Face It, Your Brain Is a Computer • Hacker News Board

Re: Face It, Your Brain Is a Computer

#21

Post by electrograv » Sat, Jun 27, 2015, 6:44 PM UTC

themgt wrote: post #9

I think the key fail is hidden beneath this statement, 'many “digital” computer switches are built out of analog components and processes'. In fact obviously, all computers are built out of analog components, in the sense that the entire universe is analog. However, this is the problem. Computers are designed to interpret or translate the messy analog world where nothing is exactly 1 or 0, into a precise, determinist…

> And I think it's highly likely that the brain does not work in this manner whatsoever, and that in fact, in the translation of reality into 1s and 0s, you've lost far too much to properly "compute consciousness"

You act as though the only datatype usable by computers is bool. That's ridiculous. Yes, computers are binary, but they can compute analog things just fine via floating or fixed point arithmetic to arbitrarily high precision, among many other things. To claim that the brain has a fundamental advantage over a Turing machine is to claim that there's some magical quantum (or otherwise) advantage in the brain's physical system that is un-simulate-able by a Turing machine.

You can't even argue that analog provides superior precision to digital representation of continuous values, because the analog part of the brain is represented with discrete units (accumulation of neurotransmitter molecules in various parts of the neuron)! In other words, our brain actually uses fixed point math (granted, incredibly complicated fixed point math :))

Re: Face It, Your Brain Is a Computer

#22

Post by linschn » Sat, Jun 27, 2015, 6:44 PM UTC

nirai wrote: post #19

linschn wrote: post #8

The only serious counter-argument I've heard against the hypothesis that the brain goes further than a Turing machine is that the brain does not implement an algorithm, with one input, some processing steps, and one output. Instead, it processes information as it comes, this information comes during the processing of previous information and the output partly loops back to the input. Short-circuits everywhere, no clo…

I'm a dualist but for the sake of argument, one approach is to claim that however the brain functions, it should be explainable by physics, and if physics is computable, then it should in principle be simulable by a Turing Machine.

The question of whether physics is computable is a fascinating one I'd really like to see solved before I die.

Re: Face It, Your Brain Is a Computer

#23

Post by nirai » Sat, Jun 27, 2015, 6:45 PM UTC

logician76 wrote: post #7

As long as science keeps thinking this, we won't have to worry about the singularity any time soon.

The problem here is that the so called singularity does not necessarily depend on machines having the so called qualia.

Re: Face It, Your Brain Is a Computer

#24

Post by rndn » Sat, Jun 27, 2015, 6:46 PM UTC

coldtea wrote: post #17

I find the metaphor bad. My brain might work like a computer (computation et al), but the word computer evokes actual existing computers to a layman -- and my brain doesn't have a von neumann architecture, nor does it run something similar to our software. A neural network does computation of course, but we wouldn't call it a "computer" in the sense we use the word when dealing with, well, computers.

I agree, calling it a computer is the origin of a lot of confusion. Perhaps "information processing" is a better term.

Re: Face It, Your Brain Is a Computer

#25

Post by jwmerrill » Sat, Jun 27, 2015, 6:48 PM UTC

themgt wrote: post #9

I think the key fail is hidden beneath this statement, 'many “digital” computer switches are built out of analog components and processes'. In fact obviously, all computers are built out of analog components, in the sense that the entire universe is analog. However, this is the problem. Computers are designed to interpret or translate the messy analog world where nothing is exactly 1 or 0, into a precise, determinist…

> Computers are designed to interpret or translate the messy analog world where nothing is exactly 1 or 0, into a precise, deterministic digital world where everything is either 1 or 0. And I think it's highly likely that the brain does not work in this manner whatsoever

But neurons totally do work by encoding information with discrete signals.

https://en.wikipedia.org/wiki/Action_potential#Process_in_a_...

Action potentials are discrete events, and just like hi or lo in a silicon computer actually correspond to a range of voltages, the exact voltage-vs-time curve of an action potential likely has considerable flexibility.

Re: Face It, Your Brain Is a Computer

#26

Post by return0 » Sat, Jun 27, 2015, 6:48 PM UTC

Osmium wrote: post #15

johansch wrote: post #10

Earlier quoted context omitted.

Who in their right mind would go to philosophers to get an understanding on how the human brain works at a physical level? Why not go to your priest instead? Same level of understanding.

I think that's somewhat unfair to philosophers. Besides the fact that science grew out of philosophy, even today philosophers are very good at defining problems and finding logical inconsistencies. As a scientist, that seems very useful indeed. To understand the brain on a "physical level", we have to know what questions we're asking first.

My experience is that philosophy of mind has almost nothing to offer to modern neuroscience. The questions being asked are the unhelpful ones and signify a general obsession with semantics, subjectivity and language tricks.

Re: Face It, Your Brain Is a Computer

#27

Post by sgt101 » Sat, Jun 27, 2015, 6:48 PM UTC

linschn wrote: post #8

The only serious counter-argument I've heard against the hypothesis that the brain goes further than a Turing machine is that the brain does not implement an algorithm, with one input, some processing steps, and one output. Instead, it processes information as it comes, this information comes during the processing of previous information and the output partly loops back to the input. Short-circuits everywhere, no clo…

I don't think that we are in the territory of having that argument at all to be honest. We don't know what the processes that the brain self evidently generates are - for example Intelligence, Consciousness. It isn't possible to discuss how these are executed (or not, depending) until we have some understanding of what they really are.

And we don't, and worse, we may not ever have access to these concepts, as Hofstadter and others have argued. It may not be possible to represent the idea of ideas in an idea.

That of which we cannot speak that thereof we must remain silent, or something.

Re: Face It, Your Brain Is a Computer

#28

Post by themgt » Sat, Jun 27, 2015, 6:49 PM UTC

linschn wrote: post #14

themgt wrote: post #9

I think the key fail is hidden beneath this statement, 'many “digital” computer switches are built out of analog components and processes'. In fact obviously, all computers are built out of analog components, in the sense that the entire universe is analog. However, this is the problem. Computers are designed to interpret or translate the messy analog world where nothing is exactly 1 or 0, into a precise, determinist…

You can approximate any analog value to arbitrary precision using a (by nature binary) Turing Machine, therefore you can bring the loss of information as low as needed. The question is whether the brain offers a more complete framework than all the models of computation we know today (which are equivalent to one another). To put it another way : is there a function that the brain can compute that a Turing Machine can…

But this is effectively begging the question. If you assume that reality can be completely described by functions that compute on natural numbers, then obviously anything within reality, including consciousness, can be simulated by a Turing machine.

Unless I've missed something, it's not actually been shown that the universe itself is Turing-computable, therefore there's no reason to assume everything within the universe could be equivalently simulated on a Turing machine.

It could also be that in theory, consciousness is Turing computable, but not in practice. For example just computing a protein fold for 1 millisecond takes hundreds of ASICs days of compute time. This might give some clue that complex physical wetware processes are not able to be efficiently simulated in real time on any realistic Turing machine.

Re: Face It, Your Brain Is a Computer

#29

Post by nirai » Sat, Jun 27, 2015, 6:50 PM UTC

linschn wrote: post #22

nirai wrote: post #19

Earlier quoted context omitted.

I'm a dualist but for the sake of argument, one approach is to claim that however the brain functions, it should be explainable by physics, and if physics is computable, then it should in principle be simulable by a Turing Machine.

The question of whether physics is computable is a fascinating one I'd really like to see solved before I die.

Physics is most probably computable, since people created it; the real question is whether the world is (Turing) computable, and from my point of view it is clearly not. I find it remarkable that most people believe it is or that it might be.

Re: Face It, Your Brain Is a Computer

#30

Post by themgt » Sat, Jun 27, 2015, 6:50 PM UTC

electrograv wrote: post #21

themgt wrote: post #9

I think the key fail is hidden beneath this statement, 'many “digital” computer switches are built out of analog components and processes'. In fact obviously, all computers are built out of analog components, in the sense that the entire universe is analog. However, this is the problem. Computers are designed to interpret or translate the messy analog world where nothing is exactly 1 or 0, into a precise, determinist…

> And I think it's highly likely that the brain does not work in this manner whatsoever, and that in fact, in the translation of reality into 1s and 0s, you've lost far too much to properly "compute consciousness" You act as though the only datatype usable by computers is bool . That's ridiculous. Yes, computers are binary, but they can compute analog things just fine via floating or fixed point arithmetic to arbitra…

> To claim that the brain has a fundamental advantage over a Turing machine is to claim that there's some magical quantum (or otherwise) advantage in the brain's physical system that is un-simulate-able by a Turing machine.

Again, please point to any research you know of indicating that the universe itself is fully Turing-computable.