The Hard Problem of Gravity

[yy2bggggs]

I spent over an hour figuring out how I might represent the concept of a "switch" using first order logic. I had to consult my old A.I. textbooks to brush up on the issue of knowledge representation before I even started, and I program A.I. for a living.

Out of curiosity... could you qualify this? Do you work in research, or industry? And what sort of applications?

 

[rocketdodger]

Rocks obviously switch on a constant basis. That's how the Grand Canyon came to be where it is. One drop of water can be sent one way or another.

Perhaps.

Yet, to say this, you were forced to devise entirely new and arbitrary definitions of the ON state for various entities. The rocks, the water, the wind, the sun, whatever.

If a rock channels water south, is it ON or OFF? If the water comes from rain, is it ON or OFF? If the sun is blocked by clouds, if the wind is warm or cold, are these things ON or OFF?

Like I said, you can come up with definitions to suite your argument. That is fine with me, who cares.

Unfortunately for you, nobody else would know wtf you are talking about unless you explained it to them in great detail.

With a thermostat circuit ... not so. ON means current is flowing in one direction above a given threshold, OFF means it is not. Very simple -- and understood by almost everyone.

You are kind of like Nick227, I think. Neither of you seem to realize that the entire point of language is to allow intelligent agents to communicate with each other and that the more efficient and expressive a language the easier it is for them to do so.

 

[RandFan]

This only matters when trying to assign some kind of mystical significance to The Great Switch, the holy grail of consciousness, the single quantum of thought and awareness.

Akin to searching for the mystical significance of feathers, wings, air pressure and momentum as they relate to flight.

There is no mystical significance to switches anymore than there is to the constituent elements of aerodynamic systems. That's just a strawman.

 

[rocketdodger]

Out of curiosity... could you qualify this? Do you work in research, or industry? And what sort of applications?

Industry. Hierarchichal finite state machine based AI for AAA games.

My crowning achievement thus far will be in full pubic view by mid summer this year. In fact you can already find them in trailer videos. I can't really go into much detail about it -- yet -- but one thing I can tell you is that as far as I know it will be the first AI in any game that, like a human player, is aware of and makes purposeful use of ranged weapon splash damage (on walls, ceilings, around corners, etc).

 

[Gate2501]

Industry. Hierarchichal finite state machine based AI for AAA games.

My crowning achievement thus far will be in full pubic view by mid summer this year. In fact you can already find them in trailer videos. I can't really go into much detail about it -- yet -- but one thing I can tell you is that as far as I know it will be the first AI in any game that, like a human player, is aware of and makes purposeful use of ranged weapon splash damage.

At least tell us the genre/platform! Pleeeeeeeeeease!

Edit: actually you gave enough clues as to the genre(in your edit).

 

[PixyMisa]

Rocks obviously switch on a constant basis.

So show me a rock switching.

That's how the Grand Canyon came to be where it is.

No, nothing of the sort.

One drop of water can be sent one way or another.

Completely wrong. Water is not quantised in "drops", nor is its direction quantised in "one way or another".

You can build a water switch out of rocks. That does not mean that rocks switch. It means you can use rocks to build a switch.

The reason that people don't analyse the switching capabilities of rocks is because they are entirely useless to us.

No, completely wrong. Rocks do not switch.

There's no point in figuring out the billions of different ways they switch water flow, or air flow, or elecrical current, or other physical events. But obviously they do.

No, wrong again. You don't appear to grasp what a switch is. A switch is a discontinuous function. At its simplest, it's a binary function.

Rocks. Don't. Do. That.

This only matters when trying to assign some kind of mystical significance to The Great Switch, the holy grail of consciousness, the single quantum of thought and awareness.

You think you're mocking us, but this is almost correct.

A binary switch is the smallest possible unit of information processing, and from that we can build a system that can compute any computable function, simulate any physical process.

That's what a computer is. That's what a brain is.

However, thought and awareness being processes and not states, the applicable quanta is not a switch but the act of switching.

 

[AkuManiMani]

Perhaps.

Yet, to say this, you were forced to devise entirely new and arbitrary definitions of the ON state for various entities. The rocks, the water, the wind, the sun, whatever.

If a rock channels water south, is it ON or OFF? If the water comes from rain, is it ON or OFF? If the sun is blocked by clouds, if the wind is warm or cold, are these things ON or OFF?

Like I said, you can come up with definitions to suite your argument. That is fine with me, who cares.

Unfortunately for you, nobody else would know wtf you are talking about unless you explained it to them in great detail.

With a thermostat circuit ... not so. ON means current is flowing in one direction above a given threshold, OFF means it is not. Very simple -- and understood by almost everyone.

Any object, or system of objects can, in principle, have their gross and fine characteristics represented in some binary language. The universe is composed of an infinitude of binary relations that can be expressed in such a way. Even so, the nature and significance of each of those relations are identical.

I think that the problem arises when individuals confuse the descriptive language for the actual object(s) of that description. Simply writing out a formal description of an entity, whether in text or machine code, is not the same as producing the thing IAOI -- especially when such a description is incomplete or flawed.

Even in the instance of phenomena that are relatively well understood [like gravity] a formal simulation of said phenomenon is not an example of the real thing. The only way to reproduce an example of a TIOAI would be to use one's formal knowledge to physically produce it and not just virtually simulate it.

In the case of consciousness, there is a severe paucity of formal description, and those that are provided are largely ad hoc shots in the dark. How can one seriously, and with a strait face, claim that they have reproduced it from such a flimsy basis?

You are kind of like Nick227, I think. Neither of you seem to realize that the entire point of language is to allow intelligent agents to communicate with each other and that the more efficient and expressive a language the easier it is for them to do so.

I'm pretty sure Nick227 and westprog realize the role and purpose of language.

What is at issue here is the lake of adequate formal language to not only describe consciousness, but to actually reproduce it. If it took you so much time and effort just to describe something as rudimentary as on/off, what in blue blazes makes you think that you have a sufficient definition of the very basis of your conscious experience? It is a problem that vastly dwarfs merely describing on/off switches -- which is why its has been historically called the hard problem.

 

[yy2bggggs]

That's what a computer is.

What about analog computers?

 

[AkuManiMani]

What about analog computers?

I find it tragically humorous that he can so strongly assert the universality of information processing yet deny it's reality in certain objects.

 

[AkuManiMani]

Industry. Hierarchichal finite state machine based AI for AAA games.

My crowning achievement thus far will be in full pubic view by mid summer this year. In fact you can already find them in trailer videos. I can't really go into much detail about it -- yet -- but one thing I can tell you is that as far as I know it will be the first AI in any game that, like a human player, is aware of and makes purposeful use of ranged weapon splash damage (on walls, ceilings, around corners, etc).

Regardless of our disagreement on what constitutes being 'aware' that sounds like some pretty cool ◊◊◊◊ right there

 

[Gate2501]

I find it tragically humorous that he can so strongly assert the universality of information processing yet deny it's reality in certain objects.

I don't think that anyone in this thread would deny that information processing is occurring in an analog computer.

 

[AkuManiMani]

I don't think that anyone in this thread would deny that information processing is occurring in an analog computer.

Oh, I don't just mean in analogue computers. Hes explicitly denied that it goes on in objects like rocks or atoms. Once again , hes completely and unequivocally wrong.

 

[Gate2501]

Oh, I don't just mean in analogue computers. Hes explicitly denied that it goes on in objects like rocks or atoms. Once again , hes completely and unequivocally wrong.

Well, I would consider information processing to be any detectable/observable changing of information. Like if a rock changed temperature, there would be thermal data pertaining to that rock, which was processed by the material that the rock is made of.

So yeah, speaking for myself, I agree with you that rocks and atoms can, and do, process information.

Does anyone here disagree with what I just said?

 

[PixyMisa]

What about analog computers?

What about analog computers?

 

[PixyMisa]

Well, I would consider information processing to be any detectable/observable changing of information. Like if a rock changed temperature, there would be thermal data pertaining to that rock, which was processed by the material that the rock is made of.

Information processing isn't just a change of information, it's a series of transactions involving changes in information.

So yeah, speaking for myself, I agree with you that rocks and atoms can, and do, process information.

Nope.

Informational transactions take place between the atoms in a rock. This does not constitute information processing, though, because there is no structure; the interactions are random. As I pointed out to AkuManiMani earlier, the result is the same as we'd get from simple thermal models, because that is what it is.

Does anyone here disagree with what I just said?

Kind of, yeah.

 

[PixyMisa]

Oh, I don't just mean in analogue computers. Hes explicitly denied that it goes on in objects like rocks or atoms. Once again , hes completely and unequivocally wrong.

You could at least attempt to show this.

Like, say, control your central heating with a rock. Or an atom. Either one is fine.

 

[Gate2501]

Nope.

Informational transactions take place between the atoms in a rock. This does not constitute information processing, though, because there is no structure; the interactions are random. As I pointed out to AkuManiMani earlier, the result is the same as we'd get from simple thermal models, because that is what it is.

I agree with you and RD that rocks are not made of on/off switches, and rocks are not switch*ing*, but I think that by the standard definition of "information processing", any observable state change on the part of the rock counts.

I would think that random processes still result in processed information, and still would count as information processing? Right?

 

[AkuManiMani]

You could at least attempt to show this.

Like, say, control your central heating with a rock. Or an atom. Either one is fine.

So central heating systems are not made of atoms? The internal dynamics of atoms and the particles that form them do not process and covey information according to logical physical laws? This is news to me.

 

[Gate2501]

So central heating systems are not made of atoms? The internal dynamics of atoms and the particles that form them do not process and covey information according to logical physical laws? This is news to me.

No, I think they are trying to express that you would need to make a switch(or series of switches) out of those atoms, in order to control the system.

 

[PixyMisa]

I agree with you and RD that rocks are not made of on/off switches, and rocks are not switch*ing*, but I think that by the standard definition of "information processing", any observable state change on the part of the rock counts.

I would think that random processes still result in processed information, and still would count as information processing? Right?

I think that by the usual physical definition, no, it doesn't count, but I'd have to check.