If a computer was powerful enough would it generate consciousness?

[littlebabyjesus]

^^ @ axon, Crispy. Thank you both.

Crispy - are you saying that one planck length can overlap with another? If so, wouldn't it be splitting that planck length into two shorter lengths, and isn't that impossible?

*head starts hurting*

 

[Crispy]

^^ @ axon, Crispy. Thank you both.

Crispy - are you saying that one planck length can overlap with another? If so, wouldn't it be splitting that planck length into two shorter lengths, and isn't that impossible?

*head starts hurting*

no, the planck "time" can overlap. I think.

 

[axon]

trouble is that in the real world, this happens asynchronously - but a computer considers a simulation in 'frames' where the state of all elements changes at the same time, every time.

Even if it is happenening asynchronously it's not like all the asynchronous bits can communicate directly with each other instantaneously.

Why couldn't you use 'time frames' for the brain? Each 'bit of neurone" is assigned a state at each time interval. Make the 'bit of neurone' small, and the time interval small and voila. Simply attach the legs and eyes and you have a human!

 

[axon]

Dunno about the 'overlap' of Planck times
But even if there were many different 'phases' of Planck time periods, for the events to be communicated across neurones would take the order of milliseconds so being out of synch by less than a Planck time period would presumably not matter.

ETA Unless of course all the events lost in the out of synch Planck time periods sum up to form a remainder term ...called Neo!!!!!!

 

[sleaterkinney]

God, slap a 'theory' on the forum and they get all arsey if you dare disagree. Have i pissed you off by disagreeing with you and pointing out that you seem to disagree with own posts? Let the intellectual deference begin. Bring your awards and certificates to the inspection committee before you post.

That could read like a critiscm of the p+p boards as well. There's a reason why I don't get involved in arguments about 1920s Russia in there....

 

[Crispy]

Even if it is happenening asynchronously it's not like all the asynchronous bits can communicate directly with each other instantaneously.

Why couldn't you use 'time frames' for the brain? Each 'bit of neurone" is assigned a state at each time interval. Make the 'bit of neurone' small, and the time interval small and voila. Simply attach the legs and eyes and you have a human!

Thing is, we have no idea how accurate the simulation needs to be.

 

[Fruitloop]

Or even if it needs to be accurate at all.

 

[littlebabyjesus]

ETA Unless of course all the events lost in the out of synch Planck time periods sum up to form a remainder term ...called Neo!!!!!!

Bugger, I was with you till then. Wouldn't all events be minutely out of synch with each other?

*Throws question open to panel like Melvin Bragg on In Our Time*

 

[bluestreak]

so in conclusion, no-one knows if it would work or not, or why.

 

[butchersapron]

so in conclusion, no-one knows if it would work or not, or why.

Yep

 

[littlebabyjesus]

so in conclusion, no-one knows if it would work or not, or why.

Oi, you, no-one knew that a photon acted like a wave and a particle at the same time until the experiment was done a few years ago. But physicists were pretty confident that it did and formulated a whole system based on the assumption. Just 'cos we can't prove it doesn't mean we shouldn't discuss it.

 

[axon]

Thing is, we have no idea how accurate the simulation needs to be.

Well we can put some limits on it. Neurones fire in the millisecond range. Ion channels (which conduct nerve impulse current) take about a microsecond to open. I think it takes about a nanosecond for an ion to physically move through an ion channel. This seems a reasonable time resolution to aim for.

eta I reckon at the very very most you'd need 1/10 nanosecond resolution. So assuming a nerve impuse travels at 100m/s, and the average ion channel is about 20 nm across, so it would take 0.2 nanosecond for an action potential to pass completley over an ion channel.

 

[Bob_the_lost]

Wasn't there a simulation of a fraction of a mouse brain not too long ago? The simulation was of a quarter of a mouse brain's worth of neurons running at around 1/5th real time speed on a massive supercomputer cluster. Last i heard about that everyone was feeling rather smug about it.

So, it's only a matter of scale to get a model that behaves (mostly) like a human brain... Of course that doesn't mean the neurons would magically start working to create a fully functional human level mind.

 

[Crispy]

Wasn't there a simulation of a fraction of a mouse brain not too long ago? The simulation was of a quarter of a mouse brain's worth of neurons running at around 1/5th real time speed on a massive supercomputer cluster. Last i heard about that everyone was feeling rather smug about it.

So, it's only a matter of scale to get a model that behaves (mostly) like a human brain... Of course that doesn't mean the neurons would magically start working to create a fully functional human level mind.

They were very simple neurons, IIRC - and there were no hormones and other non-neuron brainybits either.

 

[Bob_the_lost]

They were very simple neurons, IIRC - and there were no hormones and other non-neuron brainybits either.

Chuck in a decade each. Then another to get up to the scale where we can manage a human sized brain in close to real time. Three decades for the computational ability. Less if there was any real money behind it.

Then again how much of the brain is needed for something that will never walk? Chuck away a good portion of the neurons that (maybe) are irrelevant and you might be there in a decade or two...

 

[Crispy]

Chuck in a decade each. Then another to get up to the scale where we can manage a human sized brain in close to real time. Three decades for the computational ability. Less if there was any real money behind it.

Then again how much of the brain is needed for something that will never walk? Chuck away a good portion of the neurons that (maybe) are irrelevant and you might be there in a decade or two...

You got money you want to put on that?

http://www.longbets.org/

 

[Dr_Herbz]

Why is it that (some) humans believe that we're somehow special?

Are ants conscious or are they just organic machines, how are they different to... let's say... a venus fly trap?
How far removed are we from ants?

If ants are 'conscious' then there's absolutely no reason a computer cannot be 'programmed' to achieve a similar level of 'consciousness'.

1: Install the firmware (software enabling it to assimilate what its sensors transmit to the cpu)
2: Install a cut down version of windows 95a with no service packs installed (this would allow the computer to act in a similarly fucked up fashion to a human)
3: Install the software...

The software is the big problem but not unlike the chess computer, it is possible, you simply have to input enough scenarios and their outcome and you have a computer that for all intents and purposes, can think for itself.

The argument that a computer will never be able to make decisions for itself is fundamentally flawed. Does anyone actually believe that we make decisions based on 'feelings' or is it simply the case that we're making decisions based on past experiences etc?
We're all a product of our environment, there are still canibalistic tribes who believe it is right to eat people, they believe this because it is what they have been programmed to believe so how does this make them any different to a computer?

 

[Blagsta]

You're assuming that a computer and the human brain are directly analogous.

 

[Fruitloop]

You'd never get Windoze to run on an ant. Linux reduced kernel maybe.

 

[Dr_Herbz]

You're assuming that a computer and the human brain are directly analogous.

You're assuming that they're not...

The only difference I can see is that the hardware and software are in 1 tidy self contained unit.

If you remove a piece of someone's brain (see 'lobotomy'), they change, they're no longer the same person. Surely this pisses on the argument that it's our 'soul' (or whatever these God fearing people wish to call it) that makes us who we are?

 

[Bob_the_lost]

You're assuming that a computer and the human brain are directly analogous.

There is no science / evidence to help chose between the two theories ("spark" vs. "computer"). As such i chose the theory that i prefer and that best fits into my worldview. To sit there and say "you don't know that!" is correct but after a point, boring.

Well, slight mistruth, there are less intelligent animals that function without visible/measureable conciousness etc. All of which can be construed to support the computer arguement, but it's not exactly a proven point, just a probability.

 

[stdPikachu]

The argument that a computer will never be able to make decisions for itself is fundamentally flawed.

I don't recall anyone saying "never", but I could be wrong.

Does anyone actually believe that we make decisions based on 'feelings' or is it simply the case that we're making decisions based on past experiences etc?

Because we don't know how the brain works, we can't make a judgement on this.

We're all a product of our environment, there are still canibalistic tribes who believe it is right to eat people, they believe this because it is what they have been programmed to believe so how does this make them any different to a computer?

Because...

You're assuming that a computer and the human brain are directly analogous.

Current computer tech works in a fundamentally different way to how our current understanding of the brain works. It's highly likely that the actual structure of the brain is directly related to how it processes and stores information (heck, do we even know how information is represented inside the brain?), therefore for a computer to behave like the brain it would need to have, or at least accurately mimic, the same structure.

You also make a syntactical slip with the language there; organic brains are not "programmed" in the same sense as computers are. For a start, many organic organisms display innate knowldge from the day they're born, whereas current computer technology, once turned on, does precisely nothing until you issue some instructions to it. A computer program is fundamentally a huge list of checkboxes that may do the odd mathemetical calculation, but ultimately it can't do anything it wasn't programmed to do (except crash, perhaps ) whereas it's a fundamental function of the brain to alter its own structure based on both repetitious events and random stimuli that haven't been encountered before.

 

[Bob_the_lost]

You also make a syntactical slip with the language there; organic brains are not "programmed" in the same sense as computers are. For a start, many organic organisms display innate knowldge from the day they're born, whereas current computer technology, once turned on, does precisely nothing until you issue some instructions to it.

Not really. Think back to your days with the ZX81 or whatever that god awful thing was, if you gave it FF then it'd know to output the value in Register A (or something like that). If you press some buttons on a calculator it will do the calculation. You work with the high end, general purpose computer chip. The difference is in design, presumably a computer brain could be built with exactly the same self programing ability as the human one.

A computer program is fundamentally a huge list of checkboxes that may do the odd mathemetical calculation, but ultimately it can't do anything it wasn't programmed to do (except crash, perhaps ) whereas it's a fundamental function of the brain to alter its own structure based on both repetitious events and random stimuli that haven't been encountered before.

Code can though. Hell i've done that much and the AI course i've done was a joke.

The human brain isn't all powerful, it can only behave in certain ways, we haven't charted all of the possible behaviours (like the way it adapts to injury) but it doesn't meant that behaviour can't be incorperated into code (or even the hardware).

 

[Dr_Herbz]

You also make a syntactical slip with the language there; organic brains are not "programmed" in the same sense as computers are. For a start, many organic organisms display innate knowldge from the day they're born, whereas current computer technology, once turned on, does precisely nothing until you issue some instructions to it. A computer program is fundamentally a huge list of checkboxes that may do the odd mathemetical calculation, but ultimately it can't do anything it wasn't programmed to do (except crash, perhaps ) whereas it's a fundamental function of the brain to alter its own structure based on both repetitious events and random stimuli that haven't been encountered before.

Sorry, I should have added - Firmware == DNA

 

[butchersapron]

Why is it that (some) humans believe that we're somehow special?

Oh the layers!

 

[Jonti]

Even if it is happenening asynchronously it's not like all the asynchronous bits can communicate directly with each other instantaneously.

Ah, but you need to take entanglement into account. It's not communication that matters so much as co-ordination.

All the asynchronous bits co-ordinate with each other instantaneously.

 

[Blagsta]

There is no science / evidence to help chose between the two theories ("spark" vs. "computer"). As such i chose the theory that i prefer and that best fits into my worldview. To sit there and say "you don't know that!" is correct but after a point, boring.

Well, slight mistruth, there are less intelligent animals that function without visible/measureable conciousness etc. All of which can be construed to support the computer arguement, but it's not exactly a proven point, just a probability.

I haven't mentioned anything to do with "spark". What are you on about?

 

[Bob_the_lost]

I haven't mentioned anything to do with "spark". What are you on about?

Ok, i assumed you knew how modelling neurotransmitters in computers works and shoved you into the wrong "group".

Background/Random:You don't just use wires (or even valves or transistors) as a crude mockup. You use a fully funcitoning computer to run a simulation of each neuron. Based upon what we know about the brain and how it works it's a pretty good way to simulate things.

If you're one of the "but it might have other inputs we don't know about" camp then there's nothing to say we can't simulate that too. I merely assumed (wrongly) you were part of the "something special" group.

 

[axon]

Ah, but you need to take entanglement into account. It's not communication that matters so much as co-ordination.

All the asynchronous bits co-ordinate with each other instantaneously.

Not quite sure what "entanglement" you are talking about. I agree that all the asynchronous bits need to coordinate with each other instantaneously. I can't see why this couldn't be achieved using a computer (or two).
I'm going to the pub now.

 

[Kameron]

There is a text on this subject.

Gödel, Escher, Bach (commonly GEB) is a Pulitzer Prize-winning book by Douglas Hofstadter that discusses how animate beings can form from inanimate matter and what it means to have a 'self'.

It is a book that is really worth picking up, I've read it a dozen times, I think I get about 20% on a good day but it is endlessly fascinating whether you start from the AI point of view or from the question of how this bag of Carbon, Hydrogen and Oxygen that is me has the power to write this.