Is the brain a computer?

[INTJMom]

I disagreed.

Yes you did.

 

[redacted]

I disagreed.

if that's true, then something i stated was unclear. i agree with your posts on this thread.

 

[redacted]

We already have people who fail the Turing test, while programs pass it.

yeah, that's why the turing test is pointless, IMO.

the claim goes like this:
let P = objectpassesturingtest
let Q = objectcanthink
P=>Q

if P is false, the statement says nothing about Q.
if Q is true, the statement says nothing about P.
only if P is true does the statement say anything at all...

edit: just reread what you said. computers pass the turing test? which ones? how many questions were they asked? what kind of questions were they asked?

i'd guess that if a program passed the test, the questions were extremely narrow. my understanding of the test is that if an object can fool a human into believing it thinks, then it DOES think.

seems ridiculously unlikely that i could have a half hour conversation with a computer and not notice something strange. basically impossible.

 

[Grayscale]

the concept of computers was created by the human brain, so no. it would be more accurate to say that the computer is a simpler, static, electric version of the brain

 

[Nocapszy]

True, but take one more step and all the atomic constituents are the same.



Even if that was the case, they would still operate differently. Would you agree?

What's different about them? I'm just curious what research you've done.

Natural selection operates on everything, but in different ways. In humans, the most adaptable human survives. The most adaptable human can market himself to mates, survive competition, and replicate. In robots, you would see the same thing: best funding, best marketing, best design would lead to the most popular model. But in humans, the creative force is mutation and imperfect replication.

Until we perfect genetic engineering... it's really only a matter of time -- unless the sun explodes much soon than we think.

In robots, the creative force is human ingenuity.

Is it so hard to believe that if there were self replicating devices that there wouldn't be mutation, and the device most fit to continue replication would do so? It's my opinion that you're faith in context is too great, and too little on the nature of the systems.

So, the question is, does the way humans design and model thinking for AI really imitate the way people think? Or does it only model the way we think we think?

As poignant as that sounds, I'm going to have to call bullshit on that. The way we think is the way we think. We've implanted the same calculative systems as are found in conscious thought. Think about it for a second -- could we program them if we didn't think the same way they do? Could we even build them, and then construct the proper binary sequences to make the aggregations of blips on the screen convey what they do without first being intimately familiar with, not only how they do work, but also how we wanted them to work?

 

[Nocapszy]

the concept of computers was created by the human brain, so no. it would be more accurate to say that the computer is a simpler, static, electric version of the brain

There's nothing static about them. Well... I guess you could argue that a capacitor is static, but I really prefer not to.

There's much greater reliance on semiconductors anyway, but it would be nothing short of fallacy to say that capacitors aren't necessary, at least for the current incarnations.

Do you advocate the notion that because it's more simple, it works differently?

 

[Carebear]

Think about it for a second -- could we program them if we didn't think the same way they do? Could we even build them, and then construct the proper binary sequences to make the aggregations of blips on the screen convey what they do without first being intimately familiar with, not only how they do work, but also how we wanted them to work?

We can emulate the way they think and they can emulate the way we think, but the two ways are different. Even if we can both emulate the other way of thinking, humans do binary less efficiently than computers and computers do "organic thinking" less efficiently than humans. I guess at some point computers could be made to emulate "organic thinking" so well it would be impossible to tell the difference, but it'd still be an emulation unless it was done using neurobiological circuits etc (at which point they'd cease to be digital computers).

Whether or not it matters that it's only an emulation when we can't tell the difference any more is another matter. (Seems that's the question they tried addressing in the movie AI.)

 

[nightning]

As poignant as that sounds, I'm going to have to call bullshit on that. The way we think is the way we think. We've implanted the same calculative systems as are found in conscious thought. Think about it for a second -- could we program them if we didn't think the same way they do? Could we even build them, and then construct the proper binary sequences to make the aggregations of blips on the screen convey what they do without first being intimately familiar with, not only how they do work, but also how we wanted them to work?

Are both lines of thinking not equally bullshitting?

True, the way we think is the way we think, however we can only program something based on our understanding of how we think. How we wanted them to work is reflective of our thinking... but it's not exactly the same as our thinking. What I'm trying to say is that we really don't know how we think, we just think. And in thinking, we believe we have some inferences of how we do that. That's all.

 

[Nocapszy]

We can emulate the way they think and they can emulate the way we think, but the two ways are different. Even if we can both emulate the other way of thinking, humans do binary less efficiently than computers and computers do "organic thinking" less efficiently than humans.

Er... the brain does operate on a binary electrical system... We don't consciously know of every neuron firing, or not firing, (that would require us to have a series of neurons to be aware of each neuron Ha!) but neither does a computer keep track of every bit. It all culminates into what we speak, or a computer prints text, etc.

 

[ThatsWhatHeSaid]

What's different about them? I'm just curious what research you've done.

The following paragraph in my previous post was an attempt to explain some of those differences. In terms of research -- probably less than you have.

As poignant as that sounds, I'm going to have to call bullshit on that. The way we think is the way we think. We've implanted the same calculative systems as are found in conscious thought.

This assumes that 1) We adequately understand how conscious thought works enough to simulate it. I disagree.
2) That once we figure out how it works, we are in fact able to simulate it accurately. My point in my previous post speaks to this problem. Is thinking about thinking the same as thinking? You say yes, I say no. If what's being modeled is the thought process itself (in this case, deduction) then yes. But in AI, what's being copied isn't the deductive process (which is always thinking) but the results of those deductions.

Lets say people think like x. Someone who doesn't know much about psychology might propose that people think like y. He figures that out by using x, because, well, he's thinking. But what does he use to build the AI? If he uses y, then he's off. That's thinking about thinking, and there's no guarantee that he's right. If he uses x -- which in this case would be his deductive reasoning -- then he's right. At some point, with research and testing, x and y would converge and he might actually be able to simulate human intelligence, but there's no guarantee. Thinking about thinking is not necessarily the same as thinking.

 

[Nocapszy]

Are both lines of thinking not equally bullshitting?

Not really. My point is that there's no difference. Unless we can draw a line that doesn't use what materials went into the composition as its axis, the process is to be understood as the same.

It's either different (which has yet to be explained so far in this thread) or it's the same.

True, the way we think is the way we think, however we can only program something based on our understanding of how we think. How we wanted them to work is reflective of our thinking... but it's not exactly the same as our thinking.

What's different? The medium?

What I'm trying to say is that we really don't know how we think, we just think.

Much like the electronic device you're tapping away at right now.

 

[Nocapszy]

The following paragraph in my previous post was an attempt to explain some of those differences. In terms of research -- probably less than you have.



This assumes that 1) We adequately understand how conscious thought works enough to simulate it. I disagree.
2) That once we figure out how it works, we are in fact able to simulate it accurately. My point in my previous post speaks to this problem. Is thinking about thinking the same as thinking? You say yes, I say no. If what's being modeled is the thought process itself (in this case, deduction) then yes. But in AI, what's being copied isn't the deductive process (which is always thinking) but the results of those deductions.

Lets say people think like x. Someone who doesn't know much about psychology might propose that people think like y. He figures that out by using x, because, well, he's thinking. But what does he use to build the AI? If he uses y, then he's off.

True. But that would be if he were intentionally composing a thought structure based on his supposition of the human process.

That's thinking about thinking, and there's no guarantee that he's right. If he uses x -- which in this case would be his deductive reasoning -- then he's right. At some point, with research and testing, x and y would converge and he might actually be able to simulate human intelligence, but there's no guarantee.

So... you're argument is that we don't have enough time to figure out how?

Thinking about thinking is not necessarily the same as thinking.

 

[Carebear]

Er... the brain does operate on a binary electrical system...

Hm... in a way, but not really.

Found this easy explanation on Neuroscience for Kids - Brain vs. Computer .

"A computer uses switches that are either on or off ("binary"). In a way, neurons in the brain are either on or off by either firing an action potential or not firing an action potential. However, neurons are more than just on or off because the "excitability" of a neuron is always changing. This is because a neuron is constantly getting information from other cells through synaptic contacts. Information traveling across a synapse does NOT always result in a action potential. Rather, this information alters the chance that an action potential will be produced by raising or lowering the threshold of the neuron."

 

[ThatsWhatHeSaid]

True. But that would be if he were intentionally composing a thought structure based on his supposition of the human process.

So... you're argument is that we don't have enough time to figure out how?

My argument, if you can call it that, is that robots and humans do not necessarily think alike, despite our attempts to model AI after human thinking, because in order to do so we have to really copy how thinking works, not simply how we think about thinking. Didn't someone try to build an AI robot based on MBTI? Sounds like a horrible idea and a fantastic example of thinking about thinking.

 

[Nocapszy]

My argument, if you can call it that, is that robots and humans do not necessarily think alike, despite our attempts to model AI after human thinking, because in order to do so we have to really copy how thinking works, not simply how we think about thinking. Didn't someone try to build an AI robot based on MBTI? Sounds like a horrible idea and a fantastic example of thinking about thinking.

Who said anything about robots?

This is not about robots or AI. It's about computers.

 

[Nocapszy]

Hm... in a way, but not really.

Found this easy explanation on Neuroscience for Kids - Brain vs. Computer .

"A computer uses switches that are either on or off ("binary"). In a way, neurons in the brain are either on or off by either firing an action potential or not firing an action potential. However, neurons are more than just on or off because the "excitability" of a neuron is always changing. This is because a neuron is constantly getting information from other cells through synaptic contacts.

Neurotransmitters... yeah I understand how the neurons 'communicate' -- if you look carefully, it's very similar to a wire, except where an electron would be passed to the next atom, we have neurotransmitters passing to the next neuron. Sometimes, it doesn't work -- if the wires were that thin in computers, we would run into a LOT of the same thing. No action potential is the same has having too much resistance in the wire, or a short circuit.

Information traveling across a synapse does NOT always result in a action potential. Rather, this information alters the chance that an action potential will be produced by raising or lowering the threshold of the neuron."

Again this excerpt is the same as a short circuit. The brain doesn't shut down, because it's got dozens of other connections compensating for the one that doesn't fire. A computer would because of the amazing difference in the number of viable components.

I'll read the rest of the article though.

 

[nightning]

Not really. My point is that there's no difference. Unless we can draw a line that doesn't use what materials went into the composition as its axis, the process is to be understood as the same.

Just to let you know, I do believe a brain fits under the category of an organic "computer"... However, how we believe we think does not truly reflect what's actually going on inside our brains. That was the distinction I was making... it in no ways contradicts that the brain is an organic matrix that can compute... and can be called a computer. I was merely pointing out at inconsistencies within what you've said. The way we built computers is a reflection on how we believe our thinking ought to work. It might not mirror how our brains actually thinks. A semi-digression... do excuse me.

"A computer uses switches that are either on or off ("binary"). In a way, neurons in the brain are either on or off by either firing an action potential or not firing an action potential. However, neurons are more than just on or off because the "excitability" of a neuron is always changing. This is because a neuron is constantly getting information from other cells through synaptic contacts. Information traveling across a synapse does NOT always result in a action potential. Rather, this information alters the chance that an action potential will be produced by raising or lowering the threshold of the neuron."

So the medium is different... We can make a computer to operate on thresholds like neurons... That's not difficult to program. We just need to make a system that is different from the normal PC to account for neural circuits but still it's doable. So besides physical and wiring differences... are there any other differences between a brain and a computer?

 

[Nocapszy]

So the medium is different... We can make a computer to operate on thresholds like neurons... That's not difficult to program.

Yeah. Likely we wouldn't be able to utilize everything in a human's brain, but we could start off with just a piece of neural 'fabric' -- a filet of graymatter; perhaps from a deceased rat or something, and wire it up like an integrated circuit. You'd be surprised how tiny and precisely those things are. I think it wouldn't be too difficult, given that we can get these microscopic transistors to line up, and those enormously small wires leading to the contacts, to do something similar with the brain tissue. The trouble then would be finding out just how much voltage it could handle without destroying it, and figuring out just how to program it.

This quote in direct correlation with what I was thinking:

We just need to make a system that is different from the normal PC to account for neural circuits but still it's doable.

Upon deeper inspection of this idea we share, I hold that it would be a near hopeless endeavor, not because it would be impossible to analyze the routes of the neurons, but rather that every slice would be so different that we'd have hardship trying to get each one to work on the same language.

In order to program them we'd have to be as familiar with these gray-matter micro-controllers as we are with good friends. They're all a little bit different, and thus will respond to things differently. We won't be able to get onto someone else's terminal, and run it the same way we'd run our own.

Until we can precisely control the shape and form of the sheets (perhaps a bioengineering project here) all of this will, unfortunately, remain a dream.

So besides physical and wiring differences... are there any other differences between a brain and a computer?

I've been thinking about this. I think rather than using transistors and resistors, the brain uses different timing to get the different reactions.

Comps have several different kinds of components that give different levels or electricity, or can hold it in place until it's needed among other thigns. The brain just uses one. Wires. Of course, the wires carry different types of "electricity" which makes up for its lack of diversity in the components themselves. Unless we're thinking on a larger scale, in which case, we could say that each lobe of the brain is a different component, on an unimaginable tier of complexity compared to what the real electrons are shifted back and forth between in a computer.

 

[Nocapszy]

This is interesting too. Off topic, but interesting.
This too.

I'm not sure; how many of you people are actually into computers? Or the brain?

I hope my posts aren't too esoteric.

 

[ygolo]

As long as we're linking sources...

I recommend:

Minds, Brains, and Computers: The Foundations of Cognitive Science : An Anthology (Blackwell Philosophy Anthologies)