Explain consciousness to the layman.

[Beelzebuddy]

On the other hand, the coordination of the signature brain waves during consciousness, and the parallels of experience during the losing and gaining of consciousness, these are observable.

Sure. But correlation is not causation.

But you're right, correlation isn't causation. All we have at the moment is a correlation, but it's a damn exciting one.

But correlation is not causation. It doesn't matter how much you want the correlation to be causative, it most likely ain't. In this case we already have a dang good idea of what's causing the correlation: the same neural structures doing the thinking are making the waves. I can see how you might get turned around by the language used by some papers, but you need to read a little more critically to see that they're talking about the physical structures acting at those frequencies, not the electromagnetic waves their actions generate.

One last cite: if you hit the cortical inhibitory cells hard enough, the whole thing goes boinoioioing in the gamma band.

Computers that fly planes do so according to a real-time, interactive model, not the Turing computational model. A computer operating according to the Turing model would never be able to fly a plane, or play an MP3 file, or run a windowing system. This is the case no matter how powerful the computer. There has to be some kind of explicit temporal element - which is what Piggy has been describing in his post. It's this temporal element that the computationalists have consistently said is not needed - that it is possible to run the computation at one millionth of the speed and get exactly the same result. Clearly it's not possible to run the control program for a plane at a millionth of the speed and get the same results.

This does not, of course, imply that a real-time computer system is necessarily going to be able to produce consciousness. Just that a real-time computer system is the minimal requirement.

No there don't. An mp3 player running at one millionth speed still plays exactly the same music - just at one millionth speed. A plane-flying-program "flying" at one millionth speed in a simulation running at one millionth speed would fly perfectly well. Interfacing with the real world often imposes additional real-time constraints on particular implementations of an algorithm. That doesn't affect the algorithm itself.

Hence my signature. The field of computers is awash with language that's been lifted from the human world and applied to mechanical devices and mathematical abstractions. This has had the effect of anthropomorphising computers and their operations. As Dijkstra pointed out, this is a fundamentally childish and immature approach to a profession, and is guaranteed to warp thinking.

Dijkstra was a curmudgenly old coot, too greedy to take a random walk instead of the single shortest path.

 

[westprog]

No, it doesn't at ALL. It's like if I asked you how gravity works, and you suggest I drop an anvil on my toe and find out.

If you wanted to know "how it works" then you should have asked that. I would have replied "nobody knows, including you". Instead, you asked "what does it feel like" and I told you, in a way that you can very clearly find out.

Yes and you didn't understand the question, clearly.

I understood the question, if you mean what you wrote. Guessing at what you might have meant, obviously was a failure.

That's a cop out. You didn't give me a clear definition because you can't, not because your answer was better.

I just said that I can't. Nor can anyone else.

There's a confused idea about definitions that seems logical but actually isn't. A definition of a word is a list of words, describing the word. Put all the definitions together, and you have a dictionary.

How do you read a dictionary? Suppose that we were to find the ruins of an alien civilisation and we saw an alien dictionary. Written in recognisable symbols, on some kind of paper analogue - but with no other information to decode it. Of what use would it be? Could it be?

We would see the lists of symbols and try to interpret the relationships between them - but even the relationships themselves would be symbols. We could never possibly interpret the language.

However, if we were to find a surviving alien, he could, for example, point to the alien word for "pain", and slam his hand in a drawer. Given that, we could make a start at decoding the language. We would have meaning.

Our own dictionaries, our own definitions are no different. We have to import meaning from outside. The definitions by themselves are dead. We can't even specify the relationships between the symbols, because the relationships themselves have to be specified as symbols.

So when we find that something in the dictionary cannot be simply expressed by other words in the dictionary, we shouldn't be upset or worried. It's an essential element of the dictionary that such words exist. Otherwise, the dictionary would be entirely useless. It's the conduit via which meaning becomes attached to language.

 

[Beelzebuddy]

However, if we were to find a surviving alien, he could, for example, point to the alien word for "pain", and slam his hand in a drawer. Given that, we could make a start at decoding the language. We would have meaning.

Except that was the alien word for "sex," that wasn't his hand he slammed in the drawer, they just like it freaky back on Orgasmulon VI.

 

[westprog]

But correlation is not causation. It doesn't matter how much you want the correlation to be causative, it most likely ain't. In this case we already have a dang good idea of what's causing the correlation: the same neural structures doing the thinking are making the waves. I can see how you might get turned around by the language used by some papers, but you need to read a little more critically to see that they're talking about the physical structures acting at those frequencies, not the electromagnetic waves their actions generate.

One last cite: if you hit the cortical inhibitory cells hard enough, the whole thing goes boinoioioing in the gamma band.


No there don't. An mp3 player running at one millionth speed still plays exactly the same music - just at one millionth speed.

Do you really think that playing at one millionth the speed is doing "exactly the same"?

A plane-flying-program "flying" at one millionth speed in a simulation running at one millionth speed would fly perfectly well. Interfacing with the real world often imposes additional real-time constraints on particular implementations of an algorithm. That doesn't affect the algorithm itself.

No, it wouldn't fly at all. It would sit in a computer and do generate heat. Whether or not the "algorithm" would work or not is not the point. The program would lack the functionality necessary to perform the activity.

Programs that simulate something and programs that control something operate in a different fashion. Adding time constraints to a program make it a very different type of computation. The algorithm needs to be specified and implemented in a different way.

It would be possible to write a complete simulation of flight in PASCAL*, for example. It wouldn't be possible to write the control software for an aircraft in PASCAL. It doesn't have the necessary language features. Some kind of synchronisation mechanism needs to be part of the algorithm.

Dijkstra was a curmudgenly old coot, too greedy to take a random walk instead of the single shortest path.

 

[westprog]

You are equating "need not" and "can not".

If the computer that doesn't and cannot control the body is entirely equivalent to the one that can, as the computationalists claim, then the fact that some other computer can is irrelevant. We've already established that the ability to control the body is not important.

If you accept the computational view, then it inherently follows that you don't view the capability of replacing a brain and operating a human body as being significant functionality.

 

[Beelzebuddy]

Whether or not the "algorithm" would work or not is not the point.

It's precisely the point.

The rest of your post consists of obtuse misinterpretation.

 

[westprog]

You are of course entitled to use any strange and idiosyncratic meaning for an arbitrary label.

As are you.

 

[westprog]

Science doesn't know so it must be magic brain waves.

I suppose it was inevitable that the computational approach would end up describing the operation of the brain as magic.

 

[westprog]

How large would the algorithm representing all of the information (within its context) from this post be?

The representation of the post itself would probably be a few thousand bytes. To represent the semantics, and all the knowledge referenced, and the way that that knowledge interacts? I doubt we have the capacity.

 

[westprog]

It's precisely the point.

The rest of your post consists of obtuse misinterpretation.

Since I refuse to accept that the "plane" in a simulated world is a real plane actually flying, if I could only see it in the right way, I suppose one of us is irredeemably obtuse.

 

[Beelzebuddy]

Since I refuse to accept that the "plane" in a simulated world is a real plane actually flying, if I could only see it in the right way, I suppose one of us is irredeemably obtuse.

Keep knocking down that strawman.

 

[westprog]

Keep knocking down that strawman.

If you persist in standing him up. Meanwhile, my substantive point - that the program which actually controls an aircraft is fundamentally different to the computation which simulates it - remains unaddressed. Just keep telling yourself that westprog is the one who doesn't get it.

 

[Leumas]

Do you really think that playing at one millionth the speed is doing "exactly the same"?

No, it wouldn't fly at all. It would sit in a computer and do generate heat. Whether or not the "algorithm" would work or not is not the point. The program would lack the functionality necessary to perform the activity.

Programs that simulate something and programs that control something operate in a different fashion. Adding time constraints to a program make it a very different type of computation. The algorithm needs to be specified and implemented in a different way.

It would be possible to write a complete simulation of flight in PASCAL*, for example. It wouldn't be possible to write the control software for an aircraft in PASCAL. It doesn't have the necessary language features. Some kind of synchronisation mechanism needs to be part of the algorithm.

A computer control algorithm that controls actual dynamic systems can be EXTREMELY sensitive as to the TIME CONSTANTS of the system.

For example the sampling speed for an Analog to Digital Converter (ADC) is of CRITICAL importance due to something called the Nyquist Limit. The digitizing resolutions of the ADC and DAC as well as their time responses can drastically affect the control parameters of a control system.

A control system that cannot match the NATURAL time constants of the dynamic physical system it is controlling will never work.

That is why most REAL TIME DSP has to be performed by specialized chips that can perform the Transfer Functions and do all the DSP required within the real time constraints required by the characteristics of the real physical system.

For example, the Inertial Navigation System (INS) that controls the ailerons and rudder and elevators of an airplane needs to measure the current Yaw, Pitch and Roll attitude parameters from all the gyroscopes and accelerometers then perform all the necessary filtering to filter out noise and then perform the necessary integration to calculate the deviations from the required set point and then calculate all the required positions of the control surfaces so as to effect any required changes to restore the attitude to the set point according to a Proportional, Integral and Differential (PID) control system whose parameters are finely tuned according to the Transfer Function required to prevent undue oscillation or lag or overshoot or latent errors.

If all this is not done within a very definite time interval (usually in microseconds if not nanoseconds) the control will fail abysmally. Even the slightest delay in all that would change the characteristics of the whole system not just the program since the program is an integrated part of the DYNAMIC system.

Think about it in more layman terms...... for instance the plane has just pitched down a little and the input required to return it to the set point is say 0.1 degrees up in the elevator angle. Now if by the time the elevator angle change is done the plane has already pitched back up due to a gust or something then the 0.1 degrees change will make the plane pitch up even further which will require a correction of say 0.2 degrees down this time on the elevator but that again came too late then the airplane will be continuously pitching up and down with an oscillation that might in fact increase in amplitude until the plane breaks apart.

So the speed with which the computer has to complete the calculations is CRITICAL and has to be IN TUNE with the SHORTEST NATURAL TIME CONSTANT of the harmonics of the system.

If the computer program is not able to perform the calculations in real time it would NEVER be able to properly control the plane.

However..... in a simulation of the system we can make all this take place in SIMULATED time rather than real time and the actual speed of the calculations then do not matter...... but only during R&D.

In a real simulator that actually has to EMULATE how the plane behaves for the purposes of training, the whole thing has to be in real time again otherwise the pilots would not get any real feel for the plane's flying characteristics.


So the speed of the computer is of crucial importance for a computer controlling any dynamic and real process.

 

[yy2bggggs]

Forget simulations. The claim that consciousness requires a certain rate simply violates relativity.

 

[Leumas]

Keep knocking down that strawman.

If you persist in standing him up. Meanwhile, my substantive point - that the program which actually controls an aircraft is fundamentally different to the computation which simulates it - remains unaddressed. Just keep telling yourself that westprog is the one who doesn't get it.

 

[Beelzebuddy]

If you persist in standing him up. Meanwhile, my substantive point - that the program which actually controls an aircraft is fundamentally different to the computation which simulates it - remains unaddressed. Just keep telling yourself that westprog is the one who doesn't get it.

You're begging the question. The algorithm with controls the aircraft is unchanged in either case. Even its implementation need not change. The only difference is its API connects to a sufficiently high-fidelity simulation instead of the real deal.

 

[westprog]

Forget simulations. The claim that consciousness requires a certain rate simply violates relativity.

I've seen this particular claim before, and I find it quite staggering. If relativity means that there are no time constraints on the operation of the brain, then we can extend the principle to the whole of engineering. Don't worry about the timing on your V-8 - relativity, you know.

 

[Belz...]

If you wanted to know "how it works" then you should have asked that. I would have replied "nobody knows, including you". Instead, you asked "what does it feel like" and I told you, in a way that you can very clearly find out.



I understood the question, if you mean what you wrote. Guessing at what you might have meant, obviously was a failure.




I just said that I can't. Nor can anyone else.

There's a confused idea about definitions that seems logical but actually isn't. A definition of a word is a list of words, describing the word. Put all the definitions together, and you have a dictionary.

How do you read a dictionary? Suppose that we were to find the ruins of an alien civilisation and we saw an alien dictionary. Written in recognisable symbols, on some kind of paper analogue - but with no other information to decode it. Of what use would it be? Could it be?

We would see the lists of symbols and try to interpret the relationships between them - but even the relationships themselves would be symbols. We could never possibly interpret the language.

However, if we were to find a surviving alien, he could, for example, point to the alien word for "pain", and slam his hand in a drawer. Given that, we could make a start at decoding the language. We would have meaning.

Our own dictionaries, our own definitions are no different. We have to import meaning from outside. The definitions by themselves are dead. We can't even specify the relationships between the symbols, because the relationships themselves have to be specified as symbols.

So when we find that something in the dictionary cannot be simply expressed by other words in the dictionary, we shouldn't be upset or worried. It's an essential element of the dictionary that such words exist. Otherwise, the dictionary would be entirely useless. It's the conduit via which meaning becomes attached to language.

Let's get back to the question. I don't know why we're getting so sidetracked:

How does it feel like to have an experience ? Can you put that into writing or not ?

 

[Piggy]

You are misunderstanding the use of the term “Critical Mass”.

Actually, it's that misunderstanding that I was objecting to. Sorry I wasn't clear about that.

I haven't seen anyone make the claim on this thread, but in others I've heard folks claim that there is some threshold of neural activity which, by sheer volume, reaches a critical mass which causes consciousness to emerge.

That idea, of course, has been disproven, since we know consciousness is not correlated with the sheer volume of neural activity in the brain.

 

[westprog]

You're begging the question. The algorithm with controls the aircraft is unchanged in either case. Even its implementation need not change. The only difference is its API connects to a sufficiently high-fidelity simulation instead of the real deal.

That is simply wrong, as anyone who's written such software knows. You cannot use the same program to control the plane as you use in the simulation.

It's possible to write a control program for the plane and attach simulation software to it - which is of course a very different thing. You are using simulation software as a framework to test - in real time - the control software. This is not the same thing as a system which simulates the plane and control software together - which can be done in a PASCAL-like program without time constraints. Such a simulation, being simpler, can be a very helpful first step in designing a control or monitoring program.

It's always possible to produce a simulation which runs in real time - but such a simulation will not be computationally equivalent to one that runs slower. One will work, the other won't.