On Consciousness

[lenny]

That's true of everything in the Universe, and indeed, of the Universe as a whole.

Digital computers can realise only a very very small fraction of the dynamics we understand;mathematically they only map one big integer into another big integer.

In short: they fall short well before normal "stuff" falls short.

In short, you have to actually demonstrate this, not simply assert it.

hi PixyMisa,

do which bit did i not say clearly enough?

1) digital computers, per se, have digital memory: a finite number of bits. thus "one big integer"

coded in this integer are instructions that lead the machine to change its state to another big integer.

as a finite state machine, it will eventually visit a previous state, and thereafter go about in cycles.

2) digital arithmetic is unable to capture a great deal of mathematics. there is a million dollar prize for showing out there just for showing it is possible (or proving it is no) for infinitesimally small perturbations in partial differential equations to blow up in finite time: that would be much worse that "chaos"! and impossible to accurately simulate an a digital machine.

http://www.claymath.org/millennium/Navier-Stokes_Equations/

3) "chaos" (of the ed lorenz variety) cannot be accurately simulated on a digital machine either. i think there was a thread on this some years ago.


was that the kind of demonstration you were asking for? if not i am happy to try again.

 

[Mr. Scott]

hi PixyMisa,

do which bit did i not say clearly enough?

1) digital computers, per se, have digital memory: a finite number of bits. thus "one big integer"

coded in this integer are instructions that lead the machine to change its state to another big integer.

as a finite state machine, it will eventually visit a previous state, and thereafter go about in cycles.

2) digital arithmetic is unable to capture a great deal of mathematics. there is a million dollar prize for showing out there just for showing it is possible (or proving it is no) for infinitesimally small perturbations in partial differential equations to blow up in finite time: that would be much worse that "chaos"! and impossible to accurately simulate an a digital machine.

http://www.claymath.org/millennium/Navier-Stokes_Equations/

3) "chaos" (of the ed lorenz variety) cannot be accurately simulated on a digital machine either. i think there was a thread on this some years ago.


was that the kind of demonstration you were asking for? if not i am happy to try again.

Why do you think that kind of accuracy is required for a machine to be conscious? Neuron activity and the brain are extremely messy. I'd think the right overall structure would be sufficient to simulate. Why is there need for precision? Lack of precision is a hallmark of biological intelligence.

 

[!Kaggen]

Why do you think that kind of accuracy is required for a machine to be conscious? Neuron activity and the brain are extremely messy. I'd think the right overall structure would be sufficient to simulate. Why is there need for precision? Lack of precision is a hallmark of biological intelligence.

Says the "biologically intelligent"

 

[steenkh]

1) digital computers, per se, have digital memory: a finite number of bits. thus "one big integer"

coded in this integer are instructions that lead the machine to change its state to another big integer.

as a finite state machine, it will eventually visit a previous state, and theretafter go about in cycles.

The same goes for conscious brains. They consist of a finite amount of atoms, each having a finite amount of states. Given enough time, brains will also go in cycles.

Looking at these debates it seems that we need not wait too long to observe such cycles!

 

[PixyMisa]

hi PixyMisa,

do which bit did i not say clearly enough?

You were perfectly clear, you were just wrong.

1) digital computers, per se, have digital memory: a finite number of bits. thus "one big integer"

So does everything else.

coded in this integer are instructions that lead the machine to change its state to another big integer.

as a finite state machine, it will eventually visit a previous state, and thereafter go about in cycles.

Let's take as our example a tiny embedded computer, running at 1MHz, with just 256 bytes of memory. The time to cycle through all possible states would be about 3669229891921952094695762193851494025314662226076779097252566228 years.

2) digital arithmetic is unable to capture a great deal of mathematics. there is a million dollar prize for showing out there just for showing it is possible (or proving it is no) for infinitesimally small perturbations in partial differential equations to blow up in finite time: that would be much worse that "chaos"! and impossible to accurately simulate an a digital machine.

There's no such thing as a an infinitesimally small perturbation in reality. Space and energy are effectively quantised.

And of course, digital computers can perform symbolic mathematics.

3) "chaos" (of the ed lorenz variety) cannot be accurately simulated on a digital machine either.

Chaos can be simulated just fine on a digital machine. You must mean something else, but I don't know what.

was that the kind of demonstration you were asking for?

No, I was looking for something relevant.

 

[PixyMisa]

Why do you think that kind of accuracy is required for a machine to be conscious? Neuron activity and the brain are extremely messy. I'd think the right overall structure would be sufficient to simulate. Why is there need for precision? Lack of precision is a hallmark of biological intelligence.

Indeed. We know that consciousness continues largely unaffected across a wide range of operating conditions, as long as you don't cause damage to the brain itself.

 

[PixyMisa]

I should have specified a machine built by humans.

I was hoping you'd be specific about the machine -- who built it, where it is, and how they determined it had subjective experiences.

1. If we ask them, how do we know they answer honestly?

Ask more questions.

2. If we open them up and look, how do we recognize it?

Functionally. Does it provide a rote answer to the question, does it perform a simple surface computation, or does it really perform deeply introspective cogitation? You can tell by inspection.

3. If we design them specifically for this, how do we know we succeeded?

They work.

4. We are machines with subjective experiences. (I know. I was asking about machines we've made.)
[/LIST]

Yes, but I wasn't, because the distinction is misleading.

 

[Zeuzzz]

Damn.

dead for more than a decade really is down hill.

When people die, they tend to lose a lot of credibility, in my book.
The more dead they are, the less credible.

He got brain cancer and a few weeks to live in 2000, yea sucked.

Some of the things he said sometimes did verge on credible to crapola (depends on who his company was) I was not claiming to argue on his behalf.

Just the words and diction this man used were incredible. True wordsmith. He said things like Epigentics decades before science had even verified that gene expression changes outside your genome can happen, the sort of speaker that combines and makes up words as he goes along yet you know intuitively what they mean anyway.

Which is half the reason I have brought this up, not to defend him in particular but to address Dlordes comments about language, creativity, music and syntax having no relationship to ancient psychedelic use. His life is a polar opposite case.

 

[Mr. Scott]

Says the "biologically intelligent"

Would you like make a positive contribution to the discussion?

Precision is a discipline. It does not come naturally. We a born naturally to be rather imprecise, sloppy if you will.

I think creativity needs imprecision. I've observed that when I compose music, I try random things, flail around, make mistakes, change this note and that to see how it sounds, etc. The variations that sound good (hit the target emotions) I keep, those that sound worse I discard (reminds me of evolution) until a new piece is completed. This type of messy imprecision is critical to forming something new (e.g. evolution).

We build machines to be precise and accurate, but it's rather easy to add slop so they behave more like we do. A creative machine would do random things, then judge what appeared for better or worse by whatever criteria is needed, until it can't do better. This is not hard to program in a machine. I'm sure we could find someone had done it with a brief Google search.

 

[Mr. Scott]

Ask more questions.


Functionally. Does it provide a rote answer to the question, does it perform a simple surface computation, or does it really perform deeply introspective cogitation? You can tell by inspection.


They work.


Yes, but I wasn't, because the distinction is misleading.

Exactly how is the distinction between our naturally evolved brains, and machines we build to do what our brains do, misleading?

 

[PixyMisa]

Exactly how is the distinction between our naturally evolved brains, and machines we build to do what our brains do, misleading?

Because people might think it's relevant, and that's impossible. A machine is a machine; it does what it does.

 

[Zeuzzz]

Because people might think it's relevant, and that's impossible. A machine is a machine; it does what it does.

Our brains are not logic number crunching machines.

However, yes, they will continue doing what they do despite this fact.

Most people if given the chance to live like they are a machine would; as this option is the easy one, it takes less effort and conscious thought.

 

[quarky]

Our brains are not logic number crunching machines.

However, yes, they will continue doing what they do despite this fact.

Most people if given the chance to live like they are a machine would; as this option is the easy one, it takes less effort and conscious thought.

Funny, seconds before I read your post, there was a news clip on the radio about a new tread-mill desk. There's something iconicly ironic about a treadmill.

 

[Mr. Scott]

Because people might think it's relevant, and that's impossible. A machine is a machine; it does what it does.

You're right, but I'd still like to know how the internal subjective experience arises in machines.

 

[quarky]

um, it doesn't?

 

[Zeuzzz]

You're right, but I'd still like to know how the internal subjective experience arises in machines.

Me too. To watch this I wonder if could use it computer consciousness and ask the 'conscious' robot to imagine me hovering a $500,000 diamond between my fingers. Some people cant' some people just can, apparently.

But also be great what colors the sense notes in syntax and language end up in a robot. Even bizarre neural working may come into play, so A minor tastes of custard to them, synesthesia.

Or would this machine be malfunctioning in this consideration? Just like that other one that gets conscious experience from a constant flux of neurochemistry it? That would surely the closest we could get, all the way down to the exact molecular binding profiles for each. Yet the claim here that robot that users brain chemistry too like our is more likely to malfunction when normal food and traditional tinctures with consciousness clouding effect are added in.

Can we please say they are no more malfunctioning, than not perceiving things as the average person does.

If you maintain malfunctioning you need to broader your definition of it in humans, and in AI.

 

[Zeuzzz]

Because people might think it's relevant, and that's impossible. A machine is a machine; it does what it does.

Yet is still does not rival human consciousness very well, even if at some specialized mental challenging aspects I can.

Brought a present before and the chess setting was impossible to beat level 8 or over. But I them considered if you asked it what looked the hottest, the King or the Queen, and explain why, it would had not no human conscious reply and totally misunderstand it.

Would need to have a warning button for when it starts behaving too like a number crunching logic machine to tell a joke, do something spontaneous like playing prank to remind it's self of it's own free will and bigger role than just in a dead End end job being nothing more than a calculator.

 

[Aepervius]

hi PixyMisa,

do which bit did i not say clearly enough?

1) digital computers, per se, have digital memory: a finite number of bits. thus "one big integer"

coded in this integer are instructions that lead the machine to change its state to another big integer.

as a finite state machine, it will eventually visit a previous state, and thereafter go about in cycles.

Whereas it is true that a computer datastructure will have a finite number of possibility, like say 4 byte will only have a ~ 4.2 billions its representation allow for a quasi infinite numbers. For example you can decide to use a fixed byte precision, or even a floating byte precision, 8 byte of exposant, the rest for a number between 0 and 1, and tada, you can even represent reals beyond what a human can even dream to think of.
Sure you do not have infinite precision, but jsut by deciding to go from a representation to another you map a lot more than 4 billion possibilities.

Furthermore the neuron are not fully analogue. At some level they will be working by quanta, be it 1 neurotransmitter molecule or whatnot.

2) digital arithmetic is unable to capture a great deal of mathematics. there is a million dollar prize for showing out there just for showing it is possible (or proving it is no) for infinitesimally small perturbations in partial differential equations to blow up in finite time: that would be much worse that "chaos"! and impossible to accurately simulate an a digital machine.

http://www.claymath.org/millennium/Navier-Stokes_Equations/

Congratulation you proved computer are not able to go thru infinite precision.

But neither is our brain. We have a maximum precision in our vision field, in our hearing, and others field.

Soooo. what ?

3) "chaos" (of the ed lorenz variety) cannot be accurately simulated on a digital machine either. i think there was a thread on this some years ago.


was that the kind of demonstration you were asking for? if not i am happy to try again.

Fancy that chaos cannot be simulated on a machine. I guess I will have to call my pre-master discussion back and tell them it was all a lie.

 

[Mr. Scott]

Congratulation you proved computer are not able to go thru infinite precision.

But neither is our brain. We have a maximum precision in our vision field, in our hearing, and others field.

Soooo. what ?

Yea, I'm still waiting for someone to explain why infinite precision is needed for consciousness.

 

[Zeuzzz]

Infinity is just a symbol on a piece of paper. Nothing in the universe has ever been measured or proved to have an infinite anything.

In real experimental science the only real use it has is maybe extrapolating curves data or numbers to a similar standard.