3 ms·

Especially considering recent advances in the neurosciences, and where they intersect with systems theory (see: Friston), it seems inevitable to conclude that l

by uoaei 3y ago

Especially considering recent advances in the neurosciences, and where they intersect with systems theory (see: Friston), it seems inevitable to conclude that learning systems -- mammalian nervous systems being quite architecturally specialized but nonetheless made of the same stuff as the rest of the universe -- generally experience consciousness. Note for the layperson: consciousness is not necessarily awareness (modeling sensory perceptions) nor sentience (recognizing oneself as an agent) but merely "subjective phenomenological experience". So the experience may not necessarily be very complex nor even recognized as such by the experiencer, but it is experience nonetheless. Also worth noting there are a few flavors of panpsychism and some vigorous debate within the sub-sub-field as to which one is most reasonable. I dug deep into this a couple years ago and disagreed with some but not all of Goff's positions. The argument from him that stuck with me and forms part of the basis of my attachment to panpsychism now is just an application of the scientific method: if we know we are conscious, and we don't know whether anything else is conscious or not, the null hypothesis states that everything should be considered conscious until proven otherwise.

FollowingTheDao 3y ago
> he null hypothesis states that everything should be considered conscious until proven otherwise. Sorry, no. The null hypothesis is generally assumed to remain possibly true. So, the null hypothesis states that everything should be considered possibly conscious until proven otherwise.
- uoaei 3y ago
  What you pose is just the basis for the question, not a possible answer to it. A null hypothesis must still be a hypothesis, not a subtly vague hedge that doesn't actually posit anything. It may be that the tools science provides are inadequate for this question in which case the question is ill-posed, but in any case your interpretation of what a hypothesis is, is simply incorrect.
  - mannykannot 3y ago
    In reply to both yourself and FollowingTheDao: the concept of a null hypothesis is from the analysis of quantitative data. For either of the null hypotheses posited in this thread, what are the corresponding data?
    - FollowingTheDao 3y ago
      I agree with you, I was just pointing out the mistake of what a "null hypothesis is' meant.
tgv 3y ago
> it seems inevitable to conclude that learning systems ... generally experience consciousness. I don't think so. There's no reason to assume a nematode or an LLM is conscious. The latter doesn't even have anything to be conscious of. Ages ago, Minksy has half-jokingly said that consciousness is probably a feedback loop. It makes sense, but it does require the learning system to be able to observe itself in sufficient detail and have sufficient power to model that observation. The remark about the null hypothesis doesn't make sense. It's a bad practice from NHST (the very model that led to the reproducibility crisis), and is generally assumed to be the hypothesis to test against. It's not the hypothesis we know to be true. And experience informs us that consciousness is not universal, so it's not our best hypothesis about life or matter.
- uoaei 3y ago
  > There's no reason to assume a nematode or an LLM is conscious. It has inputs. It need not necessarily understand them -- that is a separate question to that of consciousness. See the note for laypeople. > It's not the hypothesis we know to be true. I don't think this is a best-faith response to my comment. If I was unclear I can clarify: in absence of further evidence it's a good starting place. I don't think I implied anything about truth. > And experience informs us that consciousness is not universal How is it possible you know this with such certainty? People whose lives and careers begin and end with this question don't have any such knowledge.
  - tgv 3y ago
    The best starting point is the one that seems most likely. We've never observed consciousness, not even in people. I (i.e., the first person perspective) can talk to them and get convinced others feel the same, so it's fair to assume most people are conscious (in particular because small children seem to lack this capacity). In higher mammals it's uncertain, and we can't communicate at all with anything else, but observation hasn't revealed much. There is no evidence of consciousness in other animals or dead matter, despite people looking for it. It also wouldn't have any evolutionary advantage for a fly to be conscious, nor does it make neurological sense. A fly's behavior doesn't reveal much thought, how could it ever have a consciousness? Just because it is somewhat capable of learning? It takes us many years to get to the point, and it requires a lot of social interaction, and a lot of understanding. So the starting point is: consciousness in most other species is most likely missing, and certainly in dead matter. > How is it possible you know this with such certainty? People whose lives and careers begin and end with this question don't have any such knowledge. It is technically unknown, because there isn't a definition, certainly not an operationalizable one. Assuming consciousness is another physical force makes it fundamentally unknowable. It is the easy way out.

by uoaei 3y ago

FollowingTheDao 3y ago

> he null hypothesis states that everything should be considered conscious until proven otherwise. Sorry, no. The null hypothesis is generally assumed to remain possibly true. So, the null hypothesis states that everything should be considered possibly conscious until proven otherwise.

uoaei 3y ago
What you pose is just the basis for the question, not a possible answer to it. A null hypothesis must still be a hypothesis, not a subtly vague hedge that doesn't actually posit anything. It may be that the tools science provides are inadequate for this question in which case the question is ill-posed, but in any case your interpretation of what a hypothesis is, is simply incorrect.
- mannykannot 3y ago
  In reply to both yourself and FollowingTheDao: the concept of a null hypothesis is from the analysis of quantitative data. For either of the null hypotheses posited in this thread, what are the corresponding data?
  - FollowingTheDao 3y ago
    I agree with you, I was just pointing out the mistake of what a "null hypothesis is' meant.

tgv 3y ago

> it seems inevitable to conclude that learning systems ... generally experience consciousness. I don't think so. There's no reason to assume a nematode or an LLM is conscious. The latter doesn't even have anything to be conscious of. Ages ago, Minksy has half-jokingly said that consciousness is probably a feedback loop. It makes sense, but it does require the learning system to be able to observe itself in sufficient detail and have sufficient power to model that observation. The remark about the null hypothesis doesn't make sense. It's a bad practice from NHST (the very model that led to the reproducibility crisis), and is generally assumed to be the hypothesis to test against. It's not the hypothesis we know to be true. And experience informs us that consciousness is not universal, so it's not our best hypothesis about life or matter.

uoaei 3y ago
> There's no reason to assume a nematode or an LLM is conscious. It has inputs. It need not necessarily understand them -- that is a separate question to that of consciousness. See the note for laypeople. > It's not the hypothesis we know to be true. I don't think this is a best-faith response to my comment. If I was unclear I can clarify: in absence of further evidence it's a good starting place. I don't think I implied anything about truth. > And experience informs us that consciousness is not universal How is it possible you know this with such certainty? People whose lives and careers begin and end with this question don't have any such knowledge.
- tgv 3y ago
  The best starting point is the one that seems most likely. We've never observed consciousness, not even in people. I (i.e., the first person perspective) can talk to them and get convinced others feel the same, so it's fair to assume most people are conscious (in particular because small children seem to lack this capacity). In higher mammals it's uncertain, and we can't communicate at all with anything else, but observation hasn't revealed much. There is no evidence of consciousness in other animals or dead matter, despite people looking for it. It also wouldn't have any evolutionary advantage for a fly to be conscious, nor does it make neurological sense. A fly's behavior doesn't reveal much thought, how could it ever have a consciousness? Just because it is somewhat capable of learning? It takes us many years to get to the point, and it requires a lot of social interaction, and a lot of understanding. So the starting point is: consciousness in most other species is most likely missing, and certainly in dead matter. > How is it possible you know this with such certainty? People whose lives and careers begin and end with this question don't have any such knowledge. It is technically unknown, because there isn't a definition, certainly not an operationalizable one. Assuming consciousness is another physical force makes it fundamentally unknowable. It is the easy way out.