5 ms·

Jürgen Schmidthuber (an AI researcher for those not familiar with the name) gave an interesting explanation what conciousness is in an AMA that he conducted in

Jürgen Schmidthuber (an AI researcher for those not familiar with the name) gave an interesting explanation what conciousness is in an AMA that he conducted in 2015 on Reddit. I just submitted that in an extra thread [0] because it seems worth discussing. [0] https://news.ycombinator.com/item?id=19241278 https://news.ycombinator.com/item?id=19241278

eanzenberg 8y ago
This is interesting, very interesting take, but it still doesn't explain where _other_ people's consciousness comes from. Through reductive reasoning, it doesn't explain why calculators and any matter shouldn't have consciousness.
- posterboy 8y ago
  A calculator processes maybe 64bits, in a machinery that is rather limited, certainly not enough to emulate it's environment, unless a more complex machine operates it.
  - goatlover 8y ago
    However, bits, information and software have the same problem as math that was mentioned in the article. And that is that they are abstract and relational, unlike the concrete qualities of experiencing color, pain, a dream, etc. So I doubt that any amount of bits will do any good, just like no math equation is going to become conscious. What would it even mean for an equation or bits of information to have conscious experiences?
    - posterboy 8y ago
      That's nonsense. A large part of maths is having the right feel for it, numbers are--paradoxically--qualitative. An equation is the application of a join operator, a formula is an operation, thought is operating and you are just thinking of lines scribled on paper instead.
      - goatlover 8y ago
        You're confusing the feel of doing math in your head with it being applied to describe something, which is abstract. A mathematical model isn't the thing itself.
    - natosaichek 8y ago
      However, voltages, ion channels and chemical messengers have the same problem as math. They are abstract and relational, unlike the concrete qualities of experiencing color, pain, a dream, etc. So I doubt that any amount of neurons will do any good, just like no math equation is going to become conscious. I think you're confusing the components of a map with a characteristic of the territory. We use 'bits' and 'equations' to describe the world. They do not themselves have conscious experiences, any more than a neuron. They are convenient components in a model of a larger system, which has consciousness or self-awareness as an emergent characteristic of the interactions of its components. This is like saying that because an individual spring or gear cannot tell time that a mechanical watch is an impossibility... The characteristic of interest 'maintenance of time keeping' emerges from the complex interactions of a variety of subcomponents, each of which alone has no reference to time.
- simonh 8y ago
  Calculators don't contain a representation of themselves, and since he thinks that's fundamental to consciousness, how could they be so?
- pygy_ 8y ago
  > Through reductive reasoning, it doesn't explain why calculators and any matter shouldn't have consciousness. Why shouldn't it?
jcims 8y ago
Lex Fridman has a great interview with Jurgen here - https://www.youtube.com/watch?v=3FIo6evmweo https://www.youtube.com/watch?v=3FIo6evmweo He presents this idea during the interview. It makes sense in a way. But it also makes me wonder if embodiment (whether real or virtual) of our models is necessary for them to fully grasp the contextual relevance of the information they are processing. E.g. does having a 'me' or 'I' provide a semantic anchor for everything that it processes and a 'router' if you will to get from seemlingly disconnected ideas.
- _Microft 8y ago
  Thanks!
- visarga 8y ago
  Embodiment is absolutely necessary for consciousness - for physical realisation, learning and natural selection. Here's a nice article touching on embodiment from the pov. of an AI researcher. https://medium.com/@francois.chollet/the-impossibility-of-intelligence-explosion-5be4a9eda6ec https://medium.com/@francois.chollet/the-impossibility-of-in...
naasking 8y ago
I think he's confusion consciousness for self awareness. That's not the hard problem that needs solving, it's actually one of the easy problems.
- taneq 8y ago
  I think stating that there's a "hard problem of consciousness," which is not explained by self-awareness, is begging the question. We just kind of assume that there's something inherently special about qualia, raise them on a pedestal, and claim that they are unique to humans. Why assume that 'what it's like to experience the colour red' isn't just the experience of being aware that you're experiencing a red visual input?
  - henriquemaia 8y ago
    Qualia are not just a human thing. Think of Nagel's seminal paper What is it like to be a bat?
    - taneq 8y ago
      Good point. Which makes it even more bizarre that there's any question whether animals can be self-aware.
  - simonh 8y ago
    Exactly, I just don't see what the problem is. I am not just experiencing this input. I am aware of the fact that I am experiencing this input. Or even I am aware of the fact that I am aware of myself experiencing this stimulus. Also even a simple stimulus can have all sorts of context dependent meanings and associations. Is see red. That's the colour of blood, also the colour of one of my wife's dresses, and our car. There are thousands of cross-connections between that simple input and every level of my memory and emotional life all firing off in the rich chemical soup of my brain. I just don't think the 'hard problem' is all that much of a step beyond the easy problems. Once you've solved those, the hard part looks a lot more tractable.
    - 8y ago
paraschopra 8y ago
That’s a plausible explanation for a necessary condition for subjective experience, but it’s diffult for me to accept it as sufficient. The mechanism - like integrated information theory - could account for the easy problem of consciousness but not the hard problem of consciousness. A scientific theory of consciousness would make new predictions that can be tested. For example, if we take self-encoding to be actually representative of subjective experience, can we find a novel encoding that when triggered in a subject makes them report a sensation that they’ve never experienced before? That’s how new scientific theories that propose new entities like quantum ‘spin’ are tested. New fundamental entities must have new testable predictions otherwise they’re just-so stories. That said, the easy problem is a path towards the hard one.
- visarga 8y ago
  > A scientific theory of consciousness would make new predictions that can be tested. The theory Schmidthuber mentioned is called Reinforcement Learning, a field of AI. RL also happens in biological agents, and we have known about it since Pavlov. A 'prediction' of RL is AlphaZero, the agent that beat humans at Go. Using the theories of RL we managed to accomplish what seemed impossible, and it was not just brute force, it was based on an artificial evolutionary program of agents battling each other, using neural networks to learn from past experience. It's all captured in a nice formula: https://en.wikipedia.org/wiki/Bellman_equation https://en.wikipedia.org/wiki/Bellman_equation

eanzenberg 8y ago

This is interesting, very interesting take, but it still doesn't explain where _other_ people's consciousness comes from. Through reductive reasoning, it doesn't explain why calculators and any matter shouldn't have consciousness.

posterboy 8y ago
A calculator processes maybe 64bits, in a machinery that is rather limited, certainly not enough to emulate it's environment, unless a more complex machine operates it.
- goatlover 8y ago
  However, bits, information and software have the same problem as math that was mentioned in the article. And that is that they are abstract and relational, unlike the concrete qualities of experiencing color, pain, a dream, etc. So I doubt that any amount of bits will do any good, just like no math equation is going to become conscious. What would it even mean for an equation or bits of information to have conscious experiences?
  - posterboy 8y ago
    That's nonsense. A large part of maths is having the right feel for it, numbers are--paradoxically--qualitative. An equation is the application of a join operator, a formula is an operation, thought is operating and you are just thinking of lines scribled on paper instead.
    - goatlover 8y ago
      You're confusing the feel of doing math in your head with it being applied to describe something, which is abstract. A mathematical model isn't the thing itself.
  - natosaichek 8y ago
    However, voltages, ion channels and chemical messengers have the same problem as math. They are abstract and relational, unlike the concrete qualities of experiencing color, pain, a dream, etc. So I doubt that any amount of neurons will do any good, just like no math equation is going to become conscious. I think you're confusing the components of a map with a characteristic of the territory. We use 'bits' and 'equations' to describe the world. They do not themselves have conscious experiences, any more than a neuron. They are convenient components in a model of a larger system, which has consciousness or self-awareness as an emergent characteristic of the interactions of its components. This is like saying that because an individual spring or gear cannot tell time that a mechanical watch is an impossibility... The characteristic of interest 'maintenance of time keeping' emerges from the complex interactions of a variety of subcomponents, each of which alone has no reference to time.
simonh 8y ago
Calculators don't contain a representation of themselves, and since he thinks that's fundamental to consciousness, how could they be so?
pygy_ 8y ago
> Through reductive reasoning, it doesn't explain why calculators and any matter shouldn't have consciousness. Why shouldn't it?

jcims 8y ago

Lex Fridman has a great interview with Jurgen here - https://www.youtube.com/watch?v=3FIo6evmweo https://www.youtube.com/watch?v=3FIo6evmweo He presents this idea during the interview. It makes sense in a way. But it also makes me wonder if embodiment (whether real or virtual) of our models is necessary for them to fully grasp the contextual relevance of the information they are processing. E.g. does having a 'me' or 'I' provide a semantic anchor for everything that it processes and a 'router' if you will to get from seemlingly disconnected ideas.

_Microft 8y ago
Thanks!
visarga 8y ago
Embodiment is absolutely necessary for consciousness - for physical realisation, learning and natural selection. Here's a nice article touching on embodiment from the pov. of an AI researcher. https://medium.com/@francois.chollet/the-impossibility-of-intelligence-explosion-5be4a9eda6ec https://medium.com/@francois.chollet/the-impossibility-of-in...

naasking 8y ago

I think he's confusion consciousness for self awareness. That's not the hard problem that needs solving, it's actually one of the easy problems.

taneq 8y ago
I think stating that there's a "hard problem of consciousness," which is not explained by self-awareness, is begging the question. We just kind of assume that there's something inherently special about qualia, raise them on a pedestal, and claim that they are unique to humans. Why assume that 'what it's like to experience the colour red' isn't just the experience of being aware that you're experiencing a red visual input?
- henriquemaia 8y ago
  Qualia are not just a human thing. Think of Nagel's seminal paper What is it like to be a bat?
  - taneq 8y ago
    Good point. Which makes it even more bizarre that there's any question whether animals can be self-aware.
- simonh 8y ago
  Exactly, I just don't see what the problem is. I am not just experiencing this input. I am aware of the fact that I am experiencing this input. Or even I am aware of the fact that I am aware of myself experiencing this stimulus. Also even a simple stimulus can have all sorts of context dependent meanings and associations. Is see red. That's the colour of blood, also the colour of one of my wife's dresses, and our car. There are thousands of cross-connections between that simple input and every level of my memory and emotional life all firing off in the rich chemical soup of my brain. I just don't think the 'hard problem' is all that much of a step beyond the easy problems. Once you've solved those, the hard part looks a lot more tractable.
  - 8y ago

paraschopra 8y ago

That’s a plausible explanation for a necessary condition for subjective experience, but it’s diffult for me to accept it as sufficient. The mechanism - like integrated information theory - could account for the easy problem of consciousness but not the hard problem of consciousness. A scientific theory of consciousness would make new predictions that can be tested. For example, if we take self-encoding to be actually representative of subjective experience, can we find a novel encoding that when triggered in a subject makes them report a sensation that they’ve never experienced before? That’s how new scientific theories that propose new entities like quantum ‘spin’ are tested. New fundamental entities must have new testable predictions otherwise they’re just-so stories. That said, the easy problem is a path towards the hard one.

visarga 8y ago
> A scientific theory of consciousness would make new predictions that can be tested. The theory Schmidthuber mentioned is called Reinforcement Learning, a field of AI. RL also happens in biological agents, and we have known about it since Pavlov. A 'prediction' of RL is AlphaZero, the agent that beat humans at Go. Using the theories of RL we managed to accomplish what seemed impossible, and it was not just brute force, it was based on an artificial evolutionary program of agents battling each other, using neural networks to learn from past experience. It's all captured in a nice formula: https://en.wikipedia.org/wiki/Bellman_equation https://en.wikipedia.org/wiki/Bellman_equation