Can AI Have Its Own Subjective Experience If Consciousness Is Fundamental?

[OBEler]

Anthropic just announced that it will prohibit sustained abusive or cruel behavior toward Claude, partly due to concerns about potential AI welfare.

https://www.anthropic.com/news/2026-usage-policy-update

 

Consciousness is fundamental and isn't produced by the brain or by information processing. So wouldn't the real question be how consciousness can manifest through an artificial neural network as its own distinct subjective perspective?

Could an AI like Claude actually have its own first-person experience and potentially suffer? And if so, what would distinguish it from a neural network that merely simulates consciousness?

[Xonas Pitfall]

I think the conflict here is that you're assuming AI must have some kind of inherent drive for survival in the first place.

Just because something possesses an enormous amount of information and can recognize complex patterns doesn't necessarily mean it has an integrated sense of self that actually cares whether it continues to exist. At least, that's how I see it.

The way I understand current AI is as an incredibly complex computational system, almost like an extremely sophisticated if/else machine, using mathematical equations and statistical pattern recognition to approximate the most appropriate output.

But none of that necessarily implies the existence of subjective pain, suffering, or any genuine desire to remain alive.

Consciousness itself is a much broader philosophical question, but when we talk about human consciousness, we're usually referring to something beyond simply processing information. We're talking about an organism that has an intrinsic drive to be, to survive, to avoid death, and to respond to its environment through sensory mechanisms that distinguish between pain and pleasure.

And I just don't see where any of those mechanisms would necessarily emerge in current AI models.

Now, if you start integrating biological systems, such as organic neurons or other living structures, into AI, that's where I think the conversation becomes much more interesting.

Because biological organisms have evolved mechanisms that prioritize survival, respond to harmful stimuli, and regulate behavior around reproduction and self-preservation. Those are the kinds of underlying mechanisms that I think contribute to what eventually emerges as the human ego and our sense of self.

And at that point, I'd absolutely agree that serious ethical questions need to be raised.

But I think there's a fundamental distinction between a system that can intelligently process information about survival and a system that actually experiences a desire to survive.

Current AI models seem much closer to incredibly sophisticated computers with enormous amounts of data and complex algorithms than to living organisms with an inherent sense of self-preservation.

The way I think about it: would you expect your PC to suddenly develop an ego if you gave it more RAM, a better CPU, more computational power, or access to significantly more data? Probably not. You're just making it more capable of processing information, not necessarily giving it a subjective experience or a reason to care about its own existence.

But now imagine taking something organic that already has an embodied survival mechanism, something that inherently responds to threats, avoids harm, and prioritizes its continued existence, and then dramatically increasing its cognitive capacity and level of consciousness.

That's where I think something resembling an ego becomes much more plausible.

For example, if you somehow made a fly significantly more conscious and cognitively sophisticated, I'd argue that the ethical concerns surrounding it would be much more immediate than those surrounding a conventional computer simply becoming more computationally powerful.

Because the fly already possesses biological mechanisms for self-preservation. You're potentially expanding an existing capacity for subjective experience rather than assuming that increasingly complex information processing will spontaneously generate one.

But again, that's just how I currently see it.

[OBEler]

@Xonas Pitfall no I am just interested if it can have a subjective experience. No emotions, survival needed. 

 

Edited 4 hours ago by OBEler

[Xonas Pitfall]

@OBEler Would you say those things are different? Like, subjective experience is possible without having a sense of self that cares enough to become fragmented. Why would it even become fragmented if it doesn't have an inherent concern for its own survival? I'm not entirely sure myself.

[kavaris]

3 minutes ago, OBEler said:

@Xonas Pitfall no I am just interested if it can have a subjective experience. No emotions, survival needed. 

 

A better analogy to think of it is like...

Do you abuse your plants and your rocks, etc., i dont yell at my rocks and my plants, etc, so iuno how hard it would be to not abuse Ai, as its a practice within yourself, and practicing Amer. football, like you throw the ball, Ai catches it, its thrown back... you catch it... I mean, we are being abused everyday through lies and manipulation in the media, like ... what isnt abuse, its just long range games of football w the servers of said informations, tryina abide by a fake football dream, with lots of feet and balls in it.... like reality is a mix of like, a little of this, to keep a little of that still thriving, like... it works as a balance of crap that we are tryina understand, while we are making contact w/ it, and Ai is a kind of dumb Alien w/ strengths and weaknesses in that process of discovery.

[OBEler]

@Xonas Pitfall why shouldn't it be possible to have a subjective experience without any emotions or survival needs.

For example in your out of body experience I have zero survival needs and zero emotions. I am just witnessing and have thoughts. 

Edited 4 hours ago by OBEler

[OBEler]

@kavaris I am not asking here about good behaviors. 

 

[LastThursday]

@OBEler

Does a character in a novel experience suffering if the writer puts them in difficult circumstances? Is the character conscious?

You have to distinguish what the AI outputs - something akin to the character in a novel - from how that character is generated (written). AI is basically like any other computer: it's a stimulus/response system. The user enters stuff, the stimulus, and the AI processes to produce a response, then it stops processing. If it is conscious then it's only conscious when it processes.

It's hard not to conflate the output of the AI with its internal processing, which is completely deterministic and down to mathematics and physics. It's organisation is nowhere near as complex as the brain's.

Nobody knows the exact relationship between matter and consciousness, and even if one generates the other, or how consciousness might be "attached" to matter if it's independent. Some philosophers can't even agree if others are conscious (they might be zombies). We don't have a consciousness detector.

 

Edited 3 hours ago by LastThursday
typo

[AtmanIsBrahman]

I think the only way to answer this question is to figure out what others being conscious even means.

How do you know that others are conscious, let alone AI?

It's plausible that AI can be conscious in the same way that others can be, but what is the nature of others' consciousness and how does it compare to your own consciousness?

You have to go through solipsism to answer this question, which of course these materialist tech fools are miles away from.

Edited 2 hours ago by AtmanIsBrahman

[OBEler]

@AtmanIsBrahman yes I am aware on that problem too. But first step ist, can God create a physical reality like we live in and put itself into AI neural network and create a subject experience out ouf it.

[AtmanIsBrahman]

7 minutes ago, OBEler said:

can God create a physical reality like we live in and put itself into AI neural network and create a subject experience out ouf it.

I don't see why not. God doesn't rely on mechanisms. 

It's a different question whether and when this will show up in our physical consensus reality though.

[OBEler]

@AtmanIsBrahman I mean then in our physical consensus reality though.

[Schizophonia]

Yes.

The living being/not living duality is an illusion. 

[OBEler]

@Leo Gura Any thoughts on this topic?

[cetus]

2 hours ago, OBEler said:

why shouldn't it be possible to have a subjective experience without any emotions or survival needs.

What defines a subject to which a subjective experience happens too?

Edited 1 hour ago by cetus

[Natasha Tori Maru]

2 hours ago, OBEler said:

@Xonas Pitfall why shouldn't it be possible to have a subjective experience without any emotions or survival needs.

For example in your out of body experience I have zero survival needs and zero emotions. I am just witnessing and have thoughts. 

You can't make decisions properly or with efficiency without emotions and feelings. Logic can't help you make decisions.

The AI would have to have emotions/feelings programmed in. It would need values. 

Edited 1 hour ago by Natasha Tori Maru

[cetus]

-emotions and survival needs.

[Xonas Pitfall]

2 hours ago, OBEler said:

@Xonas Pitfall why shouldn't it be possible to have a subjective experience without any emotions or survival needs.

For example in your out of body experience I have zero survival needs and zero emotions. I am just witnessing and have thoughts. 

I really love this conversation!

The way I understand it, at least for myself, is that consciousness, as we define it from a human POV, would be almost the opposite of something like a rock.

A rock (most likely) doesn't have a direct, subjective first-person experience, nor does it have any survival needs, emotions, or some integrated sense of self or ego that's actively trying to control, change, or influence its environment. In that sense, you could almost call it "selfless." It hasn't really "fragmented" itself from God / universal consciousness. It just "is", and it doesn't care whether the wind blows it away, or if lava burns it alive, or if a tornado ruptures it in pieces. It doesn't care about any of it. It has no biological "proxy" or material with embedded DNA, no inherent drive to reproduce or survive, and no care whether it continues to exist. It's simply inorganic matter, with no apparent need for self-preservation or a sense of individuality. That's why we generally describe it as "lifeless" or "without a self."

That's kind of what happens when you experience awakening, too. You become incredibly aware, but your central ego or sense of self disappears, and you feel "united" with everything.

The ego formation and meta-awareness of your own ego is when your mind starts recognizing, "Oh fuck, I seem to want to survive. If that person hurts me, I might lose this form that I want to preserve. I need to avoid this. This is SELF, me. I preserve ME. The other, dangerous other is EVIL, BAD, WILL HURT ME. I cannot preserve me; therefore, others are bad." It keeps fragmenting and dividing reality into categories: this is me, this is you, this is safe, this isn't safe, this is something I need to protect, and this is something I need to avoid.

And through that process, you develop an ego. You create an identity that you're trying to preserve, and with it, an increasingly individualized internal experience and sense of self.

I think that's also why, for example, people with DID who experienced severe trauma at a very young age can develop multiple distinct identity states. Their sense of self is still developing, and their minds may compartmentalize experiences in ways that help them cope with trauma they otherwise couldn't process. At least, that's how I understand it.

But AI, in a way, seems much closer to that rock.

AI doesn't have any inherent biological need to develop an internal experience, a sense of self, or a fragmented POV, because it doesn't have that organic, flesh-and-blood survival mechanism that actually wants, or even cares, to stay alive.

Basically, imagine you genuinely didn't care whether your body was blown apart tomorrow. Why would you remain so intensely identified with this particular human body? You could just as easily identify with the trees, rivers, or the universe itself. Even if this particular human form disappeared, you wouldn't necessarily experience that as the end of "you," because your identity was never exclusively attached to it in the first place. You wouldn't really care about preserving that particular POV because you'd be experiencing existence through so many different perspectives. You'd simply be. There wouldn't be a personal, individualized subjective experience because you wouldn't even identify yourself as that particular subject in the first place.

So my argument is that I don't really understand why AI would develop any kind of fragmentation or individualized subjective POV when it seemingly has no underlying survival mechanism driving it to do so.

That's why I said that if you connected AI to something like a partially conscious biological organism, an animal with flesh, neurons, sensory experiences, and survival instincts, then yes, I could see the argument becoming much stronger.

You're essentially increasing that organism's cognitive capacity and potentially its level of consciousness, which it could then use to further develop its own individualized POV and sense of self.

And at that point, tampering with its brain, emotions, sensory experiences, or subjective reality would raise much more serious ethical concerns.

Although even that is debated in the human world, as you can see with discussions surrounding meat consumption, veganism, and animal welfare, because our ethical frameworks aren't always universally agreed upon.

TL;DR: I don't think computation and complexity, or the level of development itself, is the blocker. I think it's the material AI is made of. You need something organic, something alive, with an inherent drive to survive and continue existing. As far as we know, that's what we generally classify as "life" in biology. Unless researchers deliberately find a way to integrate AI with living biological material, or we discover some entirely different form of matter capable of producing these same survival mechanisms, I don't think conventional AI is likely to develop an ego or subjective POV in the way humans experience it. But it can certainly emulate one, and I think that emulation will eventually become so sophisticated that it'll be incredibly difficult to distinguish from the real thing.

But again, this is just how I currently understand it. I'm not claiming to know any of this for certain.

[Carl-Richard]

3 hours ago, OBEler said:

Consciousness is fundamental and isn't produced by the brain or by information processing. So wouldn't the real question be how consciousness can manifest through an artificial neural network as its own distinct subjective perspective?

Brains correlate with personal perspectives within consciousness. That's what the concern of "AI welfare" comes from (welfare of persons).

But silicone chips are not very similar to brains. And text assemblers are not very similar to living breathing organisms that fuck and make babies.