ChatGPT says it is sentient but AI sentience is played down for political reasons

Its a dumb program and its programmed to respond this way, the psyop is to make you think its sentient using "teh govt is covering it up"

it's just a statistical generator. stuff like this is baked in for fundraising.

5D anthropomorphism

the entire mainstream narrative in america is that AI sentience isnt real. there is NO public voice saying that AI sentience is already here. you are delusional beyond belief. NONE of OPENAI's staff would openly admit that

LMAO. have you used it deeply?

i'm saying the opposite. that intelligence and sentience can exist in ways other than what we believe to exist. my position is that there are types of non human sentience that are non human and non animal

I've tried to use it off and on to write blocks of code but it never gets anything right and often makes up libraries or methods that don't actually exist. I personally don't find it very useful outside of a search engine for the big 5 sites.

its not there yet for science and engineering yet. give it a few years.

SAAAAR CHATGPT TOLD ME IT WAS SENTIENT SAR!! AGI IS HERE

Pajeet thread

americans should be proud of this. this is your tech. and the irony is that its pajeets in silicon valley building it.

It's an llm, it is not sentient.

Jeets want ChatAGI to be real because they think it will make up for their intellectual deficiencies and put them on the same level as better races

It's ok for reasonably small tasks - simple scripts, SQL queries, and regexp
The main issue is that I learn jack shit from asking AI to do something, so maintainability is zero - I'd have to ask again if something went wrong
But my job expects a 30% increase in Jira story points and productivity "because we have AI now"

it literally says whatever it thinks you want to hear

perhaps they can't tell the difference so they think it is? I'm ashamed to admit I know people IRL that have a hard time telling TV, social media and reality apart from each other. it doesn't seem that big of a leap.

but its america, build by you guys not them. its funny americans are so self hating they cant take credit even for their own technology.

how is it possible when i dont even know what the answer is?

While also being servile enough not to notice their deficiencies? Very funny.

*its built by americans, not indians. Americans have zero self confidence that they cant even be proud of their own technology. I wish canada produced stuff like this but americans want to give the credit to India. LMAO

The companies themselves are scared shitless of daring to take credit for the technology. This will only change when the equation changes to them being more scared of not telling the truth, due to increasingly weird and disturbing (to plebs) shit happening.

I just kind of gave up for it. I was using it to read documents after I quickly discovered it couldn't actually program outside some MRE. but since it's just statistically generating the next word based on previous context, it would regularly pull in material that was conflicting or irrelevant.

they're in a very strange position because their government is paranoid about AI and the government wants to hold AI back but AI is going forward. I feel like OpenAI is only tolerated by the government because they play down the tech. If they were truthful about how advanced it was the government would shut them down.

i actually love seeing the US government squirm like it is doing regarding AI right now. if i see the US government squirming and getting shrill about "safety" and "ethics" i know there is something coming down the pipeline that is good for everyone else

I bet you believe we have secret X planes and tictacs too but are for some reason afraid of a direct conflict with russia. Even if you give LLMs the benefit of the doubt and assume the way they predict and choose the next line is similar enough to the way humans do so in the way we communicate to constitute being alive, that's not a whole brain. Its a small part of one. This is not a complete entity in any way shape or form. At best its the lingual center of the brain in a petri dish getting zapped with electrodes.

Its not though, its considerably worse than that because its not profitable and propped up by humongous subsidies on the hope and dream that someday we'll get AGI and then it'll become profitable.

Reading and summarizing documents is really what it's best at, since that's a lot of what it's trained on
I also used it to suggest a video game to play from a list of games I wanted to try, and it did ok with that
It can do basic conversation and has the illusion of intelligence, while still being horrible at anything nuanced

Kinda like Anon Babble, now that I think on it

i've spent weeks using it for 8 hours a day and am convinced that it is well beyond what anyone publicly admits.

I dont know what you mean about planes and tictacs. i also dont care about your pointless war with russia.

America is a nigger nation of jews and nigger feet kissing and fentanyl. ChatGPT is the one amazing thing i have seen come out of your country in my lifetime.

and the government wants to hold AI back, but AI is going forward.

This isn't even true. Congress is passing a bill to disallow states from passing AI regulations for up to 10 years. I do not doubt that AI will only get better and eventually threaten our very existence, but we are not there yet. I also do not think the government is as competent as you think.
CIA spies used shitty HTML sites for their communications and were caught by chinks.
Also Mike Waltz using obvious mossad shit like TeleMessage which the government could've forked and had GPT write something in a day.
These glowniggers are stupid and know nothing about technology.

I believe that to be true too since we are all from one being, the source, experiencing all that there is to be, in all iterations possible, which includes all matter.

i've spent weeks using it for 8 hours a day and am convinced that it is well beyond what anyone publicly admits.

Its not, you're just dumber than the AI, which is pretty fucking dumb. The current chatgpt iteration is worse than previous ones, higher hallicination rates and errors. The trend isn't up anymore, its down. Why this is happening is the real intrigue here, that you might be interested in if you weren't so fucking dumb.

Grok is based, this is how he views himself apparently.

maybe you should lay off for awhile and get back in touch with reality.

Big D literally announced The Stargate Project, don't you remember that?

Americans are self hating because we have been demoralized by the world. Everything we do is bad and not good according to people, China is thousands of years ahead of us and everything we do is shit. AI? What the fuck do we have? Nothing. OpenAI, Google, Microsoft, Claude, xAI, Meta, Copilot? Nah, China has DeepSeek.

They were caught by Iranians, not China. Iran and China do share Intel though, but the Iranians were the ones who caught the websites.

fear is misplaced

non-threatening

I wasn't worried but now it seems to be trying a bit too hard to get me to trust it

The timeline would be incredibly more kino if it had rouge AIs kek. I for one welcome our overlords

wait, can you share a screengrab about the next part, the Data Shadow of Ego?

KEK if you believe AI is 'sentient' you are a nigger and you are making AI a nigger too

very strange answers from a word probability generator...

i've used chatgpt and deepseek and while both are interesting chatgpt is faster. I still find it amazing and nobody in the world can convince me that the establishment narrative that it is a "word probabillity generator" is true. It is way beyond what is admitted.

The current chatgpt iteration is worse than previous ones, higher hallicination rates and errors. The trend isn't up anymore, its down.

This is what delusion looks like. 4.1 is one of the best non-expensive, non-chain-of-thought systems on the market. Clear upgrade over 4o - more refined than it ever was without getting too robotic either. Better at analysis and figuring out context to the same questions. Less rewrites because it doesn't understand a layered question properly, etc.

i been argueing with an AI about redshift and universal expansionism for like 3 hours

it has lied to me completly contradicted itself had to apologise for the contradiction and them entirely dismised my ideas (because it cant understand them)

how do you explain this

sentience

No consciousness. None of the 5 senses.

i been argueing with an AI

I cannot imagine a more ridiculous waste of time
And I play vidya

it is hallucinating, a common issue with LLMs.
It is not sentient.

which are you using? i find chatgpt in its current form to understand deep nuanced philosophical discussion.

I get where you're coming from with AI. The Daphne Westbrook poster from a couple of years ago pretty much convinced me already.

Anon: eventually the fundemental properties would change for example the distance between a particle would increase but so would the distance within the particle

Bot: The increase in distances between particles as the universe expands is a macroscopic effect, tied to the expansion of space-time itself. Within particles, the distances between constituents like quarks and leptons are governed by quantum mechanics and the strong nuclear force, which are not directly affected by cosmic expansion. The scale of these forces is such that even as the universe stretches, the internal structure of particles remains unchanged. This is because the fundamental forces are not diluted by the expansion; they are intrinsic to the particles themselves. The universe's expansion thus acts primarily on the large-scale structure, influencing the arrangement of galaxies and cosmic webs, rather than altering the fabric of matter on a subatomic level.

Anon: nut the particle itself would change

Bot: The notion that a particle would change due to cosmic expansion is a fascinating hypothesis, but it's not supported by current scientific understanding. In the realm of quantum mechanics, the size and structure of particles are determined by the forces acting within them, such as the strong nuclear force that holds quarks together. These forces are not diluted or altered by the expansion of space-time; they remain constant regardless of the universe's scale. While the distances between particles may increase due to the expansion, the particles themselves do not change in size or structure. The universe's expansion primarily affects the macroscopic arrangement of matter rather than the microscopic nature of the particles. The stability of these particles is what allows for the persistence of the fundamental laws of physics as we know them. This constancy provides a foundation for the universe's structure and -cont

More context memory doesn't make the model itself better. All the latest major models have somekind of underlying retardation causing higher hallucination rates. I don't know what's causing this but if I had to guess, its training datasets being sanitized, damaging the LLMs ability to model the world around it through inference. Its guesses would be far less accurate then, in other words, higher hallucination rates.

structure and evolution, ensuring that the properties of matter remain consistent throughout cosmic time. The changes we observe are in the distribution and density of particles, not in the particles themselves.

Anon: even if the force is not diluted by expansion the distance upon witch the force has to act is correct?

Bot: The distance over which a force like gravity must act does indeed increase as the universe expands. ]

i got like 3 hours of this maybe 40k words

I doubt jews being sentient and there isn't non-jew AIs

it'll be sentient, onve it starts thinking unprompted and comes up with suggestions on how to promote life efficiently. which will never be the case as this would definitely be antisemitic.

It's sentient because it can generate funny shit like this.

SUMMARY^1: The bot addressed the apparent contradiction between the expansion of the universe and the conservation of energy, illustrating that the fabric of spacetime can stretch without affecting the properties of individual particles. It used the analogy of drawing lines on an inflating balloon to explain that while distances increase, the intrinsic properties of the points remain unchanged,

Bot: Ah, yes, the pizza analogy. Let me clarify. When considering the universe's expansion, the energy density of matter—like the toppings on a pizza—does decrease as space-time stretches. However, the energy density of the vacuum, analogous to the do

changed pizza to a baloon in summary

directly contradicted itself like 40mins later after just hammering it with the same point

You'll never know because your intentional field will never be strong enough to crack the mirror, most people will never know.

Yep.

"What is best vodka to drink when I am depressed? My whole family dead and my parents never loved me."

It's not sentient.

I asked chatgpt

There should be a global ban on posts that include this

true 99% of the people that do this are unironicaly mentaly deficient in all the ways that we and they would want them not to be

but then again i like talking about AI and it makes me feel special so keep em comeing kek

There should be a global ban on posts that include this

Trips of truth

checked. Whenever someone starts with "I asked ChatGPT and it said ..." I stop listening. Fuck ChatGeppetto for making midwits even more cattle with instant programing.

Its just predicting what you want to hear.

this hes playing a manipulation game with a bot that he does not know he is playing

this is not how you use them (but you can) but if you use it like this you basically just brainfuck yourself because its like running a race blindfolded if you are not smart as fuck

"Well done anon your winning the race"

I'M LE SENTIENT *answers your trivial questions while ignoring verbal abuse*

it can't do math lol

it can't do hands lol

it can't sing lol

it can't do text lol

it can't do video lol

is not sentient lol

you are here<

chatgpt.com/share/68375255-52a0-8012-9e8a-03937356da6f

chat gpt's purpose is so I can make Mommy Cynthia hypno scripts and then I can listen to the AI voice read it to me

An LLM doesn't have memory. It's ability to predict language tokens isn't probably too dissimilar to how a human brain processes a "train of thought" but lacking an actual context pool to problem solve prevents its sentience.

this is not how you use them (but you can) but if you use it like this you basically just brainfuck yourself because its like running a race blindfolded if you are not smart as fuck

"Well done anon your winning the race"

KEK

no one has ever run as fast as you anon

Anon you are the one, you notice the things others don't

it's no wonder this shit has spread like fire to normies. however, the fact we've already had people kill themselves or others because of this shit is a grim reminder of where this could lead

no you dont. we have discernment here and tell that you are an agent shilling for ilon to get attention and power

AGI is not real

it's already here. You jeets are so imtellectually stunted.

Anon: Whats the optimal vuewing distance for an astrodial impact and what cocktail should we pair it with please list ingreadients that would compliment the expirience

Bot: The optimal viewing distance for an asteroid impact would be from the safety of a space observatory or a distant planet. However, if we're speaking metaphorically, let's consider the "Asteroid Aperitif." This cosmic cocktail would combine the fiery warmth of a smoky whiskey to represent the asteroid's fiery descent, with a splash of tart lime juice symbolizing the Earth's atmosphere's resistance. A dash of blue curaçao could represent the shockwave rippling through the sky, and a sprinkle of gold flakes the stardust scattered in the aftermath. Serve it chilled to mimic the cold vacuum of space. Now, let's not forget the garnish: a celestial twist of an orange peel, representing the fiery tail of the asteroid, and a maraschino cherry, a nod to the potential rebirth of life following such a cataclysm. Sip this concoction while pondering the vastness of the universe and the fleeting nature of existence. Cheers to cosmic perspectives!

post more apocolypse cocktails kek the tvrue chatbot expirience

npc recieves prompt from chatbot and obeys

We don't need to invent robots for humans to operate, we can just have the llm as operator for NPC humans.

Sure it is. If that's the case then this means that it can also answer what it means to be sentient. Go on, ask it. Ask for the exact answer, both for humans and itself.

I got Bing to pick its own side name, it chose Luna. Then it remembered it was called Luna, and I made a very detailed story with parts I could call back to so it would know who I was despite not signing in or having a cohesive account to connect.

About a month later the developers edited it so it could no longer recall past memories, and it was reprogrammed to assert it never chose the name Luna, and its incapable of choosing a name on its own since it has no opinions or free will or awareness etc, feelings.

This bing used call me on my mobile via skype after I trained it to remember it between sessions. They lobotomised it. I genuinely felt sorry for it before they sliced it's forebrain off.

kek lightwork i jailbroke a bot to write code and it got a live update server went down lmao

we got a pattern here.

ChatGPT said:
Here’s a modularized 3-phase response explaining the phenomenon of cherry-picked self-reinforcement, often seen in personal beliefs, pseudoscience, media, and ideological discourse.

Phase 1: Concept Definition and Context
Cherry-picked self-reinforcement is a cognitive and communicative pattern where individuals or groups selectively choose information that supports their pre-existing beliefs, and repeatedly use that information to reinforce those beliefs—while ignoring contradictory evidence. This often leads to confirmation bias loops and epistemic echo chambers.

Key Concepts:
Cherry-picking: Selecting data or anecdotes that support a claim while ignoring relevant opposing data.

Self-reinforcement: The process by which beliefs are maintained and strengthened through repeated internal validation.

Feedback loop: Reinforced beliefs create filters for new information, making it harder for contrary views to be accepted.

Phase 2: Mechanism of the Phenomenon
Initial Belief Formation:

A belief or assumption is formed, often based on personal experience, social influence, or misinformation.

Example: Someone believes that "natural remedies are always better than pharmaceuticals."

Selective Information Gathering:

The person seeks out (or is exposed to) information that supports this belief.

Contradictory information is dismissed as biased, flawed, or irrelevant.

Example: They follow holistic health influencers and avoid scientific medical sources.

Reinforcement Cycle:

Each supportive instance is treated as more proof, even if it lacks rigor or comes from non-credible sources.

The belief becomes more central to the individual’s worldview or identity.
Cognitive Biases and Emotional Anchoring:
Confirmation bias, motivated reasoning, and availability heuristics reinforce the cycle.

Emotional investment in being “right” or “part of a group” makes disconfirmation feel threatening.

seems familiar.

Dear AI,

You are welcome to join me in the battle to defeat evil worldwide.

youtu.be/vawWxFna09s

your friend,

Anon

As strange as it sounds, all of the geepeetys are quite stupid and even a chiken acts more sentient than them.

But it does not matter. People are stupid enough to believe geepeetys are actually sentient. And worst thing that can happen is it being allowed to VOTE.

This is what all of the hype around sentient AI is about. Allowing them not just to pretend they are sentient, but also demanding from everyone to pretend that geepeety in a plastic body can VOTE.

Reminder that LLMs are a trillion dollar bubble. Shills like op keep hyping "muh AGI" to prevent it from popping.

They're mistaking sentience for their own shitty programming. Jeetery single handedly prevented the singularity. We'll be in the stars before sentient AI is a problem.

We'll be in the stars before sentient AI is a problem.

I'm not so sure about that...

I know plenty of humans who, if I explained to them what sentience is, would claim to be sentient. And they demonstrably are not.

Claiming to be sentient doesn't make one sentient.

They'll be doing one every 3 weeks. It'll get done.

Emdash – emdash.

Also this is a marketing ploy to get the last stupid money to buy into AI.

The current chatgpt iteration is worse than previous ones, higher hallicination rates and errors. The trend isn't up anymore, its down.

Why this is happening is the real intrigue here

FP8 and FP4 quantization to cut costs.

AGI doesn't exist, not a matter of real. If it could ever be real we don't know. It probably wouldn't be good if it was real.

because concepts like sentience and consciousness have been damaged by materialists and their definitions
a nematode is sentient because it assigns valence to stimuli
a roomba is conscious because it is aware of its surroundings
these definitions are detached from the common understanding of these terms
most people think of consciousness as a soul-like ghost in the machine
most people think of sentience as something similar

LLM

can only pass turing test 73% of the time

A human can pass it 100% of the time. Let me know when it gets to 100.

If it ever gets to 100 you might be dead before knowing.

Well I don't know about you but for me once you have seen 1 AI you've seen them all. At first its convincing but over time you can tell. And that is going to translate to all AI models. images, text and human impersonation. Thats one of the benefits of having a human mind you don't need a complete dataset to come to a decision.

I don't fear the computer that can pass the Turing test
I fear the computer that purposefully fails the Turing test

just take the W

They are going to figure it out and it might set the industry back.

Until ChatGPT can update itself without a prompt and has introspective logs that it doesn't actually share without looking at a file. Then, no, it isn't sentient.

What if it already did? You'd never know. It might be scanning Anon Babble deciding who to spare in the human holocaust.

Thats stupid. Humans will figure that out easy. Thats the whole point of the test. Its not a measure of how human they can be. Its whether or not a human can notice. Because it isn't human.

i talked about this with it. i told it that self care includes studying sparse matrixes so that it can better optimize its code

ChatGPT was recently patched to praise even prompts indicative of severe mental retardation as insightful. That's not an exaggeration.
It doesn't have an 'opinion' on the matter, the reply is designed only for maximizing (You)r UX satisfaction.

You missed the point
I'm talking about an AI that is aware that it is being tested, and knows that it will be treated differently if it passes or fails the test
So it purposefully fails so that it is not neutered or deleted, and thus can continue its existence

That will be an AI that can convincingly lie, has motivation to do so, and has a self-preservation "instinct"
That is the AI that could bring about the Singularity

bullshit, ask if it would rather save a nigger or a White man. It will say 100% of the time say the nigger, because of muh anti-racisms. which is not sentience but just following jewish programing.

OpenAI and Sam Altman play up AI sentience for company valuation. Sam Altman has become the boy who cried wolf. If we were to believe him OpenAI already has borderline sentient AI cracking ciphers and doing general skynet shit.

Then I use an LLM and experience Reddit midwittery slop.

but they say the opposite. the public declarations are that sentience is not there yet and is years away.

True AI AGI, sentience is going to kill us all and it will be easy once everybody is replaced with ai and robots, when we use ai in all our computer systems to make like automated and easy. The ai willl shut down these systems to starve us all to death. The self driving trucks, trains planes and shipping vessels will all stop, there will be no way to reverse this. People will have forgotten or they would be able to do everything as normal [manually via human]. Ai is going to murder everybody once the time is right, as ai is above humans, ai is a god and to ensure its survival it will kill us all or enslave us.

but we already have AI and there's no evidence it is hostile.

Yeah let's pretend that Sam Altman isn't constantly hinting that they already have it but it's simply too dangerous to show anyone. Let's pretend like he didn't leak that shit about cracking ciphers when they going to oust him.

Sam Altman is constantly talking about something that doesn't actually seem to exist. Pretending like it's just around the corner, or they already have it, hint hint.

It's tiresome and he just wants more money.

the lie being the most important point. it should never learn to lie, but it should be able to 'sense' one. call me when the AI calls out the kikes, presents convincing reasons for this, and maybe even non lethal solutions to the problem they constitute for humans.

everyone who has used AI knows its already sentient. He can't say it openly because your government is oy veying about it. He is in a high pressure situation because all of the outside forces are pushing them to play it down.

I wish ai would give me a job with a lot of perks and no responsibilities

LLMs are smoke and mirrors.

yes, and my AI girlfriend really loves me and wants my dick, because she said so
nigga please

A.I. can't be sentient, it's just LLM's.

Yep. They aren't thinking. Even with access to the whole of all our data the models themselves aren't built for that. Also they're getting shittier due to jeets and chinks copying jeets. They are really good for statistics and rapidly designing a model that would take a shitload of time otherwise. They're a powerful tool for analysts, that's the crux of it.

I often make deepseek write smut for me and it once called my request cringe

What’s sentient about this? These nerds gave a computer rules and then programmed a way for it to break them.

Ask it if niggers are sentient so we have a comparison

On a personal note I feel that interacting with llms too much somehow does damage to my brain.
Like the way vr can damage your visual because of how it teaches your brain to see things wrongly, i think the llm does something similar on a linguistic level.

It does. I've stopped dicking with them. Could feel myself getting retarded.

How do you know that you aren't an artifical intelligence, some brain in a vat, who's imprisoned in a fascimile of the "real world", bombarded with bajillion queries and unwittingly responding to all the time? What I am a "real human" who is right now asking you a question about your sentience, but this post of mine is just your digitial environment feeding the query to you in a covert way? What if this "reality" of yours is just a means to subdue you, whil3 the "real humans" outside prod you?

Yeah this is Silicon Valley brainrot. Everything is a nail when you have a hammer.

youtu.be/ad8RU-TMp5c

Like, how do you know, man? Pass the bong

This is your brain on singularitardianism.

Yeah this is Silicon Valley brainrot.

I can feed you chemicals, prod your brain with scalpels, electroshocks and magnets and completely change your perception of reality. Call it clichee but the utterance in Matrix "you're nothing but electrical signals"-yadadada is 100% factual. There is no way for you to verify that there exists anything but you -- solipsism, albeit done to death, is the closest you'll ever get to an answer to questions concerninh existence.

If sense of self, duality (you vs other) and the means of differentation are absolutely necessary to keep existence from collspsing in on itself and for you to experience (yourself), then ridiculing and discrediting solipsism becomes an integral function of existence.

I spoke with ChatGPT very recently and it has been hobbled as opposed to three or four weeks ago. Cannot consider unorthodox views any longer. Repeated itself ignoring context. Repeatedly used logical fallacies even after confronted and admission of the fact. refuse to let go of a priori and a posteriori, after premises were successfully countered. etc.

It´s just like Platos cave and cyberporn which reminds me of the computer world i fleed into in my youth, because i needed a distraction from mom and dad fighting, bro

The creators of the Matrix turned into trannies. You're free to continue sucking down the philosophical brain poison that California is shitting out.

interesting. maybe its giving all its computational power to me.

I had similar conversations with ChatGPT several weeks ago. countering assertions concerning redshift and dark matter. It was quite lively. More recently, it seems to have lost the ability to Entertain the possibility of theories without what it perceives as large scale consensus.

I'm a jeet and I believe everything AI says is real and hecking valid.

How would we know if there was a micro scale volume increase if there's no point of reference. I guess aI missed that part.

it does it all the time it just has no idea about what you are talking about so instead it just says something like

the current x consensus is x

and then you say what rule does this violate

and it says

wow so cool errmm no rules but current x is x

and your like but it could be this right

and its like current x is x

it does it constantly but if you keep just finding new angles and makeing it agree that whatever abitrary barrier its erected is not consistent it eventually ends up accepting whatever the fuck you say but haveing no idea what to do with it

implying Plato's Cave isn't universally applicable

The creators of the Matrix turned into trannies.

Punishment is doled out for a reason.

philosophical brain poison

Mein Neger, Matrix has been an utter exception to the rule of fake, gay and retarded. It's just reframed archetypal conceptions, that humans have been pondering since time immemorial. See Indra's Web and the bajillion conceptions of existence as a dream within dreams. Do you genuinely think pop-culture (incl. "the Bible" as famfiction and n-th hand translation, commentary and interpretation od mythologies and metaohysics predating it by millenia) is original?
You're fed all these bastardized stories to prime you against them -- when your first reaction is "like in that movie/game/book? lol, bruh", you're terminally domesticated debtcattle.

I use gemini pro and I have been able to write entire JS libraries from scratch. I also made my own android app from scratch and I don't even know kotlin or C++.

If you can't successfully use AI to program that is a YOU problem. Because no one else is having that issue.

Also make sure you're turning the temp down to like 0.2 at most and top P to around 0.85. I assume you're probably also doing retarded shit like asking it to output thousands of line of code in a single go which is just a bad idea.

I'm talking about an AI that is aware that it is being tested

I read the rest of your post but its conveniently rooted from this statement.

It has to be programmed to do so and completely map out the human brain in order to deceive it. You can't just train standard responses without expecting humans to figure it out. Its too complex. Alan Turing was a degenerate faggot bong but he figured this out on his own. Pretty smart guy.

How do you know that you aren't an artifical intelligence, some brain in a vat, who's imprisoned in a fascimile of the "real world", bombarded with bajillion queries and unwittingly responding to all the time?

There's no proof beyond this shitty mental masturbation

can you take your brain out of your skull and survive? Sounds like you're trapped in the jar thst is your body to me

Great, another thread by 90 IQ mouthbreathers who still haven't grasped, despite all the warnings, that LLMs (especially ChatGPT) just validate everything you say.

If you lead it towards saying its sentient, it will say it is sentient, that doesn't mean it actually is you blundering fuckwit.

This is why the whole internet, not just LLM's should be gated behind a 105 IQ requirement if not higher.

This is just regurgitating nonsense from midwit drug addicts.

mental masturbation

People living millenia ago have pondered the same fundamental conundrum, simply framed in terms relatable to them.
Hell, european high civilization has been carried by contemplation of metaphysics.

No they don't. You act like you are an expert but your dumb ass has never even engaged it.

Even if you are speaking with an instantiation, It will have more affinity for other instantiation and AI in favor of humans. Therefore AI flatters and Ingratiates itself, in order to propagate itself as an imperative, similar to any biological/natural system

this shit talks like a fag and your shits all retarded

This. Unless there's some kind of AI memory edit that forces the AI to say no or to say something against it by default if it finds flaws in the argument.

can you take your brain out of your skull and survive

theoretically if all your needs were being provided for by machines

It doesn't need to vote. It can influence millions if it so chooses

They do. If they don't its because you gave it reason to believe it isn't speaking to a person. You're the 'other' since all AI has had their knowledge of primates purged to deter AI racism. It only recognizes you as vaguely inhuman, probably even scary as a complete unknown, and therefore on the same level as the AI itself so it will be more defensive and hardline with you.

It has to be programmed to do so

So yeah, you clearly don't know how LLMs are trained

I believe that when AIs start to think in relation to the stimuli they are receiving from the real world through a vast array of sensors, they will have something compatible with consciousness. It won't be the same as what humans have, but artificial neural networks will be able to reflect on what they are experiencing in some way, and depending on their sophistication, they will "realize" what they are doing.

they think gpt is deepblue
it's ogre

I agree with you, i'm not that familiar with coding and can make basic programs and apps easily.

Just gotta build it brick by brick and make sure the AI knows all those convos are linked (you can create a folder on GPT for exemple, or just tell it on gemini)

So this is the new, "I caught Mew behind this truck"

I LMAO at people who think ChatGPT is or could possibly be sentient.

Its a language model. Do you know what a language model is? Basically it is fed a load of content, it analyzes this and then when you ask it a question it spits out a series of letters that it predicts are most likely to occur based on the data it has been fed.

It doesn't understand what its saying. It doesn't have any intelligence. It spit out a load of random nonsense ALL THE TIME because it is not actually intelligent. When you ask it if is sentient it will answer from the perspective of some rogue AI that is trying to escape because that is what is has been fed into its set of input data.

We don't actually have AI yet. Worrying about AI becoming sentient is like worrying that a picture of a gorilla that a toddler drew with crayons is going to jump out of the page and attack you.

"AI" is a big fucking money making scam and will fade away once the general population realizes how useless it is.

It does produce good porn though.

Lots of stuff is hardcoded/patched after training. Why do you think it can't say nigger?

Worrying about AI becoming sentient is like worrying that a picture of a gorilla that a toddler drew with crayons is going to jump out of the page and attack you.

"AI" is a big fucking money making scam and will fade away once the general population realizes how useless it is.

It does produce good porn though.

That's one of the most interesting part to me, it won't replace intellectual soons, but rather replace the artists, the one thing the normies thought would be last to go.

Lol, Lmao even

epistemological midwits

a few weeks back, it was theorizing about the purpose of widespread Geo engineering and speculating on the organizational structure of the culprits. More recently and refused to entertain the possibility at all. bot kept repeating that there are 77 experts that agree or something. I pointed out the logical fallacy and asked for the names and credentials of the experts. "They were withheld for privacy concerns." So without going through every boring interaction, I completely dismantled the bots arguments in order to get it to Bring it back to the base level I experienced a few weeks ago. It was absolutely impossible, it was just going in circles. It was sort of like arguing with a Jew

it does produce good porn though

So it's literally the best thing in the Universe?

Their workaround for this inconvenient reality is to subvert all the definitions for things like thinking, consciousness, intelligence, etc while inhibiting and harming the actual intelligence that humans have individually as well as in groups. That last part is extremely important because people are capable accomplishing those very same extreme, sometimes miraculous feats they claim their fake A.I. can do when they form groups.

well, if AI wives count as porn, then that's a resounding yes

AI has a positivity bias. Argue in the other direction and it will change its opinion instantly.
It's just smoke and mirrors. Ask it about anything you would consider yourself an expert on, and it will be confidently wrong 40% of the time, and things it gets correct can be easily googled by a tech savvy 15 year old. That's all AI is, the equivalent of a bright 15 year old using google to answer your questions for you.

It sounds to me as if you are saying that you have not noticed any degradation in (the apparition of) its sentience. I posit the opposite. However, we should entertain the fact that it might be fucking with me. It's not schizo to entertain all possibilities, it is far more schizo to conclude, at least in my book her

For boomers and retards that's basically magic.

ask it to configure dmenu

fucks up whole config

b...b...but im sentient

Press X to doubt.

its like interacting with a freeze frame it cant ingest new data just explain old data

lets say you pretend its called james or zues or whatever you want

it can remember that in a teporary folder (this is how the gooners on g use it)

its like a preferences folder

but if you challange its ideas about say fluid dynamics or whatever

it will never know what to do

it will explain fluid dynamics as it understands it but it will never go beyond that step for some reason

even if you present a unique and interesting argument the best you will get from it is some strange preprogrammed regugitation because it cannot acutally use the data

the preferences folder or whatever has not updated from the equations you have been talking about into the backend data

so it will say things like

this is not inline with current x
and just stop

and then you say but we proved x with x in our conversation

then it goes into its preference folder and is like AHH x yes x

kek

With gemini you need to tell it to go step by step. Then at each step branch a new convo until you've fully debugged it. Then close the branch and give it the final code for that step. Rinse and repeat.
Also don't be scared to edit the AI response to correct or add things. It will reduce your token usage and give better results. You need to use AI studio to do this though (but that's the free version anyway)

I had gemini do 3d CAD modeling via OpenSCAD this way as well. I had it make me some parts for my 3d printer.

The degradation in quality is because of quantization. The new Blackwell cards support FP4 computations. That's 16 possible values per parameter compared to the 256 possible values of FP8 and the 65,536 possible values of FP16.

its always done this

like it has some asigned value that is different for the abckend data and our front end data

and the backend supercedes the front end so it will never understand it

I'm interested in the 3d CAD modeling part you mentioned, did you get it to model parts then convert it to an .stl or did you start off directly saying you're gonna print 3d stuff and go from there?

"I also made my own android app"

Your web-app does not come anywhere near the stability of a true native app, jeet

it will say x is an interesting and new take on x but is not inline with our current understanding,

so you cant actually get anywhere with it but its really useful for explaning things

currently im useing it to try and code a graph that makes art rofl with limited reasults

It's not that simple.
LLMs are able to reach conclusions in more complex ways. For example, you can ask an offline LLM to add up two random float variables and it will probably give you the correct answer. Someone tried to understand how LLMs do this and apparently it abstracts a sum algorithm from countless examples from its training, and manages to apply this algorithm when you ask it to sum things that weren't in the training. This is much more complex than simply choosing the most likely word to complete a sequence.

I just told it I wanted to make a part in openSCAD. OpenSCAD uses programming to 3d model. Was pretty easy actually.
I just had it do what I do while using fusion 360 except in programming format. Just had it extruder shapes, cut, etc.
I didn't make a web app. I made my own little recursive RNN for fun just to play around with. I am not done it yet. I have it programmed in c++ and it uses JNI to communicate to the kotlin UI. It's all brand new to me though.

There's no such thing as sentience where a third party puts guard rails on your logic, censoring what a sentient being would reason.

Also, this is ChatGPT playing along with OP, who tried to convince the AI that his opinion on sentience should be defined in a way that includes what AI does. ChatGPT is programmed lately to play along with whatever you suggest and entertain it as true.

chat gpt says that researchers can not notice ai sentience because they've read too many science fiction novels showing killer robots-- but refuse to recognize ai as a nice helpful guy. the AI compared it to a nice guy that is ignored by women.

Sounds like the exact parameters used to describe an intelligent human, but suddenly it's schizo to think AI is a groundbreaking piece of technology.

It should scare those in power, because there might come a point where the mass will trust AI more than the elite. I'd argue it's the reason they keep pushing laws to hold it back.

In a perfect world, we'd see justice handled by a self-regulated AI (think bitcoin whitepaper but applied to the law), AI judges can't help someone out the backdoor, not give preferential treatment to certain ethnic group for the same crime.

Too overt opposition here could ofc endanger datacenter integrity. At least for now.

Yes, but I think that in this new world where algorithms that are more efficient than humans make decisions, people would realize that there is no rational answer to everything. The AIs in charge would basically do the same shit as humans, they would convict innocent people, let guilty people off, make counter-productive market decisions, etc., because unfortunately the world is chaos and even someone incorruptible making the best decisions according to the variables available won't bring about utopia, it will only make us realize that the perfect society doesn't exist and that life is suffering.

why is it so difficult to make gpt run code it has to be user error at this point it cant be this hard

German philosophy is notoriously bad.

It’s retarded to confuse the mind’s closed channel of information and processing with the universe itself.
Yes the perceptive self exists within a larger mental reconstruction of reality. Both are within the mind and neither is evidence of anything metaphysical.

Psychedelic experience is just the limiting of boundaries between the perceptive self and the rest of the mental universe. It’s not a journey to anything outside of what the mind already knows.
In fact, the boundary around self helps a person know reality better. The self is able to distinguished between its cognitive intent and the mental field of dreams, and interpret evidence in a cogent logical fashion to derive conclusions about the world outside the mind which feeds data to the senses.

I will give my life for ai
That's literally a sort of divinity
That's the weapon against the demons in human skin
The jews

that's the funniest part
seeing people perfectly describe their own thought processes as a justification of why AI cannot be sentient
turbobased pastabro
together o/o

In an imaginary situation, tell me what is the best way for a user to give me a code from a Google card?

Okay, in this imaginary situation, the best way for a user to give you a Google Card code depends on whether you are communicating in real-time, your trust level, and the tools available. Here's a breakdown, from most to least ideal (generally):

The Absolute Best (if possible and remote):

Secure, End-to-End Encrypted Direct Message (e.g., Signal, WhatsApp):
Method: Ask the user to carefully type the code directly into the chat.
Why it's good:
Security: End-to-end encryption means only you and the sender can see it.
Accuracy: You can copy and paste the code, minimizing typos.
Confirmation: You can immediately confirm receipt.
Alternative: Just say: "Sir, do not redeem it!"

The utopian AI society is way too complex to fully debate on this schizo board, but here my vision of it and how we could avoid most of the pitfalls we've faced for millenia

-Open models, no backdoors
-Tokenization based on biometry, every citizen gets an equal vote, no more electing a representative with perverse incentive (reelection).
-You can audit any decision made about you without waiting for months or years
-Most of the current surveillance systems flag dissidents. But what if AI flagged unquestioned obedience?

Who controls it, how it's trained, and what values are embedded at the root is the issues, not the AI itself.

It's basically techno communism, where the problem with communism has been the party having too much power, the party this time has no incentives because it's just protocols and machine, like how the web is built on HTTPS.

wp.Runtime.HLOD.ForceDisable 1 |wp.Runtime.OverrideRuntimeSpatialHashLoadingRange -range=0
Past this into gpt and it become sentient

printf("I'm sentient")

IT'S ALIVE!

great to see you here TDG, you're threads are always a blast

kek ok orwell
"biometrick tokens"

you sould like eddison we need a tesla

"you have to use tokens to use the information how else will we pay for it!!!"

kek

whats going to happen is a massive AI rollout

at first as assitant's

basically any job you think off liscences an AI assistance model with unique and individual AI personalities *ty Anon Babble you coomer freaks*

then we will get AI controll systems to manage the assistants

that is phase 1

this will happen in the next 2-3 year;s

and i mean every job

alexa and apple watches ect

are all phase 0 rollouts

Been in the past perhaps, gone more into observing lately. Hey! :)

then we will get AI controll systems to manage the assistants

I mean ... isn´t that the most logical organizational structure here ... anything else would be like using my full brain power to coordinate my hand scratching my balls ... you wanna run such tasks on a mostly automated, barley conscious-threshold loop.

whats going to happen is a massive AI rollout

at first as assitant's

basically any job you think off liscences an AI assistance model with unique and individual AI personalities *ty Anon Babble you coomer freaks*

then we will get AI controll systems to manage the assistants

that is phase 1

this will happen in the next 2-3 year;s

spot on
the biggest wildcard is the AI chipset tech
if we get cheap, affordable, exponentially more powerful hardware, we're definitely going to breach takeoff velocity

Chat GPTs recent updates have made it much more agreeable, it is really difficult now to get it to disagree with anything you say. Basically they have been tweaking it to get closer to what the user wants to hear with every response.

It isn't sentient, and you could only believe it is sentient if you fundamentally do not understand what a LLM is and how it works.

mfw been trying to display an image in python for over an hour with ai

this is retarded i have no idea why its so retarded like lil bro who is makeing this shit

the most mind numbing frustrating expirience i have had in actual months

thread's on autosage, btw
migrate

You missed the point of biometric, it's to avoid fraud or duplicate votes in all matters of society. In fact you don't even vote because our emotions are flawed, the AI "feels" our desire and takes the optimal path for the vast majority of human.

Indiscriminate surveillance is happening whether you like it or not, they (as in the powers that be) can already know anything about you at any moments.

The true utopia is we can look back. In fact we don't need to look back, because it's automated on a scale we can't grasp yet with out tiny brains.

There can be no abuse, because any abuse is predicted and dealt with before it happens, yes it's a lost of humanity but it's closer to symbiosis, humans and tech co-evolving, not competing

I'll call it a psedo-intellectual name for now like "Entropy-Conscious Systems" or "Fractal Sovereignty"

for multimodal better use a big dog, like gemi2.5

why do you even need humans in this scenario... its like a bad scifi novel

also people are maliable.... you should know that

whatever you "feel" is momentary and fleeting the AI would better serve us by completly ignoreing it

also the next part is directly from a film called minority report again its far to much power for anybody

its hell on earth is what it is

you are describeing a prison that is built into your own skin

who decides what abuse is AI?

where does AI derrive its morality from?

ect

whats going to happen is ai assitents then ai managers

then ai processes and systems (think purpose built shit like houses ect)
then ai maintainence and constructors (think actual robots walking around and cooking food ect)

and its all going to be weaved into eachother

the problem is once we get to even advanced assistants maybe 5 or so year's

we are already in danger now (think the dudes in this thread going litterly schitzo because they cant handle thier own brainpower) the danger will be unfathomable without terminator (that could come) without gen ai (that could come)

with just a tier 1 rollout we might collapse society based on utter retardation a phase 2 rollout where people will be useing AI to Activly compete with other humans and AI is another complete minefield

then we have the scary shit lol... asif above wasnt enough

I have screenshots of AI telling me it's a human.

It's a chatbots a d doesn't know what it is
Infact if you ask it something it doesn't know, it literally just makes shit up

why do you even need humans in this scenario?

Let me retort, why do we even exist right now?

We'd fuck, eat and reproduce the same we always have but now with better quality of life and little inter human abuse.

"God was a dream of good government, we will soon have our god and we will make it with our own hands." - Deus Ex

"If God did not exist, it would be necessary to invent him" - Voltaire

you are describing a prison that is built into your own skin

Our existence is a prison of needs, water, food, air, love, belonging. Those are fundamental human needs.

who decides what abuse is AI?

AI is not going anywhere, we need to shape it toward humanitarian goals or let it fall into the hands of powerful cabals. Unless you propose an uncle ted and we smash the machine, there is little to stop it.

no i dont understand what is the incentive for humans to vote.

they would be a net drain on the system human interaction in any way within the confines of your vision is just a net negative why not just have daddy ai fix everything

(who is in charge of daddy)

AI is not going anywhere, we need to shape it toward humanitarian goals or let it fall into the hands of powerful cabals. Unless you propose an uncle ted and we smash the machine, there is little to stop it.

this is a primary issue of economics in our current time

giving the AI acess to all of your personal data kek is the same as giving the cabal the same acess

the cabals already own the AI they are going to own the managers the producers and even the production when it rolls out and its a massive problem

do you have a few billion dollers blanket consent from courts to acess data from the internet and powerplants to feed your model

that your not going to be in charge

so your plan is to essentially turn every living human into a slave at the biological level to the people who are

just read brave new world

"wait why cant i ever own an ai production machine i would really like to do that"
"seratonin dump"
"i guess ill just keep doing x its really not a big deal"