The Brain Is Not a Computer (2016) • Hacker News Board

Re: The Brain Is Not a Computer (2016)

#21

Post by ChrisSD » Sun, Mar 10, 2019, 6:36 PM UTC

naasking wrote: post #20

skywhopper wrote: post #11

Earlier quoted context omitted.

Yes “computer” here means “digital electronic computer” and yes the storage, retrieval, and processing all happens very differently in the brain than digital computers are capable of. You may think these are dumb points for the author to make, but it’s not clear to me at all that the media or the VCs who buy the hype about machine learning, AI, and self-driving cars realize just how different they are.

> Yes “computer” here means “digital electronic computer” and yes the storage, retrieval, and processing all happens very differently in the brain than digital computers are capable of. No, it's incorrect to use "digital computer" here. More correct might be a von Neumann architecture computer. But then that shows that the author is attacking a strawman: people comparing brains to computers aren't limiting themselves…

Are you sure? This mostly comes up in pop sci explanations where von Neumann architecture may as well be synonymous with "digital computer" because it's all the vast majority of readers are familiar with.

Re: The Brain Is Not a Computer (2016)

#22

Post by behringer » Sun, Mar 10, 2019, 6:40 PM UTC

TheOtherHobbes wrote: post #19

baddox wrote: post #3

Oh dear. This isn’t my field at all, and maybe I’m just doing a bad job of understanding their point, but this sounds completely bogus. Really...the brain doesn’t create representations of visual stimuli or store memories? Under what possible definitions of those words can this statement be sensical? Surely the author believes that visual stimuli cause measurable changes in brain state, and that people can indeed rem…

I'm not sure I understand the piece either, but I think it's trying to say that human memories are associative, sequential, and distributed rather than localised. So there isn't "a representation" in the discrete sense. It's more like the entire system changes, and it's impossible to physically scan specific elements of it to retrieve selected content. You can trigger selective recall, but you're triggering a complex…

I think researchers are working on this very thing, and it seems like the brain does store "visual representations" or at least it can't be ruled out at this point.

https://www.theregister.co.uk/2013/08/20/mindreading_mri_spo...

And of course, why wouldn't the brain store a visual representation of what you remembered. That would be the easiest way to store and retrieve it, which is why we do that on computers as well.

Re: The Brain Is Not a Computer (2016)

#23

Post by jacobush » Sun, Mar 10, 2019, 6:41 PM UTC

TheOtherHobbes wrote: post #19

baddox wrote: post #3

Oh dear. This isn’t my field at all, and maybe I’m just doing a bad job of understanding their point, but this sounds completely bogus. Really...the brain doesn’t create representations of visual stimuli or store memories? Under what possible definitions of those words can this statement be sensical? Surely the author believes that visual stimuli cause measurable changes in brain state, and that people can indeed rem…

I'm not sure I understand the piece either, but I think it's trying to say that human memories are associative, sequential, and distributed rather than localised. So there isn't "a representation" in the discrete sense. It's more like the entire system changes, and it's impossible to physically scan specific elements of it to retrieve selected content. You can trigger selective recall, but you're triggering a complex…

It is very hard to introspect into what cannot be introspected... but with such things as how many flats the key of F major has (which I have a very vague understanding of what the question even means) ... I have a hunch that such data recall is incredibly crosschecked and error-corrected by correlating to many many other known facts. You may associate to maths lessons in elementary school, visual charts, piano lessons, math knowledge, that special smell of the paper and number theory (not necessarily fancy, just simple stuff) that altogether makes it impossible to corrupt upon recalling, how many flats the key of F major has.

While as for the kinds of memories we easily do corrupt upon reading (was the perpetrator tall or blonde?) there may be no such data to draw from. Or worse, generalised "data" which is good for pattern matching but not for actual recall. (The stuff prejudice is made of.)

If so, then there is no contradiction or problem, just a very intricate mesh of data.

(Also I have an idea which is not even a hunch, but pulled out of thin air, that our "logic" and "memory" are much more intermixed in our wetware than in computers.)

Re: The Brain Is Not a Computer (2016)

#24

Post by traject » Sun, Mar 10, 2019, 6:44 PM UTC

TheOtherHobbes wrote: post #19

baddox wrote: post #3

Oh dear. This isn’t my field at all, and maybe I’m just doing a bad job of understanding their point, but this sounds completely bogus. Really...the brain doesn’t create representations of visual stimuli or store memories? Under what possible definitions of those words can this statement be sensical? Surely the author believes that visual stimuli cause measurable changes in brain state, and that people can indeed rem…

I'm not sure I understand the piece either, but I think it's trying to say that human memories are associative, sequential, and distributed rather than localised. So there isn't "a representation" in the discrete sense. It's more like the entire system changes, and it's impossible to physically scan specific elements of it to retrieve selected content. You can trigger selective recall, but you're triggering a complex…

The encoding assumption in neuroscience is also nicely argued against in this preprint. https://www.biorxiv.org/content/biorxiv/early/2018/07/13/168...

Re: The Brain Is Not a Computer (2016)

#25

Post by ChrisSD » Sun, Mar 10, 2019, 6:45 PM UTC

TheOtherHobbes wrote: post #19

baddox wrote: post #3

Oh dear. This isn’t my field at all, and maybe I’m just doing a bad job of understanding their point, but this sounds completely bogus. Really...the brain doesn’t create representations of visual stimuli or store memories? Under what possible definitions of those words can this statement be sensical? Surely the author believes that visual stimuli cause measurable changes in brain state, and that people can indeed rem…

I'm not sure I understand the piece either, but I think it's trying to say that human memories are associative, sequential, and distributed rather than localised. So there isn't "a representation" in the discrete sense. It's more like the entire system changes, and it's impossible to physically scan specific elements of it to retrieve selected content. You can trigger selective recall, but you're triggering a complex…

Surely retrieving an individual fact is more like a conditioned response than it is storing some bytes on a harddrive?

When I say "what is 9 times 9" your brain activates all the pathways (probably trained in childhood) that lead to you thinking "81" in a similar way to Pavlov's dog.

Re: The Brain Is Not a Computer (2016)

#26

Post by ozy » Sun, Mar 10, 2019, 6:47 PM UTC

Imagine a database that stores strings using a common prefix method, one could make the same claim: this database does not store or retrieve strings. And yet it does.

The model of what something does is implemented by an underlying mechanism. But for many reasons the mechanism doesn't have to be, and often isn't, a naive translation of the model.

Re: The Brain Is Not a Computer (2016)

#27

Post by naasking » Sun, Mar 10, 2019, 6:52 PM UTC

ChrisSD wrote: post #21

naasking wrote: post #20

Earlier quoted context omitted.

> Yes “computer” here means “digital electronic computer” and yes the storage, retrieval, and processing all happens very differently in the brain than digital computers are capable of. No, it's incorrect to use "digital computer" here. More correct might be a von Neumann architecture computer. But then that shows that the author is attacking a strawman: people comparing brains to computers aren't limiting themselves…

Are you sure? This mostly comes up in pop sci explanations where von Neumann architecture may as well be synonymous with "digital computer" because it's all the vast majority of readers are familiar with.

A "digital computer" is simply one in which information is represented in binary. The encoding of information does not alter the process of computation, so that's an irrelevant detail to compare to.

We have plenty of Harvard architecture computers (eg. DSPs), and there are plenty of other computational architectures (DNA computing, optical computing, quantum computing).

Re: The Brain Is Not a Computer (2016)

#28

Post by russdill » Sun, Mar 10, 2019, 6:55 PM UTC

TheOtherHobbes wrote: post #19

baddox wrote: post #3

Oh dear. This isn’t my field at all, and maybe I’m just doing a bad job of understanding their point, but this sounds completely bogus. Really...the brain doesn’t create representations of visual stimuli or store memories? Under what possible definitions of those words can this statement be sensical? Surely the author believes that visual stimuli cause measurable changes in brain state, and that people can indeed rem…

I'm not sure I understand the piece either, but I think it's trying to say that human memories are associative, sequential, and distributed rather than localised. So there isn't "a representation" in the discrete sense. It's more like the entire system changes, and it's impossible to physically scan specific elements of it to retrieve selected content. You can trigger selective recall, but you're triggering a complex…

Closest equivalent in CS would be lossy storage, Monte Carlo algorithms, or of course, neural networks.

Re: The Brain Is Not a Computer (2016)

#29

Post by ChrisSD » Sun, Mar 10, 2019, 6:57 PM UTC

naasking wrote: post #27

ChrisSD wrote: post #21

Earlier quoted context omitted.

Are you sure? This mostly comes up in pop sci explanations where von Neumann architecture may as well be synonymous with "digital computer" because it's all the vast majority of readers are familiar with.

A "digital computer" is simply one in which information is represented in binary. The encoding of information does not alter the process of computation, so that's an irrelevant detail to compare to. We have plenty of Harvard architecture computers (eg. DSPs), and there are plenty of other computational architectures (DNA computing, optical computing, quantum computing).

Sure but my point is they aren't what people think about when you say "digital computer" or use that as an analogy. People think about what they are familiar with in their every day lives.

Re: The Brain Is Not a Computer (2016)

#30

Post by russdill » Sun, Mar 10, 2019, 6:57 PM UTC

naasking wrote: post #20

skywhopper wrote: post #11

Earlier quoted context omitted.

Yes “computer” here means “digital electronic computer” and yes the storage, retrieval, and processing all happens very differently in the brain than digital computers are capable of. You may think these are dumb points for the author to make, but it’s not clear to me at all that the media or the VCs who buy the hype about machine learning, AI, and self-driving cars realize just how different they are.

> Yes “computer” here means “digital electronic computer” and yes the storage, retrieval, and processing all happens very differently in the brain than digital computers are capable of. No, it's incorrect to use "digital computer" here. More correct might be a von Neumann architecture computer. But then that shows that the author is attacking a strawman: people comparing brains to computers aren't limiting themselves…

A von Neumann architecture does not limit one to a single method of information storage or retrieval, it just limits the efficiency of different methods.