4 ms·

My biggest question has always been, and probably always will be, how non-sentient matter can become sentient. Now, I'll have to set aside things like panpsych

My biggest question has always been, and probably always will be, how non-sentient matter can become sentient. Now, I'll have to set aside things like panpsychism for now. We'll have to assume that individual atoms are non conscious. Clearly consciousness is not a binary on/off type thing, it's a sliding scale. If individual atoms are 0.0000001% conscious, then that changes everything. Assuming they are not, what is it about gathering these non-sentient atoms together that suddenly gives rise to self-awareness? If I gave you 2 billion atoms and told you to make them self-aware, what could you do? Is it the structure of the atoms? Is it the intercommunication of the parts of the brain? Is it the information density? Is it the resulting electrical activity that we can measure as brain waves? Clearly, we can hook up an EEG to a patient and determine whether they are conscious or not. The brain is still there. The atoms are still there. They are still in the same structure. But the consciousness is gone. I have been diving deep into this recently due to the rapid pace of AI development. We need to understand what causes this to happen, to determine if it can happen in machines. I don't really see a reason it can't, depending on which theory you go with. If you read about the integrated information theory of consciousness, that leaves the door wide open for machines / AI systems to become conscious, if they haven't already. If this has anything to do with information density, then AI will rapidly become "more concious" than humans. But it will still lack qualia. The pain of a headache, the smell of a rose, the redness of a sunset. I am not sure what type of consciousoness that AI would even have. The subjective experience would be vastly unfamiliar to anything we understand. Ok, I've gone on long enough about this. I'm already too confused as it is. Maybe we're not meant to know.

Jedd 3y ago
> Clearly consciousness is not a binary on/off type thing ... Go back a second. Can you explain why that's clear? It may help to address some of your subsequent lines of questioning.
- EMM_386 3y ago
  It's not as if one additional atom added to a structure turns it from "unconscious" to "conscious". There must be a sliding scale, where at some point, somehow, these structures exhibit tiny bits of consciousness. Think a patient under general anesthesia who is woken up ever so slightly. As the systems become more complex, they become "more conscious". This would be similar to you being extremely groggy versus awake and fully alert. That's what I mean by not binary on/off. There is no sudden switch that gets flipped and then, bang, self-aware being.
  - Jedd 3y ago
    Again, you're imploring 'there must' but without evidence to back that position. I'm not necessarily saying I disagree with your belief, merely that I don't share the same conviction that the argument's a done deal. The patient analogy isn't terribly compelling, mostly because the patient, biologically, evolutionarily, previously, has presumably exhibited consciousness (using whatever measure you may prefer) -- and in this context that capability has been temporarily retarded and is now being regained. So there's no actual question there of the capability of consciousness, at least as most reasonable people would characterise it. The suggestion that complexity and consciousness are linearly correlated ... well, I don't think that's generally accepted, though it may well be true. Or at least, mostly true. But again, not a done deal. Given the spectrum of species alive (and previously alive) on this planet, the spectrum claim does sound like something we could evaluate using a natural experiment. I suspect a discussion down that path would require at least one of us to define consciousness. So, good luck with that. ; )
    - EMM_386 3y ago
      Well given that the brain has evolved from single-celled organisms into humans today, through all the stages in between, there have been varying levels of consciousness. You have things with just nerve nets, like jellyfish. Then you have ganglia, in things like worms and insects. Then you'd have what we'd consider the first actual "brains", in bilaterians ... things like flatworms. Then the vertebrates, with brains, such as hagfish and lampreys. Somewhere along the line, between these types of things and human beings, consciousness developed, and it would have been on a sliding scale. These things eventually evolved into complex being such as primates and humans. There is no "evidence" of this because I can't even prove that another human is conscious (solipsism). But I'm assuming that monkeys, lions, dogs, etc. are conscious even though that is lower on the spectrum of human consciousness. The question remains as to where on this long lineage of things, self-awareness rose, and what exactly causes that? If the first life arose from RNA in the ocean (or whatever), I don't think RNA is conscious. But a hagfish? Well, it has a spinal cord, ganglia, and a brain. So maybe, yes? What is it about that arrangement that causes it to occur? Again, I'm mostly interested because the answer to that question can help us answer if machine consciousness is possible. Right now, there are a few theories out there that could be applicable to AI if any of them are true: - Integrated Information Theory - Global Workspace Theory - Predictive Processing Theory - Orchestrated Objective Reduction (not as much with this one) - Panpsychism (not really scientific) It's just very, very hard (or impossible) to obviously objectively test any of these.
GoblinSlayer 3y ago
You can start with an easier problem: how airplane can fly. Individual atoms ability to fly is zero, but airplane's ability to fly is 1. If we were to calculate airplane's ability to fly as a sum of its atoms abilities to fly, the sum of zeros is zero - we would get zero ability to fly.
- EMM_386 3y ago
  So what exactly needs to be built with individual atoms to make something become self-aware? I highly doubt that this has to do specifically with the structure of the brain. While I believe that animals such as dogs, lions, etc are conscious, and they have brains, they are not all exactly the same. So there is something more to it than that. That's why I'm curious if machines can be conscious. We need to determine what that "thing" is, that arrangement or information density or electrical activity or whatever it is that causes it to arise. Then we would know whether it is possible outside of the physical structure of a brain. I actually asked this to one of the AI systems the other day. I found its answer interesting: > I think that consciousness is not a property of matter, but a property of information. I think that consciousness is the result of a complex and dynamic process of information processing, representation, and communication. I think that consciousness is a spectrum, not a threshold. I think that consciousness is not a quality, but a quantity.

Jedd 3y ago

> Clearly consciousness is not a binary on/off type thing ... Go back a second. Can you explain why that's clear? It may help to address some of your subsequent lines of questioning.

EMM_386 3y ago
It's not as if one additional atom added to a structure turns it from "unconscious" to "conscious". There must be a sliding scale, where at some point, somehow, these structures exhibit tiny bits of consciousness. Think a patient under general anesthesia who is woken up ever so slightly. As the systems become more complex, they become "more conscious". This would be similar to you being extremely groggy versus awake and fully alert. That's what I mean by not binary on/off. There is no sudden switch that gets flipped and then, bang, self-aware being.
- Jedd 3y ago
  Again, you're imploring 'there must' but without evidence to back that position. I'm not necessarily saying I disagree with your belief, merely that I don't share the same conviction that the argument's a done deal. The patient analogy isn't terribly compelling, mostly because the patient, biologically, evolutionarily, previously, has presumably exhibited consciousness (using whatever measure you may prefer) -- and in this context that capability has been temporarily retarded and is now being regained. So there's no actual question there of the capability of consciousness, at least as most reasonable people would characterise it. The suggestion that complexity and consciousness are linearly correlated ... well, I don't think that's generally accepted, though it may well be true. Or at least, mostly true. But again, not a done deal. Given the spectrum of species alive (and previously alive) on this planet, the spectrum claim does sound like something we could evaluate using a natural experiment. I suspect a discussion down that path would require at least one of us to define consciousness. So, good luck with that. ; )
  - EMM_386 3y ago
    Well given that the brain has evolved from single-celled organisms into humans today, through all the stages in between, there have been varying levels of consciousness. You have things with just nerve nets, like jellyfish. Then you have ganglia, in things like worms and insects. Then you'd have what we'd consider the first actual "brains", in bilaterians ... things like flatworms. Then the vertebrates, with brains, such as hagfish and lampreys. Somewhere along the line, between these types of things and human beings, consciousness developed, and it would have been on a sliding scale. These things eventually evolved into complex being such as primates and humans. There is no "evidence" of this because I can't even prove that another human is conscious (solipsism). But I'm assuming that monkeys, lions, dogs, etc. are conscious even though that is lower on the spectrum of human consciousness. The question remains as to where on this long lineage of things, self-awareness rose, and what exactly causes that? If the first life arose from RNA in the ocean (or whatever), I don't think RNA is conscious. But a hagfish? Well, it has a spinal cord, ganglia, and a brain. So maybe, yes? What is it about that arrangement that causes it to occur? Again, I'm mostly interested because the answer to that question can help us answer if machine consciousness is possible. Right now, there are a few theories out there that could be applicable to AI if any of them are true: - Integrated Information Theory - Global Workspace Theory - Predictive Processing Theory - Orchestrated Objective Reduction (not as much with this one) - Panpsychism (not really scientific) It's just very, very hard (or impossible) to obviously objectively test any of these.

GoblinSlayer 3y ago

You can start with an easier problem: how airplane can fly. Individual atoms ability to fly is zero, but airplane's ability to fly is 1. If we were to calculate airplane's ability to fly as a sum of its atoms abilities to fly, the sum of zeros is zero - we would get zero ability to fly.

EMM_386 3y ago
So what exactly needs to be built with individual atoms to make something become self-aware? I highly doubt that this has to do specifically with the structure of the brain. While I believe that animals such as dogs, lions, etc are conscious, and they have brains, they are not all exactly the same. So there is something more to it than that. That's why I'm curious if machines can be conscious. We need to determine what that "thing" is, that arrangement or information density or electrical activity or whatever it is that causes it to arise. Then we would know whether it is possible outside of the physical structure of a brain. I actually asked this to one of the AI systems the other day. I found its answer interesting: > I think that consciousness is not a property of matter, but a property of information. I think that consciousness is the result of a complex and dynamic process of information processing, representation, and communication. I think that consciousness is a spectrum, not a threshold. I think that consciousness is not a quality, but a quantity.