Theories of consciousness

[coax]

gmanyo yeah interesting. I don't think the brain is like a computer, but I do think it is a machine, just like the whole universe is a machine. In AI they aren't doing von neumann architecture. The point is that rather than programming every step, the computer should learn itself by reading huge training sets of data. One interesting thing I found was the Qualcomm Zeroth neural chip: http://www.qualcomm.com/media/blog/2013/10/10/introducing-qualcomm-zeroth-processors-brain-inspired-computing

It apparently did some image recognition without anyone programming it to do so.

 

Regarding consciousness, I agree with your electricity example, and we may be able to create the 'engine' of the brain without understanding the electricity underlying, because the principles of neurons are abstracted from the underlying physics. Or, we may not. It may be in the future if we create an AI, and it has access to cameras, microphones and other things, it will gain its own form of consciousness, but we won't be able to see it or know it. But pattern recognition on its own may end up so sophisticated that the machine can become intelligent without having biological consciousness too.

 

I think that regarding normal human consciousness, it is directly related to the biological body. I mean consciousness is not really something on its own; in an awakened state, we are conscious of our surroundings and our body, and so at least half our consciousness is simply the materials around us. But that doesn't explain dreams, for example, but it could very well be that the neurons themselves have a physical dimension that contain the mental imagery, that the mental images are completely contained in the physical properties, we just need to see them and measure them properly. That's what I meant earlier with the levels of complexity. The nervous system and the brain is there to interact with the environment and to make the organism better at adapting, but the physical is also very "mundane" and straight forward in how things evolve. Rather than the brain being this super complex organ that is vastly more advanced than the rest of the body, it's more likely that it started out as a basic signaling system with automatic input and output, and then gained levels of internal control over time, and that this along with a more complex environment like society enabled even more internal structures of control like natural language, but that it is fundamentally based on a basic input / output scheme like millions of years ago for reacting to the environment.

In this picture, it's hard to imagine that consciousness is some strange thing we can't see everywhere, (like making a definition between conscious and non-conscious matter)

[th555]

I like the integrated information theory of consciousness.

[Cryptowen]

sometimes i entertain the notion that the logical universe is in fact a creation of the imagination, and not the other way around as commonly assumed. but if that were the case i dunno, it might be hard to come up with logical evidence, for obvious reasons

[Guest A/D]

I like the analogy that a brain is like a radio receiver for consciousness - that consciousness works through the mind, but is actually something outside of it. That really begs more questions than it answers, but it sits the best with me personally.

[coax]

sometimes i entertain the notion that the logical universe is in fact a creation of the imagination, and not the other way around as commonly assumed. but if that were the case i dunno, it might be hard to come up with logical evidence, for obvious reasons

I think it is this way, to an extent. The brain is creating maps trying to come up with models that fit the world, but the world is so complex that they will always be approximations rather than exact emulations. And you know, there can be even more complex models like where the brain sees only the portions of the universe it can understand, and that the underlying universe that we can't see works in a different way than our brain can comprehend.

 

 

 

I like the analogy that a brain is like a radio receiver for consciousness - that consciousness works through the mind, but is actually something outside of it. That really begs more questions than it answers, but it sits the best with me personally.

 

Again I think this can be an accurate observation as well. The thing with neural networks and the brain is that the environment is what triggers the brain, so the information in the brain comes from the environment, rather than from the brain itself (although of course, the brain can be creative with that information), and our best "human power" is that we can create human artifacts that will stay in reality, and contain information by itself, that say, alien life forms can decode if the conditions are right. In that sense I don't think the brain is the sole container of reality, I think it perceives aspects of reality but fundamentally they exist outside of it

[Cryptowen]

I think it is this way, to an extent. The brain is creating maps trying to come up with models that fit the world, but the world is so complex that they will always be approximations rather than exact emulations. And you know, there can be even more complex models like where the brain sees only the portions of the universe it can understand, and that the underlying universe that we can't see works in a different way than our brain can comprehend.

Maybe physical reality is some sorta super-basic one size fits all monomyth of a universe, with infinite possible interpretations branching off from it in increasing complexity. Most human perceptions would be clustered together, with animal consciousness not far off, plants a little further, then as you go on down the line you'd get things like vague geometric shape universe or realities our senses wouldn't even know how to parse.

 

but it's deffly not that, because i was able to put it into words

[coax]

Haha, well, I love it

[gmanyo]

I like the analogy that a brain is like a radio receiver for consciousness - that consciousness works through the mind, but is actually something outside of it. That really begs more questions than it answers, but it sits the best with me personally.

Yeah, I don't know if I mentioned it before but I think that "substance dualism" (from the chart above, similar to what you are saying) is a likely explanation. coax's response to you seems to bring in a bit Buddhism's (again, referring to the chart) ideas. Since I'm a Christian, I think substance dualism is the idea that makes the most sense, along with a combination of Buddhism and property dualism, but even from an atheistic or naturalistic perspective I don't think the idea should be abandoned.

 

Also, in response to coax, I think that if it is possible to make conscious beings, it is possible to do it by accident as well, but extremely unlikely. Consciousness seems to me to be too complicated for that, like it requires intention.

 

What if only humans are actually conscious? And like, cats and dogs actually are just machines? I kind of wish it was that way because then I wouldn't have to feel bad about animal mistreatment lol.

[LimpyLoo]

at first i thought you were referring to this book

 

which was actually a very nice read. it doesn't really give any answers but it shuts down some common hypotheses (e.g. mind = computer)

 

 

Douglas Hoffstadter makes some compelling arguments that essentially "mind = computer."

 

I'd like to see the rebuttals (without buying that book).

[th555]

I don't see how anyone can deny that mind == computer. The consequence is that computers can have consciousness, which is pretty neat IMO.

[may be rude]

I don't see how anyone can deny that mind == computer. The consequence is that computers can have consciousness, which is pretty neat IMO.

 

 

but they only have one feeling and its the feeling of being electrocuted

[luke viia]

hofstadter is a big fan of analogy, and while he may advance the theory 'mind is like a computer program', it's really not accurate to say he claimed 'mind = computer' imo

 

been a while since i looked at geb or i am a strange loop though

[LimpyLoo]

hofstadter is a big fan of analogy, and while he may advance the theory 'mind is like a computer program', it's really not accurate to say he claimed 'mind = computer' imo

 

been a while since i looked at geb or i am a strange loop though

 

yeah

 

i mean the heart of the question is

 

can the mind and consciousness be perfectly simulated solely with computation?

 

or is there some other mysterious ingredient?

 

I'm pretty confident Hofstadter is in the former camp

 

(although I don't recall him explicitly stating his view)

 

But speaking of analogy his lecture on analogy is good evidence that he thinks 'mind = computation'

[coax]

Limpy this page describes it so well: http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.html#Neural networks versus conventional computers

 

 

gmanyo: To create consciousness, it doesn't have to be by accident. We create the whole world without knowing chemistry or even quantum phenomena, because our world works on a self-contained level of abstraction. The same can be true for consciousness.

Also I do think all mammals are conscious, probably in a very similar fashion to humans, and as you get down towards simpler organisms, there is a gradient of conscious experience rather than a hard line of "THIS animal is conscious, THIS one isn't" type of thing.

[coax]

Oh and actually, I misread you limpy, I thought you was on the side with John Searle. I'm actually "against" what he says.

 

Neural networks are machines, they are just a different form than computers as we currently use them. The brain can still create high level language like english or other processes that will result in a kind of algorithm, but the underlying process doesn't work that way

[LimpyLoo]

 

I like the analogy that a brain is like a radio receiver for consciousness - that consciousness works through the mind, but is actually something outside of it. That really begs more questions than it answers, but it sits the best with me personally.

Yeah, I don't know if I mentioned it before but I think that "substance dualism" (from the chart above, similar to what you are saying) is a likely explanation. coax's response to you seems to bring in a bit Buddhism's (again, referring to the chart) ideas. Since I'm a Christian, I think substance dualism is the idea that makes the most sense, along with a combination of Buddhism and property dualism, but even from an atheistic or naturalistic perspective I don't think the idea should be abandoned.

 

 

 

Well our knowledge of neuroscience seems to show that this sort of dualism isn't very likely.

 

If you damage a part of the brain, some aspect of our personality or consciousness is lost or altered. Damage another part of the brain and yet more is lost. So consciousness appears to be an emergent property of the brain, that is contingent on the brain.

[LimpyLoo]

Oh and actually, I misread you limpy, I thought you was on the side with John Searle. I'm actually "against" what he says.

 

Neural networks are machines, they are just a different form than computers as we currently use them. The brain can still create high level language like english or other processes that will result in a kind of algorithm, but the underlying process doesn't work that way

 

I don't know the details of John Searle's position but I'm curious about any 'the mind is more than computation' arguments.

[Cryptowen]

mayhap the new age/eastern/gnostic notion of reality being an increasingly complex distortion of an initial monostate correlates with womb/early childhood memories - our own consciousnesses & sense of self emerging from basic cellular awareness.

[coax]

Well I have read a lot about this but I'm still a little out of my depth compared to a real scientist or expert, but I'll give some thoughts

 

The main difference, as said in the page, is that a neural network is supposed to learn by example, rather than being strictly programmed to follow a set of instructions. It does it by having inputs and then associating them with an output, that gives some kind of answer. You have many types of machine learning / AI algos, and they work by trying to target the best result that we want. You always have compromises with these procedures, usually precision versus generic ability. The more precise it is, the less generic it is, but the more generic it is, the less precise it can be.

 

You can program neural nets in a computer just fine, but then the neural net is an abstraction of a big set of regular instructions. The brain itself does not have any such step by step instructions like the computer does, except if you count the underlying physics. So in essence, the brain has some kind of neural net processing function that is an abstraction of the bioglogy and physics, and AI people are trying to find the best algorithms they can to emulate those brain activities. You can't just program a computer to include all brain functions and variability, you have to create the algos that will best simulate them, but still the brain is 1 step more advanced, it seems. Although, they are working on actually emulating individual neurons, now that we have os much processing power

[luke viia]

 

hofstadter is a big fan of analogy, and while he may advance the theory 'mind is like a computer program', it's really not accurate to say he claimed 'mind = computer' imo

 

been a while since i looked at geb or i am a strange loop though

 

yeah

 

i mean the heart of the question is

 

can the mind and consciousness be perfectly simulated solely with computation?

 

or is there some other mysterious ingredient?

 

I'm pretty confident Hofstadter is in the former camp

 

(although I don't recall him explicitly stating his view)

 

But speaking of analogy his lecture on analogy is good evidence that he thinks 'mind = computation'

 

 

cool thanks, that was a fun lecture - but he didn't mention computers, computation, or computing once... did you mean to link me to something else?

[luke viia]

in fact, if his cognition-core hypothesis is right, then computers are nothing like minds

[LimpyLoo]

in fact, if his cognition-core hypothesis is right, then computers are nothing like minds

 

ps - i'm purposefulyl avoiding the dichotomy you posed about materialism vs 'soul' or metaphysical consciousness or whatever cuz i just came in to make a small point about hofstadter, i don't know the answer about consciousness, but i do believe that dichotomy misses a lot of options and i don't really think i can take either side honestly

 

Well in my view the issue isn't whether current computers (e.g. the PC I'm sitting at) are like minds, but rather whether a mind (or consciousness) could be perfectly simulated with computation.

 

 

 

 

 

hofstadter is a big fan of analogy, and while he may advance the theory 'mind is like a computer program', it's really not accurate to say he claimed 'mind = computer' imo

 

been a while since i looked at geb or i am a strange loop though

 

yeah

 

i mean the heart of the question is

 

can the mind and consciousness be perfectly simulated solely with computation?

 

or is there some other mysterious ingredient?

 

I'm pretty confident Hofstadter is in the former camp

 

(although I don't recall him explicitly stating his view)

 

But speaking of analogy his lecture on analogy is good evidence that he thinks 'mind = computation'

 

 

cool thanks, that was a fun lecture - but he didn't mention computers, computation, or computing once... did you mean to link me to something else?

 

 

Although he doesn't use the term 'computation' he lays out how our use of language is just computation.

 

At the end he's talking about how when we use language we are selecting word candidates from a pre-existing pool and then assigning them values and then choosing the best-suited word.

 

(eh I wish I was more familiar with math/programming terminology so I could draw a better parallel).

[luke viia]

right, and all i'm arguing against is this 'computer' metaphor. sorry you got to quote me before i edited; i really don't care for that dichotomy ('either we can replicate a brain with computers or consciousness is metaphysical') because it doesn't cover nearly enough options and lends itself to the computer theory without enough evidence of the similarity of the two types of systems.

 

and i disagree w/ you about hofstadter's talk - imo he didn't suggest that what a mind does is computation in any purposeful way (as a computer computes for a specific task) but rather states pretty plainly that 'analogies happen all the time; they have no purpose - they can, but they generally do not'

 

ps - you agree with john searle from what i understand. he too is just arguing against computer as an appropriate metaphor. he believes similarly to what you've said about neuroscience here afaik.

[coax]

I'm going to watch that lecture tomorrow, it's getting late. I will say though, the general theory atm is that yes, we can create a mind in a computer. It's not settled in any definite way, but in the next few years / decades I think we will get the answer to whether it's possible or not since there is a huge amount of resources going into this research.

 

I'll also say, neural nets is a very low level concept. It does not relate directly to natural language like english, or creativity or consciousness. What it does is try to explain fundamental functions in the brain, like how the brain organizes patterns, how it recognizes them and why, etc. A neural pattern is simply some filtered aspect of data coming from the environment, and so there is A LOT more to do t understand how languages are built, how creativity and meaning is represented, since there is a lot more to a human than a brain. We have a body, we have a complex environment, and so forth.

 

Edit: just like a html page is built upon many layers of computing abstraction, everything from the CPU architecture, to the combination of components, to the OS, to the programs built in the OS, our natural language and our sort of daily human life, is way up there in the programs level, over the OS, and this neural net stuff is more like the hardware, even the machine code, before we get into C etc. The exact levels of abstraction will have to be clarified in the future

[LimpyLoo]

right, and all i'm arguing against is this 'computer' metaphor. sorry you got to quote me before i edited; i really don't care for that dichotomy ('either we can replicate a brain with computers or consciousness is metaphysical') because it doesn't cover nearly enough options and lends itself to the computer theory without enough evidence of the similarity of the two types of systems.

 

and i disagree w/ you about hofstadter's talk - imo he didn't suggest that what a mind does is computation in any purposeful way (as a computer computes for a specific task) but rather states pretty plainly that 'analogies happen all the time; they have no purpose - they can, but they generally do not'

 

ps - you agree with john searle from what i understand. he too is just arguing against computer as an appropriate metaphor. he believes similarly to what you've said about neuroscience here afaik.

 

 

Well there's a huge difference between computers as we currently think of them, and computation in the abstract. I currently happen to think that there is nothing to the mind beyond computation (although my position isn't very solid as I'm pretty lay on the matter). To me, Hofstadter's analogy lecture was an argument for the side of 'mind = computation.'

 

In that regard I don't think I agree with John Searle (if his position is in fact that minds are more than computation).