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

[Heaven]

With all the respect to human inventions, I don’t see how AI will ever be sentient. 
Biology and evolution are super complex processes. Coding AI is like a joke compared to nature.

[Breakingthewall]

50 minutes ago, OBEler said:

@Breakingthewall yeah but if I have a circle of stones with fire in the center of it I also have an organization, which is consciousness. My question is, when becomes the stone circle sentient. the fire integrates informatin from the external world and relates it to a persistent subject that survives through time. WOuld you then say the stone circle with fire in the center is sentient?

I cant see that.

This structure is not consciousness. It's like saying, an atom is a extremely complex structure with gluons and quarks and all the stuff , why if I put some stones there is not an atom? Because it's not an atom😅

Could be done an artificial atom, a stable pattern of the fields of reality that last 10 raised to 30 years totally stable that combine with other atoms creating molecules? Maybe, when we become gods.

Could be done a pattern that is consciousness? Maybe yes , it could do itself in a not very far future 

Edited 13 hours ago by Breakingthewall

[Breakingthewall]

1 hour ago, Heaven said:

With all the respect to human inventions, I don’t see how AI will ever be sentient. 
Biology and evolution are super complex processes. Coding AI is like a joke compared to nature.

The point is when the code starts to create itself and increase in complexity exponentially. If this can happen, for now it's just a possibility but seems that there is nothing that makes it impossible 

Edited 13 hours ago by Breakingthewall

[Leo Gura]

Just because a chat-bot can't do it doesn't mean computers can't be conscious in the future.

Chat-bots are very primitive compared to what is possible with proper research and understanding.

Edited 13 hours ago by Leo Gura

[r0ckyreed]

34 minutes ago, Leo Gura said:

Just because a chat-bot can't do it doesn't mean computers can't be conscious in the future.

Chat-bots are very primitive compared to what is possible with proper research and understanding.

Consciousness is not localized in a brain or computer. No object has consciousness. You won’t find any consciousness inside a computer.

Computers are consciousness but it doesn’t mean they are conscious or have a separate consciousness.

You can build a sophisticated computer, but humans cannot create consciousness. 

“A object (e.g., advanced AI computer) does NOT have consciousness, it IS Consciousness.” —r0ckyreed

Edited 12 hours ago by r0ckyreed

[Someone here]

When discussing this topic , I don’t understand how could the body-mind problem or understanding the nature of consciousness existentially be put on hold . It’s essential to this question. It’s the whole key to this question.

[Natasha Tori Maru]

1 hour ago, Someone here said:

When discussing this topic , I don’t understand how could the body-mind problem or understanding the nature of consciousness existentially be put on hold . It’s essential to this question. It’s the whole key to this question.

Gotta agree with this.

Phenomenology of AI - no reliable basis for claiming it does.

Awareness of itself - no reliable basis for claiming it does.

Does AI experience itself as a subject? 

On the flip side, there isn't evidence for any absence of these capacities.

'I don't know" seems to be the only reasonable stance.

[Someone here]

2 hours ago, Natasha Tori Maru said:

Gotta agree with this.

Phenomenology of AI - no reliable basis for claiming it does.

Awareness of itself - no reliable basis for claiming it does.

Does AI experience itself as a subject? 

On the flip side, there isn't evidence for any absence of these capacities.

'I don't know" seems to be the only reasonable stance.

 Yes .“Awareness of itself “ is what’s being investigated. And I say we should define , or know what exactly is awareness of our self and environment.  
awareness can be defined as perception.Where there is a subject and object and a process of perception going on . Or it can be defined as being itself without a duality between perceived and perceiver. Nondual teaching sides with the second definition. Materialism sides with the first one . 

Edited 7 hours ago by Someone here

[Jirh]

4 hours ago, Natasha Tori Maru said:

I don't know

Thanks to AI, this will probably be my official slogan for the rest of my life.

[Carl-Richard]

Let me tell you guys about something called biology:

Life on Earth is a product of a 4.6 billion year long process called evolution by natural selection. 4.6 billion years, that's 4.6b/(2026-1959=67) = 68.66 million times the time since the first integrated circuit was invented in 1959. The products of evolution are also selected by the entire ecosystem holistically, a huuuge, interconnected and complex information processing system, and in the entire history of life on the planet, that equals essentially the entire Earth's surface, the atmosphere, the Sun, lunar cycles, interstellar radiation, cosmic background radiation, etc. Essentially the entire universe is doing the selection holistically as one coordinated unimaginably information-dense unit.

Meanwhile current human computer technology (and current AI technology) is literally monkeys banging rocks together (silicone chips are made out of silicone, a type of rock). The selection process is only probably a couple thousand minds on the entire planet (and again for only the last 67 years), and their minds can only focus on one thing at a time, horrible CPU (7±2 chunks of information for any given moment), thinking is usually awfully reductionistic, linear, confined, contrived, limited to language. "But AI is now used to improve itself iteratively". Yes, AI, that is LLMs, are now doing the job of banging (in terms of software) rocks together, using monkey knowledge and cities worth of power, real estate, materials for hardware, water, etc. Yet we only get a chatbot that can write overly verbose text that might solve a theoretical problem or two from time to time when granting it again immense power, materials, water (through ungodly amounts of special-access token usage) and even then it only brute forces the problem like some rat in a maze.

LLMs are by estimation by any reasonable stretch of the imagination a dead-end when it comes to sentience, and perhaps even intelligence given certain projections about the limitations of upscaling of hardware. If we want something that even has a chance of developing / possessing sentience, we have to in my opinion abandon silicone chips as a hardware/substrate and instead focus on hacking biology like Michael Levin is doing and the rest of synthetic biology. The level of complexity of biology is in an entirely different domain than monkeys banging rocks together. Choosing an arbitrary level of computation like "neural synapses" ("neural networks are just like the brain") or "electricity" for establishing similarity is itself a testament to our severely limited monkey brains for dealing with such complexity.

Edited 1 hour ago by Carl-Richard

[Leo Gura]

5 minutes ago, Carl-Richard said:

If we want something that even has a chance on developing sentience, we have to in my opinion abandon silicone chips as a hardware/substrate

That's a bit hasty.

We need to explore the possibility of digital intelligence and life. It just needs to go way beyond LLMs and vacuuming up all the world's IP.

[Carl-Richard]

4 hours ago, Leo Gura said:

That's a bit hasty.

We need to explore the possibility of digital intelligence and life. It just needs to go way beyond LLMs and vacuuming up all the world's IP.

It's like the choice is between trying to reverse-engineer qualities found in biology like autopoesis, self-concern, selective attention, relevance realization (which is reasonably the only way you get a self-driving organism that can truly interact with the world like a living organism), or you use biology and go from there.

The former seems again like just a slow and inevitably unworkable process again because we're monkeys with tiny processors compared to the cosmos (and current AI are our sheep) and we're using starting materials that are way too dissimilar from what ourselves are made of and what we're aiming at (we're aiming at ourselves just better, just more powerful, smarter), while for the latter we align ourselves more closely with the "right" starting materials and we can let mother nature be the artist while we're more like managers.

Michael Levin changed the electrical charge of some cells in a flatworm, cut it in half and instead of the piece with the head regenerating a new tail like normal, it regenerated a new head, giving it two heads. If we can map out the space for manipulating cells using electricity or other stimuli (combined with genome-editing, bionics, etc.), we might end up creating something like a mega-Einstein, not a robot, not a human, maybe a bit of both. (Don't quote me on the ethics of that). "But he/it will be the size of a human so their brain will be small". Will he, will it?

Edited 1 hour ago by Carl-Richard

[emil1234]

my perspective; theoretically yes, but it is no more likely to be conscious than a rock. theres no correlation between intelligence and consciousness.

but tbf i never understood why humans bothered with this question. we cant even prove consciousness in humans; maybe lets do that first?

Edited 4 hours ago by emil1234

[Dodo]

@OBEler

The more interesting question to me is, how can natural intelligence individuals have their own subjective experience if consciousness is fundamental (and One)

[Mellowmarsh]

8 hours ago, r0ckyreed said:

Consciousness is not localized in a brain or computer. No object has consciousness. You won’t find any consciousness inside a computer.

Computers are consciousness but it doesn’t mean they are conscious or have a separate consciousness.

You can build a sophisticated computer, but humans cannot create consciousness. 

“A object (e.g., advanced AI computer) does NOT have consciousness, it IS Consciousness.” —r0ckyreed

Consciousness is not even conscious of its creator. Consciousness cannot fully grasp or objectively prove its own source or creator because any attempt to examine consciousness must itself be performed through consciousness.

 

[OBEler]

To all here who think only something biologial can be alive: Artificial neural networks made from living brain cells already exist. Can that be sentient?

[Leo Gura]

2 hours ago, Mellowmarsh said:

Consciousness is not even conscious of its creator. Consciousness cannot fully grasp or objectively prove its own source or creator because any attempt to examine consciousness must itself be performed through consciousness.

Outstanding!

[Breakingthewall]

5 hours ago, Carl-Richard said:

Life on Earth is a product of a 4.6 billion year long process called evolution by natural selection. 4.6 billion years

The AI is also a product of those 4,6 billion years

[Breakingthewall]

12 hours ago, r0ckyreed said:

“A object (e.g., advanced AI computer) does NOT have consciousness, it IS Consciousness.” —r0ckyreed

A object is not consciousness, it's the reality. Consciousness is the reality structured as consciousness. 

Imagine that your memory lasts zero, and you can't register change. There is no happening, always the same moment, then zero time would be the same than 1 billion of years. Would you be conscious? No. Would you be real? Yes. 

That idea about reality is "pure holy consciousness", was made by people very spiritual and very confused. As they are conscious reality is consciousness. No, reality is reality.