On Consciousness

[Belz...]

Sounded like you were complaining that those who don't agree with SRIP aren't posting their own definitions. That was my impression, anyway. I might be confusing you with another poster.





Post 2531

So:

1. It leads to absurdities no one takes seriously (conscious cars?)
2. It is undefined (self, information, and processing are unclear terms)
3. Authorities who study brains don't define consciousness as SRIP
4. It does not explain the phenomenon of subjective experience

1. Is an opinion. It doesn't make it "not work"
2. Is only true in the sense that there are broad and narrow definitions of these terms.
3. Perhaps. There seems to be some disagreement about that, here. But it doesn't mean that it doesn't work.
4. You need to explain that one to me. What doesn't it explain, once we get rid of any dualist baggage ?

 

[Fudbucker]

The neuroscientists that Annnoid keeps quoting, when addressing the question of what consciousness is: "Consciousness has not yet become a scientific term that can be defined in this way."

"This", from Annoid's oft repeated quote, refers to: "a careful and precise definition of consciousness". They're right. We don't have a careful and precise definition. We have a bunch of definitions that are pretty vague and general.

 

[Fudbucker]

You mean you couldn't even be bothered to go back a few posts and see what you said ?



I'll take that as a retraction.

I'm not retracting anything. You made a dismissive comment regarding philosophy AFTER someone you obviously respect posted an expert opinion from a philosopher (and continues to do so). Maybe it's nit-picky, but it illustrates my point that blinders go up when someone from your "camp" posts. With your position on philosophy, you must have an opinion on citing Dennett as an authority. What is it?

ETA: Retracted. Skimming through the posts (and a few others here from R&P section), I see you've challenged Pixy on issues.

That was very nit-picky on my part. Sorry.

 

[Fudbucker]

It's difficult to define in a way that satisfies you.



How doesn't it work ?

It obviously doesn't satisfy the medical community. They don't use it. Nor have I ever seen it used in peer-reviewed journals.

 

[Fudbucker]

So:



1. Is an opinion. It doesn't make it "not work"

No, there's no evidence that cars are conscious and plenty of contradictory evidence they aren't (they don't possess any known mechanism that would enable them to be conscious). Reductio ad absurdum.

I also suspect that SRIP goes on in very basic organisms, like plants and probably bacteria. These are obviously not conscious. There are also processes that we are not conscious of (e.g., digestion) that clearly involve SRIP.

2. Is only true in the sense that there are broad and narrow definitions of these terms.

Well, we could create a whole new post on what "self" means. Is it a bundle of perceptions? Eddie in the stream?

3. Perhaps. There seems to be some disagreement about that, here. But it doesn't mean that it doesn't work.

Where's the disagreement? I spent some time on Google, and I can't find any medical research articles that define consciousness as SRIP. Now, you can complain that I'm being overly restrictive by focusing on medical research (or psychological research), but what is the only thing we're sure produces consciousness? Brains.

And how does SRIP "work"?

4. You need to explain that one to me. What doesn't it explain, once we get rid of any dualist baggage ?

I might retract that one. Subjective experience itself is so poorly defined. Again, we know it when we feel it, but how do you put in words?

 

[Fudbucker]

That's the thing about Dennett and Hofstadter's work: They clear out all the nonsense that's been cluttering up the conversation. Before, those questions weren't even meaningful. Now they're not only meaningful, they're simple, direct factual questions.

I think they've added to the discussion. Whatever consciousness is, it has to be a kind of information processing. That seems to be what brains do. I see Rocketdodger clamoring that programmers (AI, esp.?) need to be in on the discussion. Maybe he has a point.

I'm not sure it has to refer to a self (basic sensory experience often doesn't have a "self" component. I brought this up before and wasn't really satisfied with your answer- seemed too recursive).

I think SRIP may be required for consciousness. But it doesn't seem to be sufficient.

One thing I think you take some undeserved criticism for is the idea that a system of ropes (or whatever mechanical analogue you want to use for a brain) can be conscious. If it's functionally equivalent to a brain-equivalent system of neurons, how could it not be conscious? That a "rope brain" could be conscious is a radical idea that needs to be explored further.

 

[Belz...]

I'm not retracting anything.

I didn't ask for your opinion, but I'm willing to hear your plea.

You made a dismissive comment regarding philosophy

With very good reason, since philosophy focuses on convincing arguments rather than evidence.

AFTER someone you obviously respect posted an expert opinion from a philosopher (and continues to do so).

That might have something to do with the fact that I don't care about what that philosopher has to say, one way or another.

 

[Belz...]

It obviously doesn't satisfy the medical community. They don't use it. Nor have I ever seen it used in peer-reviewed journals.

So what have you seen in medical journals as a definition ?

SRIP, although a bit simplistic, corresponds to the definitions of consciousness I've heard throughout my life, so I'm willing to hear what alternatives we have. I mean precise alternatives, not vague ones.

No, there's no evidence that cars are conscious and plenty of contradictory evidence they aren't (they don't possess any known mechanism that would enable them to be conscious).

That all depends on the definition of consciousness. By Pixy's definition, a lot of things are conscious.

Well, we could create a whole new post on what "self" means.

Being a programmer, I can tell you that "self" is a pointer to an object, namely the one being aware. So when you say "aware of itself", I take it to mean that it's a feedback loop.

And how does SRIP "work"?

See above.

I might retract that one. Subjective experience itself is so poorly defined. Again, we know it when we feel it, but how do you put in words?

Indeed. I think that we are very much attached to what we feel is something very different from everything else, namely our experiences, and I think that this attachment is preventing us from defining and seeing it as just another phenomenon. When people talk about qualia, I immediately think about this. Qualia is a way of saying "yes, but it just feels more than just physical."

I don't think that "feeling" cold is more than just a convenient interpretation our brain has. We've wrapped it in more-or-less mystical language, but there is no reason to keep that baggage when studying consciousness objectively.

 

[Zeuzzz]

Thermostat = computer

Brain = thermostat

Brain = conscious

Thermostat = conscious

I find it hard to explain how a computer can will itself to heat up by the power of its conscious thoughts alone like the following person



The scientists seem pretty stumped by his readings when they assess him.

In January this year, Wim Hof ran a half Marathon (21 km) above the polar circle in Finland. He wore only a pair of shorts and no shoes. The ground (snow) temperature was 35 below. In a few months time, he'll try something similar on Everest's north side. The expedition, led by Dutch Werner de Jong, will try to set a new world record as Wim attempts to climb parts of Everest wearing only shorts.

This is not your regular stunt. Wim already has 9 world records, and has trained hard for many years to withstand cold, much like some monks do in Tibet.

Wim can actually regulate his core heat to control the temperature of his skin. Something of a medical enigma, Wim is able to withstand cold that could kill or seriously injure other people.

On Everest, Wim will also put to use his free style climbing skills - which he once demonstrated hanging between 2 hot air balloons by his middle finger at an altitude of 1500 meters. Once back inside the basket, Wim finished by climbing to the top of the balloon.

One of Wim's world record attempts took him to the North Pole, where he held his breath for about 6 minutes and 20 seconds below the ice.

 

[dlorde]

It obviously doesn't satisfy the medical community. They don't use it.

The medical community is (rightly) a conservative special-interest group. Their focus is on clinical health and well-being, not general operational definitions. For an example of their careful conservatism, check out the history and ongoing changes to the definition of clinical death.

... Nor have I ever seen it used in peer-reviewed journals.

Which ones have you consulted?

 

[dlorde]

... We don't have a careful and precise definition.

Whatever you feel about its applicability and the ground it covers, isn't SRIP careful and precise?

 

[Clive]

Being a programmer, I can tell you that "self" is a pointer to an object, namely the one being aware. So when you say "aware of itself", I take it to mean that it's a feedback loop.

See below for some code that was posted in another thread long, long ago. At that time Rocketdodger agreed this was an example of "SRIP".

// srip.cpp

// Self-Referential Information Processing (based on rocketdodger's code), 
// and therefore (according to rocketdodger) this will be conscious 
// when compiled and executed. in other words, a working implemention
// of SRIP.

class A {
   A &someReference;
   bool iMayBe;

public:
   A(): someReference(*this) {}
 
   void think () 
   {
      if( &someReference == this ) 
         iMayBe = true;  // this is apparently self referential behavior
      else 
         iMayBe = false; // while this is not
   }
};

int main()
{
   A i;
   i.think();
   return 0;
}

Do you agree the program above is a full implementation of "SRIP", and that therefore, according to Pixy and those who agree with his "definition", must be "conscious" when run?

Is this what you think being "self-aware" is?

Remember that when complied and actually running on a physical machine, there are no variable names to add implied meaning to anything, and any "pointer" will just be a value stored in some register.

 

[Clive]

simple, direct factual questions.

You appear to have missed my "simple, direct factual questions" (post 2532)?

 

[dlorde]

... I spent some time on Google, and I can't find any medical research articles that define consciousness as SRIP. Now, you can complain that I'm being overly restrictive by focusing on medical research (or psychological research), but what is the only thing we're sure produces consciousness? Brains.

I don't think medical research is the source of choice for reliable information on consciousness; I'd be more likely to agree if you went with brain dysfunction, or disorders of consciousness; like it or not, medical research tends to focus on dysfunction.

Subjective experience itself is so poorly defined. Again, we know it when we feel it, but how do you put in words?

Perhaps try describing what you experience, and/or invoking the sensations in others; isn't that what literature, art, poetry, lyrics, & music attempt to do?

For what it's worth, consciousness seems to me a compound or aggregate concept, a conceptual wrapper for a host of processes whose filtered output feeds a reflexive interpreter or narrative generator, entrained to the focus of attention, and based on a flexible model of observed/experienced behaviour.

In my view it exists as the interaction between that reflexive narrative of causal interpretation or explanation of the activities of those of sub-processes, and the sub-processes themselves. The interpreter or narrative generator originates in the predictive modelling of external behaviours that, at advanced levels, also gives rise to 'theory of mind' necessary for sophisticated social interaction. That modelling facility is used to model the self as well as others, and the causal narrative interpretation and prediction of behaviour, when applied to the modelling entity in particular, becomes the conscious self image.

So one might expect high-level co-operative predators to have particularly high levels of consciousness, as they need the flexibility and capacity to model the behaviour of both fellow hunters and prey, with the selective advantage of co-ordination of activities via communication driving greater intellectual complexity. One might expect competitive groups of highly social co-operative predators to have the most sophisticated levels of consciousness, driven by a competitive cycle of increasing complexity of social behaviours, requiring ever more sophisticated communication and predictive modelling, with explanatory/interpretive narrative generation contributing to both.

So my model of consciousness involves SRIP in a flexible predictive (and potentially explanatory) model of behaviour for both self and other, where the model for self involves at least a minimal reflexive narrative as the self image. This requirement for self-image takes my definition well above Pixy's basic SRIP definition, but fundamentally incorporates it.

 

[PixyMisa]

Remember that when complied and actually running on a physical machine, there are no variable names to add implied meaning to anything, and any "pointer" will just be a value stored in some register.

And everything reduces to atoms. So that fundamentally can't possibly matter. Whatever produces consciousness, it has to be a physical process, because there's nothing else to be. Therefore the effect can be reproduced by some other physical process.

There's no magic to meaning.

 

[PixyMisa]

Pixy, your claim is that consciousness is solely the result of computation of a particular type

Yes.

and also that the human brain is a Turing Equivalent computer.

No. The human brain is no more than Turing-complete, and its function can be reproduced by any Turing-complete system. But it can't be strictly viewed as Turing-complete because it's unreliable.

You claim that a Turing machine could (in principle) simulate all the computations of a particular biological brain (whilst also being provided all the appropriate "inputs"), and that the simulated brain will necessarily have *exactly* the same subjective experiences during this simulation as the matching real brain would have had given exactly the same "inputs".

Of course. It's impossible for it to be otherwise.

Now suppose we were actually able to do this - to simulate an entire brain along with all the relevant initial conditions and inputs from the surrounding environment for some finite period of time, let's same one minute.

Yes. (It's not physically possible even in principle to do this exactly, but leaving that aside for the moment.)

This simulation can be recorded as a sequence of integers - each integer encoding the state of the data tape and machine for each and every step of the Turing machine computation. In other words, if all your claims are correct, we can produce a finite sequence of integers that, in effect capture every detail of that simulation from start to finish.

Of course. This is possible for any finite physical system.

Now, note that for any finite sequence of integers {y_0, y_1, ... y_n}, there is an infinite set of polynomial functions P = {p_0, p_1, ...} that will all produce the same sequence of integers. That is to say, p(0) = y_0, p(1) = y_1 ... p = y_n, ...., for any p in P.

My question for you is, when we compute the sequence p(0), p(1), ... p using any particular p from the infinite set P, does that sequence of computations also necessarily produce exactly the same conscious experiences as the Turing Machine simulation?

It's conscious in the same way that you can have a conversation with a recording.

 

[Clive]

And everything reduces to atoms. So that fundamentally can't possibly matter. Whatever produces consciousness, it has to be a physical process, because there's nothing else to be. Therefore the effect can be reproduced by some other physical process.

There's no magic to meaning.

Do you agree with Rocketdodger that the program shown (when complied and run) demonstrates "SRIP"?

Your response above sounds to me like you're now getting a little wary of doing that and are now saying something like, "well, whatever, at the end of the day it's all just atoms".

Can you "reduce" the value of pi to atoms?
Does the value of pi "exist"?
Could it exist without atoms?

Is the "pattern" of prime/non prime as you traverse all integers more or less "fundamental" than atoms (or quarks or superstrings or whatever you wish to label as "fundamental" in the physical world)?

 

[Jeff Corey]

Sorry, but the lot of you are not making any sense and just asserting crap with no backup. Cars are not conscious. Fact. My laptop is not conscious and has all the social graces of an Aspergers person.
You will never get anywhere toward understanding these private events until you dump your pre-boxed assumptions.

 

[PixyMisa]

Do you agree with Rocketdodger that the program shown (when complied and run) demonstrates "SRIP"?

It's got self reference. But it doesn't do anything.

As I've said any number of times, self-reference alone isn't consciousness. Self-referential information processing is consciousness.

Your response above sounds to me like you're now getting a little wary of doing that and are now saying something like, "well, whatever, at the end of the day it's all just atoms".

Nope. I'm pointing out that no objection of the sort "it's just a value in a register" can have any relevance.

Can you "reduce" the value of pi to atoms?

See answer to second question.

Does the value of pi "exist"?

No. You can have an instance of that value, and that will reduce to atoms.

Could it exist without atoms?

Not realistically. You need some form of matter, and anything other than atoms is just impractical.

Is the "pattern" of prime/non prime as you traverse all integers more or less "fundamental" than atoms (or quarks or superstrings or whatever you wish to label as "fundamental" in the physical world)?

It doesn't exist in the physical world. "Fundamental" has no referent.

 

[PixyMisa]

Sorry, but the lot of you are not making any sense and just asserting crap with no backup.

Sorry, but no.

Consciousness is, before anything else, the ability to examine your own mental processes. To be able to respond to "A penny for your thoughts." in some form. Car ECUs can do this. Many computer programs, including some that are sure to be on your laptop, can do this.

You will never get anywhere toward understanding these private events until you dump your pre-boxed assumptions.

What assumptions?