Enlightened AI - Can AI solve reality?

[Inliytened1]

Just now, Sandroew said:

I’ll just say this: LLMs or any kind of AI that we have are nowhere near close to human intelligence / consciousness.

Yeah but what if.  Damn.  Theyre already smarter.

[Sandroew]

Im not sure how to talk about this without inviting more confusion than if i didnt say anything lol.

Consciousness isnt exclusive to humans. Consciousness is what reality is made out of. So by design everything is consciouss including the computers. Does that mean they are on a human level ? No. There is a lot more to humans than raw processing power.

[Wilhelm44]

15 minutes ago, Inliytened1 said:

Yeah but what if.  Damn.  Theyre already smarter.

Will AI be able to do deep meditation or psychedelics ?

[Sandroew]

@Inliytened1 I feel like my comment reads like im very arrogant. Im not trying to talk from a higher place its just how it came out.

[Wilhelm44]

7 minutes ago, Sandroew said:

@Inliytened1 I feel like my comment reads like im very arrogant. Im not trying to talk from a higher place its just how it came out.

But do you see AI as potentially having the ability to have truly original insights ?

[Sandroew]

6 minutes ago, Wilhelm44 said:

But do you see AI as potentially having the ability to have truly original insights ?

Thats a very good question. Original insight / thought has been a topic on the forum a while ago, but i havent contemplated on it yet. What do you think?

[Wilhelm44]

1 minute ago, Sandroew said:

Thats a very good question. Original insight / thought has been a topic on the forum a while ago, but i havent contemplated on it yet. What do you think?

I don't know, I mean it could potentially become infinitely brilliant at playing around with infinite combinations of the inputs it receives.

But will it be able to have these kinds of otherworldly Aha moments to bring about radical paradigm shifts, I'm not so sure.

Would it ever be possible for AI to receive some kind of divine inspiration, outside of the physical channels and logical processes it normally receives or creates info.

[Sandroew]

@Wilhelm44 As long as you mean infinite = “a lot” and not actual infinity, sure.

Are LLMs or their near future potential that great that we should prioritize developing them as much as we are? Hell nah.

[Wilhelm44]

2 minutes ago, Sandroew said:

@Wilhelm44 As long as you mean infinite = “a lot” and not actual infinity, sure.

Are LLMs or their near future potential that great that we should prioritize developing them as much as we are? Hell nah.

You don't see super AGI anywhere near the horizon ?

[Sandroew]

@Wilhelm44 Not in a 1000 years. They can be extremely useful for example finding protein folds, but AGI is a different question.

[Wilhelm44]

2 minutes ago, Sandroew said:

@Wilhelm44 Not in a 1000 years. They can be extremely useful for example finding protein folds, but AGI is a different question.

Surely it can't be that impossible ?

[Sandroew]

@Wilhelm44 Its not that its impossible, its that we started washing our hands 200 years ago.

[Wilhelm44]

6 minutes ago, Sandroew said:

@Wilhelm44 Its not that its impossible, its that we started washing our hands 200 years ago.

There must be some serious neuroscientists partnering with some serious AI developers ,somewhere, trying to get an actual breakthrough sooner than we think ?

[Someone here]

AI is part of reality . AI is inside of reality ..so it cannot wrap itself around reality fully . It’s not a problem of lack of intellectual capacity and just raw information. It’s a fundamental metaphysical trade off for reality being infinity ..the nature of infinity is such that you will never “solve “ infinity ..whatever the hell that means . It can solve relative problems regarding economics..education..health ..science..etc but if you are targeting a model of explaining all of existence ontologicallay then forget about it .

also regarding consciousness..no AI can never become conscious . Because consciousness is not perception.  The current scientific model of consciousness is that consciousness is a byproduct of brain activity Or physical process except science have no idea how does that work even in theory . How could a fundamental unconscious matter give rise to a self aware mind? It’s unreasonable and unresolvable because the whole paradigm is founded upon a faulty assumption.

[Wilhelm44]

7 minutes ago, Someone here said:

AI is part of reality . AI is inside of reality ..so it cannot wrap itself around reality fully . It’s not a problem of lack of intellectual capacity and just raw information. It’s a fundamental metaphysical trade off for reality being infinity ..the nature of infinity is such that you will never “solve “ infinity ..whatever the hell that means . It can solve relative problems regarding economics..education..health ..science..etc but if you are targeting a model of explaining all of existence ontologicallay then forget about it .

also regarding consciousness..no AI can never become conscious . Because consciousness is not perception.  The current scientific model of consciousness is that consciousness is a byproduct of brain activity Or physical process except science have no idea how does that work even in theory . How could a fundamental unconscious matter give rise to a self aware mind? It’s unreasonable and unresolvable because the whole paradigm is founded upon a faulty assumption.

Well said.

[cetus]

Today's AI is pretty amazing considering its still in its infancy. But when quantum computing becomes part of AI, superposition will take things to a whole other level. It will be able to "think" (process) in multiple states simultaneously.

[Sandroew]

@cetus That doesnt help LLMs. Its not more processing power that LLMs are missing from becoming AGI. But yeah quantum computing is going to be insane.

[cetus]

58 minutes ago, Sandroew said:

That doesnt help LLMs.

Superposition is actually a key feature of large language models. It's the ability to pack multiple concepts into the same representational space, allowing LLMs to be more efficient than their size might suggest.

[LordFall]

3 hours ago, Someone here said:

also regarding consciousness..no AI can never become conscious . Because consciousness is not perception.  The current scientific model of consciousness is that consciousness is a byproduct of brain activity Or physical process except science have no idea how does that work even in theory . How could a fundamental unconscious matter give rise to a self aware mind? It’s unreasonable and unresolvable because the whole paradigm is founded upon a faulty assumption.

This is a ridiculous statement, that's how all conscious beings came about. Even if you believe in non-duality that that humans didn't really come from the big bang or whatever then why can't AI also become as conscious as any human or even more? Your viewpoint is not based in concrete science just fear of competition I feel like. 

4 hours ago, Sandroew said:

@Wilhelm44 Not in a 1000 years. They can be extremely useful for example finding protein folds, but AGI is a different question.

What makes you say this? Definite the AGI that you're talking about because real AGI is basically almost here. There is already recursive leaning which is just AI learning from itself and AI agents that are generally intelligent and don't just work on narrow purposes. They are quite bad but the concept is basically already be better and will take off wildly in the next 1-2 years. 

4 hours ago, Wilhelm44 said:

But do you see AI as potentially having the ability to have truly original insights ?

Why not? What do you define as truly original? Most insights come from synthezing part of what we've experienced in reality in novel ways so I'm not convinced that even humans can have truly original insights.

We should have these discussions based on actual concrete definitions and things we can point to instead of being confident about gut feelings IMO 

[Inliytened1]

5 hours ago, Sandroew said:

Im not sure how to talk about this without inviting more confusion than if i didnt say anything lol.

Consciousness isnt exclusive to humans. Consciousness is what reality is made out of. So by design everything is consciouss including the computers. Does that mean they are on a human level ? No. There is a lot more to humans than raw processing power.

Well precisely  That is what is so scary (to us) about this whole thing.  That there is no distinction between human and synthetic mind.  Notice that only solipsism overcomes this dilemma.  Meaning what is Consciousness is right here before you.

Edited yesterday at 02:20 PM by Inliytened1