| User | Post |
|
9:12 am
August 19, 2010
|
twnf
| | Keremeos, BC | |
| Member | posts 34 | |
|
|
The alien beings in Blindsight have no self consciousness. This interests me. Could organisms evolve intelligent behavior without some sense of self? If so, why has self-consciousness evolved at all?
|
|
|
9:23 am
August 19, 2010
|
keanani
| | |
| Member | posts 155 | |
|
|
George Berger said:
"…let me say that Blindsight is so dense and rich philosophically, that I feel that one can learn just as much about the philosophical issues involved as you can by reading the thousands of philosophical papers devoted to these topics."
twnf said:
The alien beings in Blindsight have no self consciousness. This interests me. Could organisms evolve intelligent behavior without some sense of self? If so, why has self-consciousness evolved at all?
This interests me also twnf. I have yet to read Metzinger's book. Would one also take into account what exactly it means to "be intelligent"? From who's or what perspective?
|
The world is but a canvas to our imaginations ~ Henry David Thoreau
The real voyage of discovery consists not in seeing new landscapes but in having new eyes. ~ Marcel Proust
Fiction is a way to explore the problems of the human heart in conflict with itself…alone can make good writing because only that is worth writing about, worth the sweat and the agony. ~ William Faulkner
|
|
|
9:45 am
August 19, 2010
|
twnf
| | Keremeos, BC | |
| Member | posts 34 | |
|
|
In the context of my question I mean intelligent behavior to be behavior that optimizes the survival of the species.
|
|
|
10:56 am
August 19, 2010
|
Hljothlegur
| | |
| Moderator
| posts 367 | |
|
|
Post edited 10:56 am – August 19, 2010 by Hljothlegur
I gotta say I think that consciousness, as in, the awareness of self as a distinct point of view, having a "theory of mind" evolved because it's really helpful to social animals. If I have to negotiate a dominance heirarchy, say, or work cooperatively to gather, hunt, etc., I can anticipate the reactions of others better if I have a theory of mind. It saves me steps. I don't have to try everything to get an effect, I can imagine the future from the past, I can model other people's behavior and anticipate it.
It just happens that this turns out to be really freaking useful in other situations. Once I can model my fellow animals, I can start modeling the seasons, the tides, the behavior of my prey.
It's not that you couldn't have an advanced group of animals which do not have a mental construct of self – bees might be an example – it's that in order to get to that higher level of organization, your genes have to randomly try many more combinations to get behavior that works effectively and cooperatively without it.
Once individual animals can model, they can run through simulations of behaviors mentally, discard blind alleys quickly, and multiply the advantage their genes gave them, many times over. And add language, so the models can move from brain to brain? Pretty soon those animals take over the planet.
|
|
|
11:34 am
August 19, 2010
|
Flanders
| | |
| Member | posts 113 | |
|
|
Have you read Charles Stross' "Missile Gap?" It touches on these topics and is worth a read.
|
|
|
|
|
11:58 am
August 19, 2010
|
Hljothlegur
| | |
| Moderator
| posts 367 | |
|
|
Flanders said:
Have you read Charles Stross' "Missile Gap?" It touches on these topics and is worth a read.
I haven't! I've enjoyed so much of his work, probably because he can tell a tale.
Lemme guess – is there a plushy cthulu doll in there somewhere? A dumpy computer guy under the sway of a powerful woman? Insurance fraud? Lawsuits? Gray goo? Computer science evolved so far as to be indistinguishable from magic? Gamers? Car crashes? Is it from the POV of a female? 
|
|
|
12:57 pm
August 19, 2010
|
Flanders
| | |
| Member | posts 113 | |
|
|
More like Cold War paranoia, gonzo interstellar architecture, and sentience as an emergent phenomenon that comes and goes as required, rather than hanging around all the time making an ass of itself. I don't remember which collection it's in, but it's worth tracking down.
|
|
|
|
|
1:54 pm
August 19, 2010
|
twnf
| | Keremeos, BC | |
| Member | posts 34 | |
|
|
Hljothlegur said:
I gotta say I think that consciousness, as in, the awareness of self as a distinct point of view, having a "theory of mind" evolved because it's really helpful to social animals. If I have to negotiate a dominance heirarchy, say, or work cooperatively to gather, hunt, etc., I can anticipate the reactions of others better if I have a theory of mind. It saves me steps. I don't have to try everything to get an effect, I can imagine the future from the past, I can model other people's behavior and anticipate it.
It just happens that this turns out to be really freaking useful in other situations. Once I can model my fellow animals, I can start modeling the seasons, the tides, the behavior of my prey.
It's not that you couldn't have an advanced group of animals which do not have a mental construct of self – bees might be an example – it's that in order to get to that higher level of organization, your genes have to randomly try many more combinations to get behavior that works effectively and cooperatively without it.
Once individual animals can model, they can run through simulations of behaviors mentally, discard blind alleys quickly, and multiply the advantage their genes gave them, many times over. And add language, so the models can move from brain to brain? Pretty soon those animals take over the planet.
I agree with what you say, especially "in order to get to that higher level of organization, your genes have to randomly try many more combination to get behaviour that works effectively…". This is why I asked if intelligent behavior could evolve without the members of a species having some sense of self. The Blindsight aliens have no sense of self and yet they are intelligent. But without a sense of self they would all act according to a single determination of the best response to a given situation. There would be no competing options.
The information encountered by individual members of the species would, presumably, be gathered into a single memory base so there would be no shortage of learned information to draw on but the processing would be done according to a single set of parameters. This would result in an inflexibility to adapt behaviorally to new circumstance (just like a limited gene pool restricts the physical adaption of a species to environmental changes). While their collective consciousness might be able to assign individuals different roles in carrying out its determination, its pool of "points of view" would be limited to one (or none actually since there is no sense of self and therefore no point of view). They would be doomed to the severe evolutionary limitations of literal single-mindedness. If this is the case then, like bees, they would not have evolved to the level of intelligence they exhibit in Blindsight.
What do others think?
|
|
|
3:14 pm
August 19, 2010
|
Giant Squid
| | Oblivion | |
| Admin
| posts 8 | |
|
|
Hljothlegur said:
I gotta say I think that consciousness, as in, the awareness of self as a distinct point of view, having a "theory of mind" evolved because it's really helpful to social animals. If I have to negotiate a dominance heirarchy, say, or work cooperatively to gather, hunt, etc., I can anticipate the reactions of others better if I have a theory of mind. It saves me steps. … Once individual animals can model, they can run through simulations of behaviors mentally, discard blind alleys quickly, and multiply the advantage their genes gave them, many times over.
But why should a theory of mind– why should any kind of modelling — require consciousness? The world is full of software designed to anticipate the behavior of external agents (chess programs, chatbots — hell, my own doctorate was built around a model that predicted harbor seal behavior), and none of it has to be self-reflective in order to function.
It seems to me you're simply assuming that anything which models another's behavior must inevitably be self-aware. I don't see why this should be; the algorithms that decide such things are the same whether or not there's a little guy watching them from center stage. Hell, most of our decision-making ops are completely nonconcious.
|
|
|
4:11 pm
August 19, 2010
|
twnf
| | Keremeos, BC | |
| Member | posts 34 | |
|
|
Post edited 4:50 pm – August 19, 2010 by twnf
Metzinger maintains that any system capable of modeling aspects of the world will inevitably create a self model. If this is so then self-awareness is a kind of side-effect of the modeling activity of the brain.
|
|
|
4:06 am
August 20, 2010
|
George Berger
| | |
| Member | posts 32 | |
|
|
@Keanani—I placed a brief discussion of colour on the "Introduce Yourself" section, since the intelligence discussion is underway here.
|
|
|
7:12 am
August 20, 2010
|
Giant Squid
| | Oblivion | |
| Admin
| posts 8 | |
|
|
twnf said:
Metzinger maintains that any system capable of modeling aspects of the world will inevitably create a self model. If this is so then self-awareness is a kind of side-effect of the modeling activity of the brain.
Again, why does self-model=self-aware? All sorts of systems model themselves — the diagnostic systems of cars and airplanes, for example — but as far as we know none of them have woken up.
|
|
|
7:59 am
August 20, 2010
|
Hljothlegur
| | |
| Moderator
| posts 367 | |
|
|
Giant Squid said:
But why should a theory of mind– why should any kind of modelling — require consciousness? The world is full of software designed to anticipate the behavior of external agents (chess programs, chatbots — hell, my own doctorate was built around a model that predicted harbor seal behavior), and none of it has to be self-reflective in order to function.
It seems to me you're simply assuming that anything which models another's behavior must inevitably be self-aware. I don't see why this should be; the algorithms that decide such things are the same whether or not there's a little guy watching them from center stage. Hell, most of our decision-making ops are completely nonconcious.
Oh! I didn't mean modeling requires consciousness!
I agree that theory of mind is a form of modeling, and that software can do some modeling with no sense of I. Idk? Software is a bad example, maybe, in that you were conscious when you programmed the model and later when you looked at the results and determined what they meant – the software was a tool, like a hammer or a bowl, not the source of the model. But I can model an object in my head with no subjective sense of I unless I rotate it, so your point is agree to there.
Could we agree that the I is just a trait that exists, like eyes, fingers or skin pigment, and was therefore is either an advantage, survival-neutral on balance, or is in the process of being selected against?
Could we agree that if I'm an animal living socially and my species already has a certain degree of mental and behavioral flexibility, that I have problems that a guy in a less flexible species does not? I.e., I have to negotiate novel behaviors in others of my kind more often.
What I'm saying is that being able to generate I isn't better in the way improved predator-evasion is better. Ispeeds up the pace of change, increasing the variety of behaviors, the ability of the animal to rearrange the environment in ways that alter his fitness and reproduction. More beneficial changes and more detrimental changes faster. It's the genes giving the animal extra license to rewrite them. It's an accelerant.
"But why should a theory of mind require consciousness?" I think that is a very excellent question. Can I know your mind is a discrete-but-similar point of view to my own without being able to notice that I have a point of view myself. Hm. Again, I like the question.
I thought that twnf's was thinking, "How would Scramblers come about?" Maybe the intermediate steps from no modeling to doing higher math with no self-awareness? Twnf, did I catch your drift? I'd be really interested in that, too.
|
|
|
8:01 am
August 20, 2010
|
sheila
| | mindsided by Blindsight | |
| Moderator
| posts 515 | |
|
|
Post edited 8:19 am – August 20, 2010 by sheila
Is there anything that selects against consciousness?
(I can't remember if this was covered in blindsight, and I haven't. reread it yet.)
aside, Hljoth the other day you pointed out that all we onserve about a magpie passing the mirror test is that it passed the mirror test and we don't know why. maybe it checked for spinach on its teeth after seeing it on someone else's. this seems inconsistent with your stance above.
(edited to get rid of the green grocer apostrophe.)
|
|
|
8:08 am
August 20, 2010
|
Hljothlegur
| | |
| Moderator
| posts 367 | |
|
|
Post edited 8:24 am – August 20, 2010 by Hljothlegur
sheila said:
aside, Hljoth the other day you pointed out that all we onserve about a magpie passing the mirror test is that it passed the mirror test and we don't know why. maybe it checked for spinach on it's teeth after seeing it on someone else's. this seems inconsistent with your stance above.
Can you elaborate on that a little?
|
|
|
8:22 am
August 20, 2010
|
Hljothlegur
| | |
| Moderator
| posts 367 | |
|
|
Giant Squid said:
twnf said:
Metzinger maintains that any system capable of modeling aspects of the world will inevitably create a self model. If this is so then self-awareness is a kind of side-effect of the modeling activity of the brain.
Again, why does self-model=self-aware? All sorts of systems model themselves — the diagnostic systems of cars and airplanes, for example — but as far as we know none of them have woken up.
But they don't create the model endogenously, do they? Someone conscious programs the diagnostic systems. Maybe diagnostic systems that spontaneously generate models of themselves would suddenly be conscious? There's a scary thought.
|
|
|
8:23 am
August 20, 2010
|
sheila
| | mindsided by Blindsight | |
| Moderator
| posts 515 | |
|
|
Post edited 8:24 am – August 20, 2010 by sheila
Hljothlegur said:
sheila said:
aside, Hljoth the other day you pointed out that all we onserve about a magpie passing the mirror test is that it passed the mirror test and we don't know why. maybe it checked for spinach on its teeth after seeing it on someone else's. this seems inconsistent with your stance above.
Can you elaborate on that a little?
It was in your comment to the "Containing Within It the Seeds of Something that Will Not End Well." post.
If you look over and smile and the person you see smiles and has spinach in her teeth, you might check your own teeth because now you’re thinking about things stuck in your teeth, and this works even if the other person isn’t your reflection.
So we have no idea if the magpie has a sense of “I” have a dot on “my neck,” all we know is what he does.
(I apologize for the re-edit but I saw the green grocer apostrophe in H's quote of my post and it drove me crazy. I'm getting old. I never used to make those mistakes. sigh)
on reread, It is not inconsistent since it is only one example. but it is an example of intelligent behavior that you think might not necessarily have an aspect of consciousness.
|
|
|
8:30 am
August 20, 2010
|
sheila
| | mindsided by Blindsight | |
| Moderator
| posts 515 | |
|
|
twnf said:
The alien beings in Blindsight have no self consciousness. This interests me. Could organisms evolve intelligent behavior without some sense of self? If so, why has self-consciousness evolved at all?
Given that humans show intelligent behaviors without consciousness driving, I would say that it is possible to evolve this.
But given that those behaviors occur concomitant in a being that shows consciousness with other behaviors, is consciousness a requirement for some frequency of non-conscious intelligent behaviors?
|
|
|
8:38 am
August 20, 2010
|
sheila
| | mindsided by Blindsight | |
| Moderator
| posts 515 | |
|
|
Hljothlegur said:
But they don't create the model endogenously, do they? Someone conscious programs the diagnostic systems. Maybe diagnostic systems that spontaneously generate models of themselves would suddenly be conscious? There's a scary thought.
perhaps things would be different if you designed a system that had to model itself as it started as well as improve upon the programming to be able to adapt to new risks.
I can't think of a great example offhand, but take the robot drone that pw brought up in the seeds of evil blog post.
consider the case where we have [hand wavingly advanced technology such that it could grow new subsystems]. maybe you'd want a drone that could run diagnostics on the new parts out in the field rather than run back to the base for repairs or to have new diagnostics added for the new behaviors.
that's a bad example, because it's probably pretty cheap for a drone in that setting to fly back to base for repairs and tweaks. so we need to increase the cost of heading back to base by a huge factor. or something.
|
|
|
8:43 am
August 20, 2010
|
sheila
| | mindsided by Blindsight | |
| Moderator
| posts 515 | |
|
|
Once I was working on a project where we wanted to test a system without needing to have humans there to set up the tests. I didn't get close to the goal while working on it, but the idea is to try and have some good enough heuristics to inspect a device to get enough information about inputs in to the system such that you could generate test vectors to run through a lot of input scenarios. of course generating the test vectors is huge, so actually you don't want to generate all of them. only the ones that are most likely to find bugs. and in the priority you want to find them in.
it's been quiet a while since I worked on or thought about that, so take everything I said with a grain of salt. it could be a bunch of crap thinking.
in any case, it is the brute forcish dumb idea of having something that can inspect a system to know what to try out to test it.
but, keep in mind you can only go so far before you are running in to an unsolvable problem.
|
|
Login 
Register 
Search 
Forum
Print this Post 
Topic RSS 
