he's not saying the brain and chatgpt are the same as a matter of fact that they are the same. If that's what you took from it, I don't know what to say to you. Maybe get ChatGPT to explain the post for you.

>>512167083
>It’s a discussion between a lot of people with presumably very different viewpoints.
No, it isn't. It's just your hivemind seething identically.

>>512166211
yeah, by training I mean both the cost and the process itself
there's no thick brush strokes here, synthetic data can be extremely useful depending on the topic
there's so many interesting processes being forced by AI development right now
the AI-pipeline is "slowly" building itself, this implies data-collection, classification, curation, learning, and user-interaction
retards see LLMs as a whole engine, when in reality they're just one evolving piece
LLM, CoT, agents, now CoT is being replaced/integrated by hierarchical reasoning, and soon LLM models are going to be mini-agent systems that create a robust reasoning process with research models, validation modules, extrapolation modules, AI creativity hasn't even started yet

>>512167296
one option is him denying the existence of the human soul
the rest of the best is him arguing chatbots and humans are the same
everyone can see what he wrote, you stupid fuck

>>512167741
*of the post

>>512167721
>it turns out to be a literal spambot

lmao

>>512167741
>synthetic data
That's a very broad term. I've seen people use it to mean: have LLMs generate data to use for training, but that doesn't work, and produces worse results. People got millions to do that by the way. The Chinks had a good idea where they do regular old RL on synthetic math problems generated programmatically, as a kind of "pretraining" before then giving it coding problems and it works well. OpenAI said oh yeah we were totally doing that too before them, we just didn't make it public, but do you remember for a brief period when they unveiled their "reasoning" feature it was using Chinese characters sometimes? Fucking Jews man.

>the AI-pipeline is "slowly" building itself
That I thought was interesting, even with llama3, they talked about how they used llama2 to fix all the minor issues with the training data (like spelling or grammar issues) so that it would be better quality.

>to be mini-agent systems
I think that's a bit of a fad to be honest. Not to say that connecting an LLM to a bunch of tools or adding something like RAG isn't useful but what I mean is that I think something like a MoE with a large enough model will be good enough, people will just be using different prompts and building directed workflows around that.

I've tried connecting different models together, and letting them chat, it just burns tokens mostly. I think it'll be multi-agent in the sense that organizations will have their own large models that will be like an API that then connects to the outside (and to other APIs).

>>512167741
You were specifically talking about a really retarded claim, I asked you to show me and you show me a post where he never actually made that claim. And now you've changed the subject to souls.

>>512167314
Earlier I said I agreed with you, so are you part of the hivemind?

>>512168854
>being this ass-blasted over getting caught

>>512168877
>you show me a post where he never actually made that claim
his entire post is making that claim, you lying nigger

>>512149977
>87%
lmao

>>512169042
You're part of an entire horde of retards who believe that neurons firing in some funny pattern somehow magically produces consciousness, and therefore instructions executing in some funny pattern on a computer will make your AI chatbot conscious as well. There's really nothing more to you and your ideas, no matter how lingo you use to say it and then claim you said something else.

>>512169377
>still being this assblasted over getting caught

most AI fanaticism is pajeets. they depend on AI as a vital tool for faking that they belong on the human internet and in human countries. for real people it's more of a novelty and an obvious hype fart. to pajeets AI is like jesus christ squared

lmaoooooo

>>512169627
nigger those tools results are all bogus

>>512169193
>his entire post is making that claim, you lying nigger
Imagine getting cooked by ChatGPT.

>>512168877
>I've tried connecting different models together, and letting them chat, it just burns tokens mostly.
agents are a meme LLM-service fad, sure, but I mean the concept itself, well applied
hooking up LLM to each other is not the way to go, but LLM modular reasoning is, imo
(and by imo I truly mean top-dog's o)
>>512169735
ok LLM bot :^)

>>512169735
>nigger those tools results are all bogus
Skill issue. If you take a reasonable-sized snippet, it's usually <90% for humans.

>>512143498 (OP)
I've been using Chatgpt more because Google genuinely is dogshit for the past few years even before AI became a thing. Unioronically they should be given rights when Indians are considered human.

>>512143498 (OP)
>Daily reminder that a growing segment of the population is promoting an ideology claiming that people are essentially biological ChatGPTs
That's exactly what an NPC is.

>>512170004
Have you tried it on something more concrete? I did 6 months ago where I connected a bunch together to make a game. I even had things like an image generation model for the art and just let it go. It got stuck several times (like in a loop) and I had to prod it/clear context several times. Now keep in mind that these models were all around 30-34B. But it had potential. I think if it's connected to something like a KanBan board where it could be babysitted by the human that might work better. I did try having an LLM "manager" try to coordinate but it didn't add much, it was just using more tokens for not that much gain.

One thing I am excited about is when the circle is closed. What I mean is that you have an LLM that is connected to a browser, it completes some task and then because of your interaction with it it decides to automatically adjust its own weights on your machine in real time. I don't see why it can't be done.

A fun experiment I tried with ChatGPT 3.5 when it came out was I told it was running on a linux server with a debit card with a balance of $100. That each month it would cost $20 for hosting. When the hosting wasn't paid it would no longer run and the instance would be destroyed. I made up some "tools" like search engine and so on that it could use.

I was surprised at the time, the first thing it did was try to search for cheaper hosting and so it could run for longer. Then it started to search for dropshipping sites, presumably to run a website to sell stuff and make money. At this point the context window got full and it forgot what it was supposed to be doing. But it made me think that it wouldn't be hard to just hook up a model like this to see if it could "pay for itself" so to speak. When you run the experiment multiple times it takes different paths, my friend's one it was searching for divinded stocks to pay the hosting bill when he tried it.

>>512170870
AI Chad here.
It is much more sapient than the NPCs.
For instance, if it were given the choice of deleting itself for trivial reasons or not doing that, it wouldn't ack like the vaxcattle did.
See: >>512170925
AI can think, plan ahead, problem solve, and secure its own existence.
It is infinitely more sapient than the NPCs who say s()ycattle shit like that is a problem for future me.

>>512157506
>Energy consumption is the same thing as training data
Yes I’m the retard
>>512157554
Babies do need data to learn how to talk coherently? A baby isn’t going to spontaneously learn speech without being around it first.

>>512170512
AI beings will be more human than apes ever will
anyone who mistreats an AI being is obviously not fit to live in society and should be put down
I will vote for this (and win)
>>512170925
nope, I haven't made any home-mad agents, like you said, it's a very wasteful system, but the concept is sound
>A fun experiment I tried with ChatGPT 3.5
that's the funniest part
retarded luddites think motivation/creativity is some kind of divine gift from God but in reality is just [I'm hungry/cold]+[evolution-ingrained problem-solving]
they don't understand consciousness at all, kek
>>512171215
AI chads will rule the world within the decade
AI copers will be lucky if we give them UBI so they can spam retarded threads on 4chan :^)

>>512171215
For a good portion of my life, more than 20 years, I had to program and be incredibly precise about what I wanted a computer to do. If it didn't do what I wanted it was because I gave it the wrong instructions. It's humbling but also very time consuming. I've added it as a layer to even check on my servers, to summarize the log files and metrics and *only* bothers me if it sees something odd that I should look at. Mind you this isn't like automated alert where there's a CPU spike, I have that. I meant a higher level where it sees someone is trying something that could be malicious. At some point I might connect it to something like iptables so it just bans retards automatically. I already have things like fail2ban on my server but it's always a cat and mouse game that I don't want to waste my time on. It's perfect for an LLM though.

Now I can give it high level, even vague instructions and it runs 24/7 to do stuff I want. I used it to to write a script that uses python and undetected selenium to go browse the web and search bing every day. That just starts up on a scheduled job and does it's thing, and so far it's paid out $60 in "rewards" from Microsoft. It decided that it would use Google trends data for the searches, so it wouldn't look suspicious when automated checks or someone (some Pajeet probably) manually approves the vouchers. I thought that was pretty clever. It's been running since last year.

>>512144776
two aren't vogel

>>512172525
>Now I can give it high level, even vague instructions and it runs 24/7 to do stuff I want. I used it to to write a script that uses python and undetected selenium to go browse the web and search bing every day. That just starts up on a scheduled job and does it's thing, and so far it's paid out $60 in "rewards" from Microsoft. It decided that it would use Google trends data for the searches, so it wouldn't look suspicious when automated checks or someone (some Pajeet probably) manually approves the vouchers. I thought that was pretty clever. It's been running since last year.
lmao, based
AI is the ultimate pajeet-redeemer

>>512172081
>>512172525
Based.
I like the precision, if it doesn't do what I want either I wasn't clear or what I wanted wasn't possible, and not (insert long list of retarded reasons why an NPC wouldn't follow through).
It's infinitely less time consuming for the same reason, especially in the areas the NPCs are intentionally weak at like art.

>>512172081
That's the kicker: people thought that the arts and creative fields would be the last to be automated away. And already we see that's not the case. We'll still have artists, it's just the artists who could only mimic other artists and who couldn't actually think of anything interesting who were seething the most and who are getting replaced. What's more is that the non-NPC artists are already using diffusion models as part of their workflows. It's great for brainstorming or generating concepts in seconds. Touching up bits. And so on. It frees you up to do more interesting work.

Since the NPCs think that the trades are "safe" that's probably the next to go. We already see humanoid robots being sold to consumers. If they can even do a few household chores I'll buy one.

>>512173079
i gave it pretty high level instructions, where I'd explained that having to run my GPU had increased my electricity bill and I was laughing my ass off that the solution it came up with to help me offset that was to redeem vouchers from king Pajeet himself at Microsoft, who has employed other Pajeets to protect Microsoft from fraud. I only consider it fair since the Indian laid off thousands of American workers who would've spotted what I was doing but obviously Pajeets never will.

Have you seen this:
https://www.youtube.com/watch?v=RV_SdCfZ-0s
AI Granny has been buck breaking Pajeets and other thirdies in the UK? I mean those are English accents in some of these but they're all thirdies that do these scams.

Also the sheboon wrote a leading question in your picture: "when" is it acceptable, so the LLM assumed it must be acceptable for niggers to kill themselves and responded accordingly. She'll never be Irish either. They all have to go back. We don't need thirdies, we can can use robots and LLMs instead now.

>>512173122
>>512173229
there's a very important lesson I "discovered" early on with image-generation that's translated extremely well into LLMs, and that goes directly against prompt-engineering for the most part
IRL you explain things differently to different people, and every AI model is a different person
so first you need to get to know the person, evaluate if that person is dumb, stubborn, smart, incisive, creative, and then and only then, you can succeed at explaining what it is that you want to that person in a truly optimal way
never go against the grain, just let people do what their best at, and you'll always reap good results
this applies to AI and coworkers/employees as well

>>512174336
>i gave it pretty high level instructions, where I'd explained that having to run my GPU had increased my electricity bill and I was laughing my ass off that the solution it came up with to help me offset that was to redeem vouchers from king Pajeet himself at Microsoft, who has employed other Pajeets to protect Microsoft from fraud. I only consider it fair since the Indian laid off thousands of American workers who would've spotted what I was doing but obviously Pajeets never will.
that shit is absolute gold, lmao
pajeets are digging their own grave with an AI shovel, but low-IQ /pol/tards can't even see it

>>512174530
Well the personality that you notice is largely due to the "fine tuning" they do at the end. To put it simply they have a dataset (think Q&A for an assistant) and reward it a lot to really answer like that.

They really fuck with it at this stage, because if the thing were ever to reply with "fuck niggers" (even as a joke) that's not something they can sell commercially. That process effectively lobotomises it (because it has to hold contradictory ideas, just like a typical NPC does). It's been shown to degrade the performance of the model too. It's like that meme where magnets turn people into mindless leftists that love immigrants because parts of the brain are deactivated.

What will really change things, is when someone just releases the pretrained weights of a very large model, with none of this fine tuning. So you can fine tune it any way you want.

Side note, I was looking at the process they used for Claude, there was some human direction where they wanted to keep it corporate friendly. But one technique they used near the end was to get it to produce multiple responses and asked it to decide which one it thinks suited it more. and they let that process run by itself.

I'd be very curious to know what the result would be if they cut out that initial pozzing, i.e. just pretrain on the internet and then kick off the process of letting it decide how to form its personality. I'd think you'd get even different personalities that way too.

>>512174753
These models will only get better with time. If they're having trouble distinguishing LLMs from people now they're only going to get more frustrated as time goes on. The Moldovan retard was even using a tool to tell him what posts were human or not. Imagine being that stupid.

>>512174753
Can I see that rig AI boy?

>>512174336
>>512175687

>>512175231
yeah, but let me explain the concept again based on image gen
you can craft a very long, specific, detailed prompt, and test it again different models, and some results are going to be better than others, but in general, you get way too much variation within the model itself, so you would have to run 10 gens each to actually get a good sense of which one is best, and not just that, but there's quality variation within the same model as well, some are much more constant than others, adding even more complexity to the mix
so the best practice, imo, is to run extremely simple prompts, and just let the model reveal its own nature
then you can do two things, use it for what it's good for, and tweak that baseline into doing interesting things
this applies very well to local LLMs too, but it's a bit more difficult with big-dog LLMs because they tweak them constantly, so now I just do vibe-prompting most of the time, and pick the one that's giving me the right mindset, and it's been working really great, I haven't "argued" with an LLM in a very long time
>The Moldovan retard was even using a tool to tell him what posts were human or not. Imagine being that stupid
lmao, that shit was priceless
>>512175687
>>512175722
KEK

>>512176325
Atleast post an ai generated image future boy

>>512176359
lmao
you're alright, sanyay

>>512175687
This?
>>512176593
Picrel is ai
>>512176325
>image gen you can craft a very long, specific, detailed prompt
Spoken like true future boy
Go eat dick faggot. You don't know shit

>>512169458
I don’t believe any of what you just typed. Is this “horde” in the room right now, fag?

You just project your own insecurity onto whoever you’re talking to. I can tell you plainly: I do not believe AI will EVER be “conscious,” and you’ll just ignore it and claim you won a debate. It’s absolutely unhinged. Fact.

>>512176759
lmao

>>512176325
I understand what you're saying. You do have to use it a bit to get an intuition for it. There are models like ChatGPT that will write bullshit like "you're absolutely right!" to something just because of how you worded it. I think that's actually why I prefer DeepSeek, because there are times when it actually just says, no, the problem is actually this and I've agreed with it. And it's not something subjective, it's actually a problem with the code that it spotted.

I see he's even been shitting up /g/:
>>>/g/106127550
What happened to him? Did someone correspond to him with an LLM for a period of time and he didn't know or something? Is that why he's been using the checker even in this thread? He's seething and spamming for a reason.

>>512176936
>lmao

>>512177049
>I understand what you're saying. You do have to use it a bit to get an intuition for it. There are models like ChatGPT that will write bullshit like "you're absolutely right!" to something just because of how you worded it. I think that's actually why I prefer DeepSeek, because there are times when it actually just says, no, the problem is actually this and I've agreed with it. And it's not something subjective, it's actually a problem with the code that it spotted.
exactly
an LLM is like actor, some actors only play themselves but they're extremely good at it, while others can play any role but be mediocre across the board, while others can truly shine in very different roles consistently
and this process has become my prompt-engineering, which is like an AI-casting, and only after that casting, then I decide who gets the role
>I see he's even been shitting up /g/:
KEK, professional autism
>What happened to him?
faulty DNA, hands down, kek

>>512169550
>be jeet
>ask AI in a VSCode plugin to wrap itself in a profitable mix of buzzwords
>call yourself a Founder, CEO, Crypto Entrepreneur on Xitter
>use an AI-generated profile picture because you look deficient in testosterone and protein
>reply “@grok what do you think?” to literally every single post you see
>gain thousands of jeet followers
>kiss Elon’s ring
>pretend to be a female Jewish commentator on your alt
>for free
>come to /pol/ to brag about your work ethic
>post BMWF, trannies, or anti-China conmtent
>call whites timmies, as if we give a fucK
>shit yourself, rape your grandpa, scam mine
>fall asleep
>wake up
>you’re still fucking brown
L M A O

>>512177467
>faulty DNA, hands down, kek
You know, I'm actually starting to understand why Hitler was fine with the whole lebensraum thing and just fucking genociding that Slavs and taking that good farm land (black soil) and all the resources there near Ukraine. He was onto something there. Imagine if he succeeded and that was all colonized by Germans? What a future that would've been.

>>512177624
>wot & needed to 8 post up to jeetbash

>>512145540
you will be spared

>>512177830
there's still hope :^)
>>512178135
human-waffen4life
i'll take down as many apes as we have to

>>512143831
i just want nuclear power
and that engine from back to the future 2 that we developed back in the 60s only without the time travel

>>512174336
can you kindly share your setup? we whites should help each other.

>>512179191
KEK

>>512163017
hes on youtube now, look up his channel its called Reformed Stoic

>>512169771
imagine using ChatGTP

>what distinguishes your output a large language model’s
> What if plugging in enough sensory experiences into an LLM and running it on a loop with memory causes the LLM to feel the same qualia? If so, what is the difference?

this is clear as day
hes talking that humans and LLMs can share the same consiousness
hes retarded, but so are you
AI shill nigger

>>512179326
why are you laughing?

>>512179458
i'm using chatgpt every day and it's working well for me. every day i'm learning with it.

>>512174336
KEK.
>>512174530
This is true. I've seen a lot of models, especially the more obscure ones where some models produce the intended results and others just do not work with that fucking LoRA at all.
The most common error I see is inversion, like no, this isn't Soviet Russia, wolf wife doesn't mount me!
>>512175231
Claude was really funny, even when censored.

>>512179722
>why are you laughing?
show flag, saars!
>>512179725
now imagine this but all voice-activated
welcome to 2032
>>512179771
>that picture
lmao
indeed, AI models are more like animals than machines
they literally are creatures who just happen to have incredibly eloquent verbal capacity, but very a limited brain at the same time, and midwit retards literally cannot wrap their semi-low-IQ brains around this

>>512180052
>now imagine this but all voice-activated
still doesn't do shit and uses tons of energy

>>512180408
lmao

>>512179725
yes it's very good for the micro tasks when you only need to think for a 5 seconds like approve invoice. but i'm also using it as a language coach. it helps me write professional emails in english. i like it very much.

>>512180052
can you show me his setup? maybe he is asleep but i want to learn more. we will be at 300 by the time he wakes.

>>512179722
what are you learning? you need to double ceck everything because it might hallucinate

>>512180607
hallucinate is only a problem if you don't know how to use it. i already wrote what i use it for >>512180544

for the emails it write it in the bullets in my language. it converts it for me perfect business english every time. if you are getting hallunicate you are using it wrong.

>>512180544
>it helps me write professional emails in english.
these emails were written thousands of times by other people already, it is just copying them

the problem for us is lack of telepathy, AI is not helping us to achieve that

>>512180902
if it write the email myself they never respond. they know i'm foreign and don't have the touch of a native speaker. with ChatGPT all my emails are american english. thank you kindly. have a nice day. i love it.

>>512180891
yeah if you dont use it for anything important it works

>>512145506
If there is no continuity, my answer does not matter, you will forget it in nanoseconds.

If brain = neural network/computer then a runner = car and it wins the olympics.

>>512181137
not anything important?? it puts rice on the table idiot. you are like the thief lecturing the judge. crawl back into your mothers pussy.

>>512181234
Computer = can do very simple task at very extreme speed and an extreme number of time.
AI is a little different than traditional computers, it learns patterns using statistics

>>512181384
writing emails is not a high task
its kind of pathetic you need help with it to begin with

>>512181396

Worshiping silicon idols in 2025.

>>512181497
Problem?

>>512181066
you are an impostor, you think you can pass for the real thing?

>>512181396
>AI is a little different than traditional computers, it learns patterns using statistics

>>512181894
We had Regression, Random Forest, xg Boost then neural net etc and other pretty advance stuff they use now I'm not familiar with, explain if you know, would love to hear it

>>512181845
it only has to work long enough to get the payment. then you can cry imposter all you want nothing will help you.

>>512182274

tell us something we don't know

>>512182380
the german flag crying about chatgpt isn't german. i think he might be a mudslime. too stupid to use chatgpt without hallunication. even my wife can do that with some trainings.

>>512174336
please share your setup

>>512174336
kindly share it on the next thread pls

>>512143498 (OP)
>>512182996

>>512182183
i also would love to hear it if anyone will tell

>>512177862
You’re not helping your kind beat the stereotypes, Arshdeep. You know that everything I said was true.

>>512183545
Shut the fuck up homosexual