4 ms·

To me "consciousness" feels mostly like early chemists talking about pholgiston. Or like early biologistis discussing elan vital. Or even current physics' "dark

To me "consciousness" feels mostly like early chemists talking about pholgiston. Or like early biologistis discussing elan vital. Or even current physics' "dark matter". These are words that don't really point to single externalities with sharp borders, instead they are terms we apply to disparate, but seemingly (somehow) related phenomena. > The real answer is that we don't yet know enough about how the brain works... I don't think the "problem of consciousness" is one of missing empirical evidence so much as simply a fuzzy (ill-posed?) question. Though enough evidence might be sufficient to forcefully dissolve an irreal question. On the flipside, a lot of the current ML research does a good job at addressing consciousness by breaking it into concrete, communicable actions. Instead of asking, "What is consciousness?" try asking, "What actions could XYZ take that would convince me it's conscious?" A related question is "What actions (in minute detail) are involved as I believe/think/feel I'm conscious?" Those two questions are similar, but tend to evoke quite different sets of "external" experiences and actions. It might turn out that there is some mathematically invariant property of things that are capable of acting as convincing conscious agents, a la the physically precise definition of heat turning out to behave kind of like a phlogiston. In such a case, we might in fact find a "thing" that deserves the label "consciousness"---such as Hofstadter's idea of a strange loop---but for now I think the term is used pretty much used in a way that's synonymous with "magic".

Throwaway_4285 7y ago
>A related question is "What actions (in minute detail) are involved as I believe/think/feel I'm conscious?" I think we can define a generalized version of what we call a "conscious thought" as a thought that can be "consciously reflected" about. With "conscious reflection" I mean using some representation of the thought as an input for another thought in such a way that the new thought, including the usage of the original thought, can be used as an input for another thought in the same way. The representation doesn't have to correspond particularly closely to the execution. (We remember to think in natural language, but maybe the thought process is just converted into natural language after the fact for the purpose of saving and reflecting.) Computer programs can also be conscious according to this generalized definition if the process of executing it is saved in some way and can, depending on the circumstances, be used in some way, which is saved in the same way. That different "ways" of using a thought as an input for another thought are possible means that there can be different consciousnesses in our brain. One of them (if there are multiple), which is the one responsible for our output, at least on a higher level, or does at least significantly inform the output, is what we call "my consciousness". The reason why we don't know whether "my consciousness" is responsible for the output, despite having the appearance, is that it may be mostly a "post-hoc rationalization engine" for decisions made on another level, possibly for other reasons. But it does at least inform the output, since thoughts of "my consciousness" in the past inform our later output. For example, if we are asked about our past thoughts, we talk about the ones that are remembered in "my consciousness". This is the thing that puts it in such a special position compared to other hypothetical consciousnesses, which don't inform our output in this way, and are therefore invisible to other people, and obviously also to "my consciousness".
- SubiculumCode 7y ago
  This concept is called meta-cognition.
- Throwaway_4285 7y ago
  Think about it like this: What would have to change to call an conscious thought unconscious or an unconscious thought conscious? The ability of "my consciousness" to reflect about it. (Answer since edit doesn't seem to work.)
goatlover 7y ago
> On the flipside, a lot of the current ML research does a good job at addressing consciousness by breaking it into concrete, communicable actions. I wasn't aware ML had anything to do with consciousness. > In such a case, we might in fact find a "thing" that deserves the label "consciousness"---such as Hofstadter's idea of a strange loop---but for now I think the term is used pretty much used in a way that's synonymous with "magic". So you consider your experiences of color, taste, pleasure, etc. to be akin to "magic"? Because those sensations are what make up our conscious experiences.

Throwaway_4285 7y ago

>A related question is "What actions (in minute detail) are involved as I believe/think/feel I'm conscious?" I think we can define a generalized version of what we call a "conscious thought" as a thought that can be "consciously reflected" about. With "conscious reflection" I mean using some representation of the thought as an input for another thought in such a way that the new thought, including the usage of the original thought, can be used as an input for another thought in the same way. The representation doesn't have to correspond particularly closely to the execution. (We remember to think in natural language, but maybe the thought process is just converted into natural language after the fact for the purpose of saving and reflecting.) Computer programs can also be conscious according to this generalized definition if the process of executing it is saved in some way and can, depending on the circumstances, be used in some way, which is saved in the same way. That different "ways" of using a thought as an input for another thought are possible means that there can be different consciousnesses in our brain. One of them (if there are multiple), which is the one responsible for our output, at least on a higher level, or does at least significantly inform the output, is what we call "my consciousness". The reason why we don't know whether "my consciousness" is responsible for the output, despite having the appearance, is that it may be mostly a "post-hoc rationalization engine" for decisions made on another level, possibly for other reasons. But it does at least inform the output, since thoughts of "my consciousness" in the past inform our later output. For example, if we are asked about our past thoughts, we talk about the ones that are remembered in "my consciousness". This is the thing that puts it in such a special position compared to other hypothetical consciousnesses, which don't inform our output in this way, and are therefore invisible to other people, and obviously also to "my consciousness".

SubiculumCode 7y ago
This concept is called meta-cognition.
Throwaway_4285 7y ago
Think about it like this: What would have to change to call an conscious thought unconscious or an unconscious thought conscious? The ability of "my consciousness" to reflect about it. (Answer since edit doesn't seem to work.)

goatlover 7y ago

> On the flipside, a lot of the current ML research does a good job at addressing consciousness by breaking it into concrete, communicable actions. I wasn't aware ML had anything to do with consciousness. > In such a case, we might in fact find a "thing" that deserves the label "consciousness"---such as Hofstadter's idea of a strange loop---but for now I think the term is used pretty much used in a way that's synonymous with "magic". So you consider your experiences of color, taste, pleasure, etc. to be akin to "magic"? Because those sensations are what make up our conscious experiences.