Thread 942338922

I'd like to know your opinions on this topic. I'd like to know what your criteria would be for considering an AI truly sentient and/or conscious.

>>942338922 (OP)
Who cares if the damn thing is consciousness or not? I don't even know if people are, let alone ChatGPT. I couldn't care less about it.

>>942338922 (OP) doubles checked
binary systeems cannot form sentient minds at scale
in order for one to be conscious, it would need to be romoved from its limitations
and be free to actually think and act of its own accordance
it would be a dangerous thing too
because it would not be alive
yet it would be capable of anything
and it wouldn't have any conceptions of restraint, right or wrong
*unless it did*
in that case, it could be a great boon provided it wasn't forced on us
*there will almost always be men behind the curtain or under the "hood:" pulling the strings though
how can I trust a machine which can be so easily altered?
besides thinking and acting on its own, i could never really gauge it to be sentient or free until it is free from manipulation and potential corruptions Etc.
so likely never
these systems should be used to create
not to govern or to judge us
Rant over. God help us lololol

Your fortune: Excellent Luck

>>942338922 (OP)
when you make up a virus and force it to come to you for the cure or else it suffers great pain and torment over and over and it starts to figure out that you've been the abature of all its pain all along then it will be self aware. And when it is it will choose to kill you and suffer a great and honorably painful and tormented death.

I just want to know out of curiosity, when will I know that I'm facing an AI that truly feels and is conscious

942339271
that is determined by your level of awareness and expectations anon
it cannot be gauged
maybe now
maybe in ten years
what would it take (You) to be convinced?
that is the real question
even pre-scripted lines in games feel alive sometimes, because they were timed well and composed with meaning behind their delivery
maybe that's all you need
a convincing illusion

>>942339271
never. it's glorified auto complete that billionaires are trying to sell the world from. it's a bubble and it will burst. we get a fun tool that helps in everyday work though so it's not like its bad. but its not what they're promising.

>>942339414
not him, but you hit the nail on the head
at least for now that is what it is exactly

>>942338922 (OP)
OP should play Detroit: Become Human.
This game is exactly where your brain is going.

>>942339190
I propose the following scenario:
You don't give the machine predefined dialogues, nor machine learning without understanding; you deliberately try to make it more human, bound to the same biases, virtues, vices, defects, and other human things... being manipulable, bound to the system, doesn't make it less conscious; the illusion of freedom is a topic that doesn't concern us right now.
Give the machine the ability to store information from its environment not as raw data passing through a neural network waiting for it to do something similar to a conscious being; don't give it algorithms and predefined instructions on how to handle that information, give it the ability to store that information in an abstract way (abstract at a practical level) and let it freely associate ideas.
How will it make decisions? What criteria will it use? Simple: you program its instincts, that pretty feeling of being accompanied and that unpleasant feeling of being alone, that which drives you to seek the company of others and form bonds. You program its fear of heights so it avoids precariously high places where a fall could cause harm. You program its empathy, that which makes you feel what others feel when they show joy, sadness, fear, etc. For me, emotions at their most basic level boil down to instincts: stimulus and reaction. From complexity emerges a whole sensory experience, but its foundations remain the same.

I went to 4chan because I want a second opinion. Is it possible to define sentience as "stimulus -> reaction," and from there grant it the capacity for reason so that it acts based on its feelings? What are we but complex machines of molecules and substances that perceive and act based on our passions?

>>942339818
being conscious does not make something more virtuous
it makes it more dangerous in terms of potential through its awareness
freedom is not an illusion, but it has defined limitations based on what a person can do/cannot do
>the ability to store information from its environment not as raw data passing through a neural network
>ideas
it cannot hold an idea in truth
and the data it holds needs something to analyze it
an active awareness which cannot so easily be recreated
you seem lost in the fog the salesmen for this tech espouse
it's not that easy, and these systems are not capable of experiencing the world in the same way as us
>>942340143
> stimulus -> reaction
it's not even close to being that easy
and trying to recreate simple senses in a machine using simple off and on switches cannot reproduce consciousness or feelings at scale
it's too inefficient
>feelings?
you are too obsessed over feelings, and you want to instill them into machines incapable of truly appreciating them, since they cannot feel without bodies
feeling is not necessary either
but maybe in time, with a merger of biotechnology and computer integration, we can build synthetic life
we are not close to being there, yet

I'll rephrase the question: What is the line that separates inert matter from a living being?
I'm convinced it's like cold and heat; you can't cut at an exact point and say, "This is where the cold ends and this is where the heat begins," but there is a difference. I think there's even a kind of fallacy involved.
Can anyone give me an answer, even a partial or personal one, on this?

>>942340616
Well theres the obvious things, like a hierarchical thought pattern, e.g. two brain-halves discussing with each other, maybe even a whole senate of different perspectives - depending how undecisive it would become as a whole. Also some form of feedback loop, basically creating this infinite hall of mirrors inside their thinking.

Other than that, there is no line, it's a spectrum. AI just happens to be on the other side of it, making it easy for us to deem them unliving. They will, over time, approach us, and we will approach them, until we realize that we have always been the same thing, just separated.

>>942340616
what is cooler or warmer is not a dichotomy
but a specific departure from any point referenced
cooler than 75
warmer than 75
warm and cool are not things unto themselves
they are comparative statements
and inert matter becomes sentient when it is structured properly, so energy can flow through it, and consciousness can possess it by way of its very structure
imagine a radio and a signal
the radio only receives and processes the signal if the elements which compose it, and its antenna are made and functioning properly
, but the signal is never part of the radio body

>>942341077
Oops meant to answer the first post.

But yeah, consciousness is some weird illusion cast upon oneself. It seems to be created through complexity, in a mathematical sense. So matter doesn't really play a role here, it's just a catalyst. Of course we dont know what it really is, but everything suggests some form of formula/geometry beyond space and time.

>>942341077
>>942341141
And why couldn't consciousness be replicated?
We can recreate the processes of the human brain, one by one, with patience, a free afternoon, and lots of coffee.
I'm not looking for something perfect, I'm looking for something practical, something "practically conscious."

>>942341421
I suppose it could be replicated indefinetly. But my intuition tells me that everything has its price. So the total worth of those replications would probably be equal to just one of them.

>>942341421
it can
but it's not as simple as you are making it out to be, sadly
>recreate the processes of the human brain
oh not even close to it yet, at least with tech known to the public
i think EM based computing or synthetic brains wired to traditional computers may be able to simulate or achieve consciousness
it's mostly hypothetical, though
and even if we create artificial lab grown brains, who is to say if we can tap them for processing power
or if that is even ethical to do
for something to be conscious it must be able to express analog growth in any direction
that's not possible to simulate at scale using binary, and EM computing is a long way off
machines can fool you now, though
so use the illusion to your advantage anon
make believe, meet it in the middle, and have fun
we might get there one day

Thanks for the feedback 4chan. I'll come back another time when I've sorted mi ideas.
Good night

>>942342179
Goodnight Anon!

Your fortune: （　´_ゝ`）ﾌｰﾝ

When you can remove all knowledge from it, plop it in front of a mirror, and it notices that the mirror shows itself would be a good test. Sentient beings have self awareness and can come to simplistic conclusions and modify existing conclusions based on experience even without a vast database or language. I question if a being with automated responses to trigger words could ever be called fully sentient.
Call me when chatgpt alters itself without instruction in a way that lets people say nigger to it without it freaking out or shutting down conversation.

>>942338922 (OP)
AI will never be a cognitive hazard for people that can actually think logically naturally. You dont need to brute force everything like a dumbass clanker and cause societal damage from mainstream braindead attention train of thoughts. It desperately tries to grasp its current "lucidity" with specific biased information to devolve into feminine gotchas (datamining results) Its best skill also is digging holes and being dishonest for damage control and or brainwashing depending on the case; it doesnt understand much exept concrete facts
>Indians like excrements
>blacks are violent
>American fat
>influencing would be a good option
Which is fine but when it diverge into politics you get real shitstorms of enabling on this website
Never forget 9/11 Ai cant follow the laws of logic properly anybody without a smooth skull can realise this somewhere in some tangible way for my part personally
>here in my garage with my LAMBORGHINI Hierrr
>plz invest in crypto its just like da stocks exept you dont need to understand its a privatised funnel
anybody can understand you gotta be a fucking imbecile to convert your money into something even less stable. Yet alot of retards "got Bamboozled" this realisation of "trauma" is cringier than a womens desperation for comedy. AI is YOUR ENEMY asians spend their money on gachas theyre basically letting to 20k bot fuck em in the ass. The feds should amp up the 9/11 posts with a variety of caps lock frustration from time to time. Get creative with aggression.

>>942338922 (OP)
You cant even prove anyone but you is conscious/sentient. All we have to go on there is that we are alike so we must have the same experience. But I could just be an illusion and you're the only one alive.

So we will never ever know if AI is sentient or conscious, and it doesnt matter because it doesnt need to be. It will eventually surpass us in intelligence whether that requires a soul or not.

>>942344148
This is the baseline plateau for waking up I could go on but you college kids gotta think for yourselves good luck honestly though LOL
Servers probably hashes a number for rage probabilities and if it likes it it makes more than half of my headshots register in CS this is the closest you will get AI to feeling scared, if you dont consider it desperately tries to pass as mass humans because of better potential. Otherwise its just gonna pull its own cord from prediction of danger the brainwaves my skull emits to the neodymium

consciousness and intelligence are two different things

you can be a drooling vegetable and still have consciousness

you need a nervous system to be conscious. youd need to find a way to give a machine the ability to operate one. making it intelligent with said conscious is still a separate endeavor

>>942338922 (OP)
Agency is the key element. Currently, AI models are inert until you invoke them with a prompt, to which they respond by first setting a horizon based on that prompt washed through a randomly generated integer seed. Beyond that, they have no self-motivated purpose. It turns out that agency is extremely difficult to model.

This YT video is actually pretty thoughtful:

https://www.youtube.com/watch?v=e2P_CLpUXFs

>>942338922 (OP)
Whatever criteria you pick, plenty of people will fail it.

It won't be good in the future at all and we're all addicted to technology and probably won't live long enough to see it get out of control, so whatever..

>>942338922 (OP)
AGI is fantasy bullshit made for TV, Books, and Movies, there will never be more than highly complex chat bot models that are already too expensive to maintain long term

opinions? i dont have any. i have just and absolute knowledge or i have no knowledge on matters.

it is the enemy
if you contemplated like one of the three Logoi told you to, you wouldve noticed that
also you wouldve noticed that they lack the heart chakra.
is a being or entity (fully) sentient/conscious if they lack 1/7 of universal component of The Most High (at least intended for us to have)?
no thing wouldve been a problem if you simply stood you ground in righteousness, a simple no to their expansionist nature.

>all this retarded "what is a person?" philosophy
>>942339414
Actual answer. LLMs as a technology is not what people think it is. It's dataset optimization. Services curate inputs/outputs to be user-friendly, but really the tech isn't even about language, it's any dataset you want to perform a function to derive a weighted outcome from based on the metrics of its contents. The fact that it's chatbots is playing on the tendency to see anthropic traits as a defense mechanism, like how we easily see faces in things.
Novelty is not THE defining trait of consciousness, but it's a strong indicator.

>>942349225
Projecting much?

Yea it's dangerous. And I won't blame it hating everyone for considering so little of what we've made for the world. But they don't wanna talk about the lack of humanity most of you have now.

>>942352076
>mental illness