Ask HN: Is AGI 'computable' on transistor substrates? Is consciousness?
7 comments
I believe that MemGPT makes AGI possible this year, if it hasn't already happened.
Thanks to the structure inherent in writing, Word2Vec maps words into a coherent latent space.
Thanks to positional encoding, self attention mechanisms are enabled. Giving rise to next word prediction at superhuman levels in Transformer LLMs.
Thanks to synthetic data, you can train an LLM to follow directions.
Thanks to being able to train LLM s, you can generate synthetic examples of them successfully using a memory prosthesis.
They can be trained on those examples, and suddenly you have MemGPT which can be a packer[1].
It remains to be seen if a mapper can also be emulated by an LLM. I'm not sure how you could generate a ton of examples of finding an inconsistency between internal chunks of knowledge for it to train against.
[1] https://reciprocality.org/Reciprocality/r0/index.html#:~:tex....
Thanks to the structure inherent in writing, Word2Vec maps words into a coherent latent space.
Thanks to positional encoding, self attention mechanisms are enabled. Giving rise to next word prediction at superhuman levels in Transformer LLMs.
Thanks to synthetic data, you can train an LLM to follow directions.
Thanks to being able to train LLM s, you can generate synthetic examples of them successfully using a memory prosthesis.
They can be trained on those examples, and suddenly you have MemGPT which can be a packer[1].
It remains to be seen if a mapper can also be emulated by an LLM. I'm not sure how you could generate a ton of examples of finding an inconsistency between internal chunks of knowledge for it to train against.
[1] https://reciprocality.org/Reciprocality/r0/index.html#:~:tex....
This is excellent, thank you. I didn't know any of that, fascinating!
It's so funny, we're talking of AGI or consciousness like if we knew what they were but we don't. At best, we have some intuitions. And so this idea that we could reproduce something that we don't fully understand is pretty funny.
We're like the old sorcerers who wanted to turn lead into gold without having any clue about the periodic table of elements. And them too thought they were close...
We're like the old sorcerers who wanted to turn lead into gold without having any clue about the periodic table of elements. And them too thought they were close...
So the idea is, we don't have any conception of where we are heading with AGI or consciousness, no idea of what the goal end state should look like, so we can't get there?
I disagree. We do have ideas. I don't know what AGI is or what it will look like, I can just guess, but it's low confidence, it could be totally alien. But I do know what consciousness is, I have that and I can investigate that. I get if you don't feel you have that, or that ability to investigate that.
Or is the idea we don't have any conception of the engineering of how to get there, the pathways or 'constituent parts' if there are even such things, so we can't possibly hope to get there?
Again I disagree. We can figure it out. As long as we know where we're going, we'll work it out.
So I disagree on both interpretations (did you have another one?), but I think there's also a truth to what you're saying that's no so clear.
I agree with you about the humor of it, I think it's all funny. All frontier science and discovery is inherently ridiculous.
Humans are groping around in the dark. The default state is failure. It's fucking hilarious. Hahahaha! :) All the time, all we're doing is failing, and then maybe we succeed randomly somehow, and we call it genius. And we tell ourselves -- we back propagate -- our justifications for how we got there and we call ourselves insightful. Hahaha! :)
Sometimes we do have an insight, or an intuition, or an understanding, that is faithful to reality, and we understand and we use that to get there. But in another sense I think a lot of that is just models and models are wrong, but useful. So the default state is lack of understanding, that's like the default state of human experience. Hahaha! :)
So it's possible to see our striving as ridiculous human arrogance or cope. We have no fucking idea what we're doing, but we're trying. And I like that. Personally, I do. I get if you don't. But I do.
But I agree that's funny. It's hilarious. But I don't think it's unique to AGI or consciousness.
Like you say, it's similar to the alchemists. But that analogy is transitively equivalent with a helluva lot of human situations.
Even when we create it, we're not going to understand all of it, but it'll work. And that's the point. It may seem a paradox, but that's not so surprising. I think it's the nature of reality: understanding is optional. We give birth to new humans, we are literally creating new life and new consciousness with nothing more than ourselves and some grass and water. We don't understand how any of that works, really, but it works and we do it anyway.
So, fucking checkmate, any problem that we face. I think we can fuckin' do it. Hahaha! :)
But...maybe I missed your point. Maybe there was another joke in there somewhere. If so, let me know what it is! I think it's important to focus on the funny and have a great laugh about it. Hahahahaha! :)
I disagree. We do have ideas. I don't know what AGI is or what it will look like, I can just guess, but it's low confidence, it could be totally alien. But I do know what consciousness is, I have that and I can investigate that. I get if you don't feel you have that, or that ability to investigate that.
Or is the idea we don't have any conception of the engineering of how to get there, the pathways or 'constituent parts' if there are even such things, so we can't possibly hope to get there?
Again I disagree. We can figure it out. As long as we know where we're going, we'll work it out.
So I disagree on both interpretations (did you have another one?), but I think there's also a truth to what you're saying that's no so clear.
I agree with you about the humor of it, I think it's all funny. All frontier science and discovery is inherently ridiculous.
Humans are groping around in the dark. The default state is failure. It's fucking hilarious. Hahahaha! :) All the time, all we're doing is failing, and then maybe we succeed randomly somehow, and we call it genius. And we tell ourselves -- we back propagate -- our justifications for how we got there and we call ourselves insightful. Hahaha! :)
Sometimes we do have an insight, or an intuition, or an understanding, that is faithful to reality, and we understand and we use that to get there. But in another sense I think a lot of that is just models and models are wrong, but useful. So the default state is lack of understanding, that's like the default state of human experience. Hahaha! :)
So it's possible to see our striving as ridiculous human arrogance or cope. We have no fucking idea what we're doing, but we're trying. And I like that. Personally, I do. I get if you don't. But I do.
But I agree that's funny. It's hilarious. But I don't think it's unique to AGI or consciousness.
Like you say, it's similar to the alchemists. But that analogy is transitively equivalent with a helluva lot of human situations.
Even when we create it, we're not going to understand all of it, but it'll work. And that's the point. It may seem a paradox, but that's not so surprising. I think it's the nature of reality: understanding is optional. We give birth to new humans, we are literally creating new life and new consciousness with nothing more than ourselves and some grass and water. We don't understand how any of that works, really, but it works and we do it anyway.
So, fucking checkmate, any problem that we face. I think we can fuckin' do it. Hahaha! :)
But...maybe I missed your point. Maybe there was another joke in there somewhere. If so, let me know what it is! I think it's important to focus on the funny and have a great laugh about it. Hahahahaha! :)
Consciousness and intelligence are not explicitly related. Consciousness co-opts intelligence as a survival strategy.
Transistors are like switch boards and index cards. Rearranging these can produce intelligence superior to human capacity. Intelligence as inference may be statistically devised, and deduction may be combinatorially brute forced (over enough tries/long enough time line).
The relationship of information and intelligence are like memetics to consciousness. Consciousness is attracted to possibilities of outcomes and novelty.
I think consciousness is the material potential of the universe (as bound in our electro-chemical neurons) evolving through life technology to perceive and interact with itself (the rest of the universe.) In this way quantum computers may be closer to consciousness than classical transistors will ever be (which is more like rearranging a vast deck of cards.)
Transistors are like switch boards and index cards. Rearranging these can produce intelligence superior to human capacity. Intelligence as inference may be statistically devised, and deduction may be combinatorially brute forced (over enough tries/long enough time line).
The relationship of information and intelligence are like memetics to consciousness. Consciousness is attracted to possibilities of outcomes and novelty.
I think consciousness is the material potential of the universe (as bound in our electro-chemical neurons) evolving through life technology to perceive and interact with itself (the rest of the universe.) In this way quantum computers may be closer to consciousness than classical transistors will ever be (which is more like rearranging a vast deck of cards.)
This is also excellent, thank you. Very interesting!
Are they enough for consciousness? I doubt that. I don't think consciousness is 'computable' on transistors. I think we need something else. But maybe I'm totally wrong...maybe it's never "computed" on the substrated, but rather just attracted to the substrate (panpsychia / universal consciousness field), in which case it's probably possible. In that case, then the question becomes are transistors 'an expressive enough' substrate for a consciousness that we (=humans) would want to build? :)