it's tech

[Wrath of Dagon]

A machine can not be self-aware, thus no true AI.

[Humodour]

A machine can not be self-aware, thus no true AI.

 

lol. I guess I should expect something like that from you.

[Wrath of Dagon]

What do you do, Krezak, sit in some basement and read about other people's accomplishments on the internet all day?

[Humodour]

What do you do, Krezak, sit in some basement and read about other people's accomplishments on the internet all day?

 

a) I haven't even been on here in a week.

b) Yes, I read scientific journals. Does that offend you?

 

But let's get to the meat of your inanity:

 

A machine

 

Define 'machine'.

 

can not be self-aware

 

Define 'self-aware'.

 

thus no true AI.

 

Define 'true' in this context. From a technical perspective, 'true' AI is metaheuristic, while 'false' AI is symbolic. I fully concede symbolic AI (i.e. generally pre-90's AI) has no capacity for true learning.

 

I have one whooping big counter-example of a machine which is both self-aware and intelligent: humans. Other examples exist, though. Most of them in the mammal or dinosaur family (birds, specifically).

 

Oh, and hey - let's keep religion out of this OK? Or at least tell me if you're approaching this from a religious perspective so I can ignore you.

[Wrath of Dagon]

A human is not a machine. If you need me to define those terms for you, you need to study up on computer science, I'm not a tutor for whiny brats. And yes, I do approach this from a philosophical, you could say religious, perspective, so having no further discussions with you is an extra bonus.

[Humodour]

A human is not a machine. If you need me to define those terms for you, you need to study up on computer science, I'm not a tutor for whiny brats.

 

Oh, no. I understand what 'machine' means. But it's proper in a discussion to define what your terms are to avoid confusion. If you're unwilling to do that, I will:

 

Two significant definitions exist: a physical process, and a mathematical one.

 

Either way, you have no basis for your claims, since neural nets are rigorously defined mathematically (they are universal Turing machines), and have been proven to exist physically (in animal brains - both human and not).

 

You tell any half-decent biologist that a 'human is not a machine' and watch them laugh at you.

 

And yes, I do approach this from a philosophical, you could say religious, perspective, so having no further discussions with you is an extra bonus.

 

Ahhh. And I bet you believe evolution is just a 'theory' (like gravity?) too?

[Trenitay]

Evolution is a theory. Gravity is a law.

[Humodour]

Evolution is a theory. Gravity is a law.

 

Actually, gravity is a theory. A law is an idealised mathematical construct (also known as an axiom) which is always 100% correct in some platonic mathematical space. We cannot say this about gravity (see below), and in fact there has been at least two different laws of gravity so far (the first Newtonian, the second relativity). To call gravity as observed in physical reality a law is a silly as calling evolution a law because there exists a mathematically idealised version of evolution (see mathematical optimisation techniques).

 

There seems to be this fundamental misunderstanding among laypeople about what the term 'theory' means in science. It means a hypothesis which has been experimentally tested and never been falsified (i.e. always turns out correct).

 

In science, since we cannot with certainty say "this theory is 100% correct" as we do in mathematics, the most powerful assertion we can make is "this theory has been correct in every case tested".

 

Why, then, is science useful? Because our theories are based upon the 'idealised mathematical constructs' (a.k.a. laws) previously mentioned. When we discover these constructs to be an insufficient approximation of the real world (i.e. they are disproved for some specific case), we ditch them and formulate a new theory based on a new mathematical construct which accounts for the anomalous case. This is what has happened to both gravity and evolution in the past.

 

So yes, evolution is, like gravity, 'just a theory'.

 

Edit: if you want another take on the formulation of theories, wikipedia has a nice subsection on it: http://en.wikipedia.org/wiki/Theory#Assump...mulate_a_theory

Edited December 19, 2008 by Krezack

[Wrath of Dagon]

Either way, you have no basis for your claims, since neural nets are rigorously defined mathematically (they are universal Turing machines), and have been proven to exist physically (in animal brains - both human and not).

Neural networks in animal brains are very different from the phony little neural networks your computer simulates. And you can't get any more universal than a Turing machine as far as computers are concerned.

You tell any half-decent biologist that a 'human is not a machine' and watch them laugh at you.

A human is only a machine empirically.

 

Ahhh. And I bet you believe evolution is just a 'theory' (like gravity?) too?

You just said it is. Probably why it's called the Theory of Evolution. Anyway, I don't mix up philosophy and science.

[Trenitay]

Actually, gravity is a theory. A law is an idealised mathematical construct (also known as an axiom) which is always 100% correct in some platonic mathematical space.

 

Are you sure? I was led to believe that gravity was a law without a sufficient theory to explain it.

[taks]

gravity is a law, along with coulomb's law, ohm's law, etc. these laws are different than mathematical "laws" true, but laws nonetheless.

 

relativity isn't a law, either, it is a theory. these can be looked up anywhere on the web as well as in any textbook on the subject.

 

taks

[Walsingham]

With respect, sir, you appear to have become intoxicated with Newtonian hubris. Krezack is correct. There is no such thing as a 'law' of gravity. There is a force which we refer to as gravity, which operates under observation according to the theory of gravity. If there were a recognisable law of gravity we would have no need of the Hadron collider.

 

I agree wholeheartedly with Krezack when he says that there is a public misconception about what constitutes science. I think this stems from a natural human desire for certainty. But if there is anything fundamental to science it is the spirit of testing scientific theories, and by inference, asserting they are not true.

[taks]

With respect, sir, you appear to have become intoxicated with Newtonian hubris. Krezack is correct. There is no such thing as a 'law' of gravity.

sorry, you are 100% incorrect. the law of gravity is one of many laws of science. you can look the term up on wiki if you'd like (hard to really reference anything else), or search on the web, but taken from the wiki intro:

The laws of science are various established scientific laws, or physical laws as they are sometimes called, that are considered universal and invariable facts of the physical world. Laws of science may, however, be disproved if new facts or evidence contradicts them. A "law" differs from hypotheses, theories, postulates, principles, etc., in that a law is an analytic statement, usually with an empirically determined constant. A theory may contain a set of laws, or a theory may be implied from an empirically determined law.

 

I agree wholeheartedly with Krezack when he says that there is a public misconception about what constitutes science. I think this stems from a natural human desire for certainty. But if there is anything fundamental to science it is the spirit of testing scientific theories, and by inference, asserting they are not true.

i agree as well, but i also think that krezack does not understand it nearly as well as he professes, and hence, makes these ridiculous statements.

 

taks

 

edit: technically the "law of gravity" is really "newton's universal law of gravitation."

Edited December 19, 2008 by taks

[taks]

btw, mathematical axioms are not laws in the same sense as laws of science. they are self-evident and assumed as starting points, though not provable in and of themselves. contrary to axioms, scientific laws are typically empirically determined.

 

taks

[Humodour]

There's very little wrong with my explanation except that it was oversimplified for lay people, taks.

 

edit: technically the "law of gravity" is really "newton's universal law of gravitation."

 

Um. You are aware that Newton's law of universal gravitation was proven false by Einstein about 100 years ago, right? It is correct only in platonic Newtonian/Euclidean mechanics, not real life. It's still used for many Earth-based physical calculations today because it's a decent approximation to the much more complex and time-consuming equations of relativity (an approximation made possible because space-time is a manifold). But for precise calculations, the theory of relativity is used (i.e. space-time curvature due to mass is considered, or at least approximated by geodesics).

 

And we're not done revising the theory of gravity, yet. Relativity, stunning as it is, doesn't play well with quantum mechanics, so as Walsingham said, we'll hopefully learn something about gravity from the LHC.

[samm]

I haven't heard of Platon or Euclid in mechanics (except for the latter maybe: that non-parallel lines cross, two points can be connected via a line, and stuff like that, i.e. euclidean geometry, is significant as well), and that Newton's law applies to Newtonian physics is... well, it's redundant As for this:

Relativity, stunning as it is, doesn't play well with quantum mechanics, so as Walsingham said, we'll hopefully learn something about gravity from the LHC.

Indeed that would interest me as well, but the LHC won't be back up and running for a while. So until then I stay tuned for more news on the memristors and hope that the next posting in this thread will not disappoint me in that respect

And: I don't want to kill this thread so don't let this post derail it further...

[Wrath of Dagon]

Newton's law is correct for normal objects, because there's no observable difference between it and relativistic calculations. Newton never said his law applies to sub-atomic particles.

[samm]

Relativity is also very difficult to claim to be maintained at sub-atomic level You're mixing it up with quantum mechanics - or I'm mistaken of course.

[Wrath of Dagon]

Relativity is maintained at sub-atomic levels, and it's important since particles often travel at or near the speed of light. Newtonian laws don't apply at all, that was my point.

[Walsingham]

I was thinking about this faster than light business.

 

The statement that you can never accelerate beyond the speed of light seems fundamentally bogus, as a naive science cretin. This statement implies that speed is solely correlated with acceleration, but I can't think of any example of a pure 1 to 1 correlation. This in turn implies that other factors have to be at work. if that is the case, what is to stop us leveraging those other factors to produce superior speed?

 

Or have i simply filled my synapses with christmas pudding?

[Gorth]

It does indeed seem bogus. The university of Vienna has been experimenting with Entangled Photons for years now, not only doing faster than light communication, but also some rather obscure practical applications. How about Quantum Encryption?

[Wrath of Dagon]

What? I never heard of anything travelling above the speed of light. Theoretically it's possible, but such an object could never go below the speed of light. According to Lorenz equations, it takes infinite energy to accelerate a finite rest mass to the speed of light, or by symmetry to decelerate. Now some day some one may prove that wrong, but so far as I know that hasn't happened yet.

[Gorth]

What? I never heard of anything travelling above the speed of light. Theoretically it's possible, but such an object could never go below the speed of light. According to Lorenz equations, it takes infinite energy to accelerate a finite rest mass to the speed of light, or by symmetry to decelerate. Now some day some one may prove that wrong, but so far as I know that hasn't happened yet.

The university built this big setup which they are expanding so they can test it's practicality on larger and larger distances. Experiments with faster than light communication and teleportation across the Danube: http://www.univie.ac.at/qfp/publications3/...les/2006-10.pdf

 

They hope to be able to get their own satellite in space one day for testing their communication technology

[taks]

Um. You are aware that Newton's law of universal gravitation was proven false by Einstein about 100 years ago, right? It is correct only in platonic Newtonian/Euclidean mechanics, not real life.

no, einstein showed that gravity does not work as newton calculated at all scales. he did not prove it wrong. it is an approximation, but all laws are approximations.

 

do a little research, krezack. you are demonstrating your ignorance again. this is basic science learned in the US at the age of 12 or 13, and any high school physics class should have gone through why it is a law and not just theory. people seem to confuse what "theory" or "law" means for whatever reason, perhaps because they don't have any formal training, i don't know.

 

"just a theory" doesn't mean something is weaker than "a physical law." generally it simply stays "just a theory" because it is too broad to be a law. the universal law of gravitation can be tested directly. relativity, OTOH, cannot, at least, not all of it. that said, calling evolution "just a theory" doesn't in any way decrease its importance. too many just don't get it.

 

taks

Edited December 29, 2008 by taks

[taks]

So until then I stay tuned for more news on the memristors and hope that the next posting in this thread will not disappoint me in that respect

the latest edition of the IEEE Spectrum has another article on the concept but i have not read it yet and, unfortunately, i cannot provide a link to it nor am i allowed to paste it anywhere since it is copyrighted material (requiring membership to access). i'm assuming it is more of the same, btw, since the original was through the IEEE anyway. i doubt we'll hear much more till someone comes up with a bona-fide application.

 

oh, and i will state again, since apparently someone is incapable of bringing up a simple wiki page, physical laws of science are not even similar to mathematical axioms, and they are indeed laws (please, try to tell me ohm's law is not a law, please). they are natural properties that are assumed true, and generally can be tested directly, e.g., gravity. as noted by wrath, all einstein did was show that gravity does not work in relativistic scales, but it works well at normal scales.

 

taks

Edited December 29, 2008 by taks