Artificial Intelligence/Conciousness

[bombus]

A thought experiment:

 

Charles Babbage is thought of as the inventor of computers with his difference engine etc. This was a mechanical device.

 

Modern computers are electronic rather than mechanical and work by having electrons flow through CPUs and other electronic devices.

 

Is it theoretically possible to replicate the workings of a computer using mechanics?

[insane_alien]

yes, it is theoretically possible to build a mechanical computer. it is also practically possible as evidenced by the difference engine.

 

in theory, if you had enough time you could do everything a modern computer can do on babbages difference engine.

[bombus]

yes, it is theoretically possible to build a mechanical computer. it is also practically possible as evidenced by the difference engine.

 

in theory, if you had enough time you could do everything a modern computer can do on babbages difference engine.

 

In that case, if it is possible for computers to become conscious (and many think they can), could a purely mechanical engine of enough complexity become conscious?

Edited June 21, 2009 by bombus

[insane_alien]

if an elctronic computer can do it, then a mechanical computer can do it. however, with a mechanicla computer complexity will need to be greater to achieve the same result and it won't be capable of being as fast.

[ydoaPs]

if an elctronic computer can do it, then a mechanical computer can do it. however, with a mechanicla computer complexity will need to be greater to achieve the same result and it won't be capable of being as fast.

 

That's what I was thinking: it'd be incredibly slow like the Chinese Room. And, since it's a mechanical device, it would be more prone to failure.

[bombus]

That's what I was thinking: it'd be incredibly slow like the Chinese Room. And, since it's a mechanical device, it would be more prone to failure.

 

It's hard for me to accept that a purely mechanical device could become conscious.

[insane_alien]

how hard it is to accept has no bearing on its reality however.

[ydoaPs]

It's hard for me to accept that a purely mechanical device could become conscious.

 

Yet you can accept that a purely chemical one can?

[bombus]

Yet you can accept that a purely chemical one can?

 

Maybe... but due to interactions at the sub-atomic level.

 

I suspect it's actually something to do with quantum mechanics. I think quantum computers could become sentient, but not the ones we use today. (I know QM is involved with 'normal' cpu's but I mean qbits rather than bits)

 

---

Merged post follows:

Consecutive posts merged

how hard it is to accept has no bearing on its reality however.

 

Very true. a current theme I keep noticing recently is how much scientific investigation is influenced by incredulity - including my own. I have been reading a brief history of time again, and research on black holes was delayed because certain influential scientists didn't like the idea and encouraged others not to bother with research. Trendy ideas get funding for research, unfashionable or unpopular ideas do not. Scientific study is often not that scientific. It is a little worrying.

Edited June 21, 2009 by bombus
Consecutive posts merged.

[ydoaPs]

Maybe... but due to interactions at the sub-atomic level.

 

I suspect it's actually something to do with quantum mechanics. I think quantum computers could become sentient, but not the ones we use today. (I know QM is involved with 'normal' cpu's but I mean qbits rather than bits)

 

But the structures in the brain are classical objects. What does QM have to do with it?

[bombus]

But the structures in the brain are classical objects. What does QM have to do with it?

 

Structures in the brain are classical objects, but thinking is done by chucking electrons around the brain. One could say that the brain is a structure to contain and 'herd' electrons, i.e., it is the electrons that are the important bit rather than the meat itself. I suspect that Quantum effects come into play. Some proposals have been put forward, but this is debated, and Max Tegmark thinks the brain too warm and wet:

 

If our neurons have anything at all to do with our thinking, if all these electrical firings correspond in any way to our thought patterns, we are not quantum computers," says Tegmark. The problem is that the matter inside our skulls is warm and ever-changing on an atomic scale, an environment that dooms any nascent quantum computation before it can affect our thought patterns. For quantum effects to become important, the brain would have to be a tiny fraction of a degree above absolute zero.

 

Hameroff is unconvinced. "It's obvious that thermal decoherence is going to be a problem, but I think biology has ways around it," he says. "Water molecules in the brain tissue, for instance, might keep tubulin coherent by shielding the microtubules from their environment. In back-of-the-envelope calculations, I made up those 13 orders of magnitude pretty easily."

 

Also, superconductors are functioning at ever higher temperatures, so nature could have got there already via millions of years of evolution. We can but wait and see.

Edited June 22, 2009 by bombus

[iNow]

Structures in the brain are classical objects, but thinking is done by chucking electrons around the brain. One could say that the brain is a structure to contain and 'herd' electrons, i.e., it is the electrons that are the important bit rather than the meat itself.

That's ridiculous. Ever heard of a neurotransmitter?

 

I'd also appreciate it if you would share a source for that text which you copy/pasted. Not only is it the rules, not only is it just good personal policy, but I'd like to read the context in which the comments were made so as to better understand them.

 

A sentence suggesting that some guy Hameroff is not convinced isn't good enough.

Finally, those superconductors operating at higher temperatures that you mention... those are made out of materials which simply aren't present in the brain.

[bombus]

That's ridiculous. Ever heard of a neurotransmitter?

 

No it's not, and Yes.

 

Neurotransmitters are just one part of brain functioning.

 

I'd also appreciate it if you would share a source for that text which you copy/pasted. Not only is it the rules, not only is it just good personal policy, but I'd like to read the context in which the comments were made so as to better understand them.

 

sure It was only an article I found somewhere, not a scientific paper.

 

http://www.scribd.com/doc/14429946/Problem-with-Quantum-Mind-Theory

 

A sentence suggesting that some guy Hameroff is not convinced isn't good enough.

 

Isn't good enough for what?

 

Finally, those superconductors operating at higher temperatures that you mention... those are made out of materials which simply aren't present in the brain.

 

No, but the brain has far more complex structures than superconducting material. I am just pointing out that temperature may not be such a factor as previously thought.

[insane_alien]

bombus, the brain is not a superconductor. it has resistance, magnetic fields can pass through it un impeded. it can barely be called a good conductor never mind a super conductor.

 

also, regarding structures in the brain, there is nothing special at the scales considered for superconductivity. pretty much indistinguishable from your skin really.

 

also, superconductors tend to be highly crystaline.

[GutZ]

You have to remember that just because you deal with elementary electrical particles doesn't mean that the quantum effects that go on down there affect what's happening at our level.

 

If that were the case quantum theory would be a normal notion.

 

It's probably a good idea too with all EM waves going around and sub atomic particles. Who knows what havoc that would bring if our brains were quantum controlled.

[bombus]

the points you raise are difficult to answer (certainly for me) but our understanding of how the brain works is still unknown at the deepest level - I doubt it's just down to neurones firing - so I'll keep my mind open on brains being quantum systems.

 

However, the idea that a purely mechanical system could become conscious really intrigues me. How presposterous could we make the 'computer'? How about one made of wood and bricks, powered by steam with water doing the same job as electricity?

[Mokele]

No, we understand perfectly how the brain works at the deepest level - they're called neurons, and there's no deeper level.

 

Have you ever seen a neuron? They've *VERY* large cells. Individual axons of the largest of them can be seen with the naked eye, and can be over 3 feet long. They're composed of literally *billions* of molecules. Transmission of a nerve impulse involves fluxes of literally tens of thousands of ions. No electrons are involved - just whole sodium and potassium ions, moving in large numbers. Transmission between cells is due to the release of thousands of neurotransmitter molecules. Any quantum effect would be rapidly averaged out over the sheer number of molecules involved.

 

Of course, that ignores the real issue - evidence. We can model neurobiology without going any deeper than the neuron, and can do so accurately. What actual evidence is there for any quantum effects?

[insane_alien]

yes, we could do that.

 

if you really wanted you could breed a few octillion slaves and give them all abacus's and use messenger pidgeons as the data bus. of course this over looks the many practical problems but in principle it is possible.

 

using a modern computer system would be much much more efficient.

 

so I'll keep my mind open on brains being quantum systems

 

you should never be closing your mind either way.

 

i think you are assuming the word quantum means any number of wacky things can occur. it is not equivalent to magic

[bombus]

No, we understand perfectly how the brain works at the deepest level - they're called neurons, and there's no deeper level.

 

You are very wrong there. However, if you do know, maybe you should write a scientific paper on it. I'm sure the science world would love to know that you know exactly all there is to know about brain function.

 

Paramecium function, including hunting out prey, without a single neuron, as do other single celled creatures. Neurones are not therefore the only things that can give an organism 'intent'.

 

Try here for some ideas:

 

http://www.informaworld.com/smpp/content~content=a788080291~db=all

---

Merged post follows:

Consecutive posts merged

yes, we could do that.

 

if you really wanted you could breed a few octillion slaves and give them all abacus's and use messenger pidgeons as the data bus. of course this over looks the many practical problems but in principle it is possible.

 

Using sentient slaves might be seen as a flaw!

 

 

i think you are assuming the word quantum means any number of wacky things can occur. it is not equivalent to magic

 

No, its just that quantum computers can do many more calculations simultaneously - even solving problems without fully running programs!

 

Using an optical-based quantum computer, a research team led by physicist Paul Kwiat has presented the first demonstration of "counterfactual computation," inferring information about an answer, even though the computer did not run. The researchers report their work in the Feb. 23 issue of Nature.

 

"In a sense, it is the possibility that the algorithm could run which prevents the algorithm from running," Kwiat said. "That is at the heart of quantum interrogation schemes, and to my mind, quantum mechanics doesn't get any more mysterious than this."

 

See here:http://www.physorg.com/news11087.html

[insane_alien]

No, its just that quantum computers can do many more calculations simultaneously - even solving problems without fully running programs!

 

this is not how quantum computers are special, they are special in that the can run algorithms that are impossible to run on standard computers.

 

the fact that they can do parallel operations is nothing fancy because computers have been capable of doing this since the 50's.

 

and i never said anything about the slaves being sentient, they could be well trained dogs for all it matters, as long as the rules are followed.

[bombus]

this is not how quantum computers are special, they are special in that the can run algorithms that are impossible to run on standard computers.

 

Yes. Fair enough, but that's how it's often described:

 

Quantum computers have the potential for solving certain types of problems much faster than classical computers.

 

Anyway, I suspect quantum computers will not be hindered by Gödel's theorem. I have just started reading Shadows of the mind, so will get back to you about that...

[Mokele]

You are very wrong there. However, if you do know, maybe you should write a scientific paper on it. I'm sure the science world would love to know that you know exactly all there is to know about brain function.

 

Paramecium function, including hunting out prey, without a single neuron, as do other single celled creatures. Neurones are not therefore the only things that can give an organism 'intent'.

 

When did I claim to know everything? Or that neurons were necessary for complex behavior?

 

Here's two claims for you:

 

1) I know much, much more about neurobiology than you do.

 

2) There is no empirical evidence for quantum effects in the brain.

[mooeypoo]

This is getting tedious.

 

bombus, there is a reason why we keep insisting on avoiding logical fallacies. It's not just our anal attempt to be "strict and logical", it's a way of making sure the argument remains consistent: That we argue on the same subject, the same point, and answer one another with valid answers. If we don't do these, the entire argument is moot.

 

bombus, you really should start paying attention to the way you relate to people's comments. It might serve you well to read a bit about what logical fallacies are, and why they are important to recognize. I don't really care if you can recognize them by name, but you should be fair enough with us (and your own argument, quite frankly) to try and argue valid points, rather than argue just for the sake of remaining correct and hope people missed the flimsy logic that was used.

 

Look here, for instance:

No, we understand perfectly how the brain works at the deepest level - they're called neurons, and there's no deeper level.

 

You are very wrong there. However, if you do know, maybe you should write a scientific paper on it. I'm sure the science world would love to know that you know exactly all there is to know about brain function.

Mokele was refering to the physical level of our understanding, in which case he is far from being wrong. By changing the meaning of "deepest level" from physical level to metaphorical level, you presented him as wrong. That's called a straw man. It's a logical fallacy because it makes the comment absolutely moot. Mokele did not say what you claim he said. Your point only seems true because you misrepresent Mokele's counterpoint.

 

If you would have answered what he actually said, rather than what you want to argue, it would have been a valid point to make. Try again.

yes, we could do that.

 

if you really wanted you could breed a few octillion slaves and give them all abacus's and use messenger pidgeons as the data bus. of course this over looks the many practical problems but in principle it is possible.

Using sentient slaves might be seen as a flaw!

That might be, bombus, but that wasn't the purpose of this argument. insane_alien MENTIONED the point you made, when he said "of course this over looks the many practical problems". It was also a semi-non-related cynical addition, because the subject wasn't (yet) the morality or practicality of actually getting robots to work on something as sentient being, the point is that robots are able to.

 

What you did in this point is shift the point of the argument and introduce a red herring. Those are two logical fallacies, and they don't make you right.

 

i think you are assuming the word quantum means any number of wacky things can occur. it is not equivalent to magic

No, its just that quantum computers can do many more calculations simultaneously - even solving problems without fully running programs!

This was a shifty little bump on your part, bombus. Here's insane_alien's FULL quote on the matter of quantum computers:

 

yes, we could do that.

 

if you really wanted you could breed a few octillion slaves and give them all abacus's and use messenger pidgeons as the data bus. of course this over looks the many practical problems but in principle it is possible.

 

using a modern computer system would be much much more efficient.

 

so I'll keep my mind open on brains being quantum systems

 

you should never be closing your mind either way.

 

i think you are assuming the word quantum means any number of wacky things can occur. it is not equivalent to magic

 

So, when you look at the full answer in context, you see that what insane_alien answered was your point about brains being quantum systems. I am not sure who said that exactly (that seems to be another strawman, though a subtle one), but regardless, insane_alien's response was directly related to your comment.

 

Your answer doesn't stand against insane_alien's answer. His point was that "quantum" doesn't mean magic, and so it's not all that insane to compare a brain to it. Your quote:

 

Using an optical-based quantum computer, a research team led by physicist Paul Kwiat has presented the first demonstration of "counterfactual computation," inferring information about an answer, even though the computer did not run. The researchers report their work in the Feb. 23 issue of Nature.

 

"In a sense, it is the possibility that the algorithm could run which prevents the algorithm from running," Kwiat said. "That is at the heart of quantum interrogation schemes, and to my mind, quantum mechanics doesn't get any more mysterious than this."

 

See here:http://www.physorg.com/news11087.html

 

.. brings an exception - a "weird thing that happened once" - with quantum computers, presents it as a usual occurance with quantum computrs (is it? I don't think so) and it seems you use this to show that quantum computers are, therefore, not like the brain. Your conclusion doesn't follow that logic: For one, you need to demonstrate what EXACTLY it means by "the computer didn't run". Was there electricity through it, but just not any software? Well, if that's the case, isn't that *exactly* how our brain works? You don't have any software, you have electrical signals going back and forth between the neurons -- which is what has been told in the thread.

 

So essentially, you've presented a case that doesn't necessarily come against the claims put forth to you, but presented it as if the obvious conclusion is that the other claims are wrong. The logic doesn't not follow the conclusion, bombus.

 

So, to summarize - everyone will benefit greatly if we all avoid logical fallacies. Not just for the rules of the forum (which state, quite clearly, against their use) but for us to have a *VALID* discussion, one we can all mutually learn and understand from.

 

 

 

~moo

Edited June 25, 2009 by mooeypoo
gah, removing extra line breaks..

[Sisyphus]

 

Using sentient slaves might be seen as a flaw!

 

Not if they're just following a set of rules.

[bombus]

This is getting tedious.

 

bombus, there is a reason why we keep insisting on avoiding logical fallacies. It's not just our anal attempt to be "strict and logical", it's a way of making sure the argument remains consistent: That we argue on the same subject, the same point, and answer one another with valid answers. If we don't do these, the entire argument is moot.

 

Well maybe you should pull up others as well for a change. e.g Mokele's claim that he knows more about brains than I. As far as I am aware that kinda talk is not allowed on this forum.

 

bombus, you really should start paying attention to the way you relate to people's comments. It might serve you well to read a bit about what logical fallacies are, and why they are important to recognize. I don't really care if you can recognize them by name, but you should be fair enough with us (and your own argument, quite frankly) to try and argue valid points, rather than argue just for the sake of remaining correct and hope people missed the flimsy logic that was used.

 

Look here, for instance:

 

Mokele was refering to the physical level of our understanding, in which case he is far from being wrong. By changing the meaning of "deepest level" from physical level to metaphorical level, you presented him as wrong. That's called a straw man. It's a logical fallacy because it makes the comment absolutely moot. Mokele did not say what you claim he said. Your point only seems true because you misrepresent Mokele's counterpoint.

 

I did not change the meaning of deepest level. Mokele states that there IS NO DEEPER LEVEL. And then goes on to say he knows more about brain functioning than me...

 

If you would have answered what he actually said, rather than what you want to argue, it would have been a valid point to make. Try again.

 

Well, you make a good point there, fair enough. I shall try harder.

 

That might be, bombus, but that wasn't the purpose of this argument. insane_alien MENTIONED the point you made, when he said "of course this over looks the many practical problems". It was also a semi-non-related cynical addition, because the subject wasn't (yet) the morality or practicality of actually getting robots to work on something as sentient being, the point is that robots are able to.

 

What you did in this point is shift the point of the argument and introduce a red herring. Those are two logical fallacies, and they don't make you right.

 

I was actually making a light hearted joke (c/f the chinese box)

 

 

This was a shifty little bump on your part, bombus. Here's insane_alien's FULL quote on the matter of quantum computers:

 

What 'shifty little bump'? I was actually agreeing with him that my original phrasing was inaccurate and clarifying what I had meant earlier.

 

So, when you look at the full answer in context, you see that what insane_alien answered was your point about brains being quantum systems. I am not sure who said that exactly (that seems to be another strawman, though a subtle one), but regardless, insane_alien's response was directly related to your comment.

 

I don't think there is a problem with his answer nor my subsequent reply.

 

Your answer doesn't stand against insane_alien's answer. His point was that "quantum" doesn't mean magic, and so it's not all that insane to compare a brain to it. Your quote:

 

Again, I see no problem here. I am just pointing out that I know QM isn't magic - but it is wierd.

 

.. brings an exception - a "weird thing that happened once" - with quantum computers, presents it as a usual occurance with quantum computrs (is it? I don't think so) and it seems you use this to show that quantum computers are, therefore, not like the brain. Your conclusion doesn't follow that logic: For one, you need to demonstrate what EXACTLY it means by "the computer didn't run". Was there electricity through it, but just not any software? Well, if that's the case, isn't that *exactly* how our brain works? You don't have any software, you have electrical signals going back and forth between the neurons -- which is what has been told in the thread.

 

I have been studying QM for since 1991, and think you are wrong to dismiss this as a one-off. It's a development of QM and quantum computing, many years in the making.

 

So essentially, you've presented a case that doesn't necessarily come against the claims put forth to you, but presented it as if the obvious conclusion is that the other claims are wrong. The logic doesn't not follow the conclusion, bombus.

 

I don't think there were claims against me. Just comments. We're having a discussion.

 

So, to summarize - everyone will benefit greatly if we all avoid logical fallacies. Not just for the rules of the forum (which state, quite clearly, against their use) but for us to have a *VALID* discussion, one we can all mutually learn and understand from.

 

OK, but I think you're scolding the wrong person, or at least, I'm not the only offender.

 

I think maybe my reputation is too tarnished to carry on as I am being prejudged and misinterpreted. Maybe I should retire bombus and start afresh - but I'd rather not, I think I've been around for 4 years now...

 

 

 

~moo

---

Merged post follows:

Consecutive posts merged

Not if they're just following a set of rules.

 

I was only joking actually. The joys of text message communication:-)

Edited June 26, 2009 by bombus
Consecutive posts merged.