The minimum estimate for the human brain's capacity I have encountered was about 3 terabytes, and the maximum as von Neumann's estimate, 2.8E20 bytes.

My question is : how are they able to make this sort of comparison? I mean, I've heard on another thread that it compresses data in an extremely efficient manner, but then occasionally we still forget things and if we try to replay something in our heads, they are always very blurry (as reconstructed from other vague clues). And that makes the data quality sorta fuzzy. Also there are some things that 'come to us' when we don't exactly intend to recall it but 'escape us' when we try to recall something intentionally.

This remindes me of a teaching theory someone once explained to me. It was called the Sponge Ttheory and it said that students have a limit to what they can learn, and if you overload them they will either have a meltdown or forget facts learned in the past in order to make room for the new information. This is not unlike a computer, you must deleat programs when it runs out of memory.

Sponge Theory (if that's what it's called) definitely makes sense. There's a sort of passive ignorance in visual perception, especially that which goes on in the more subtle details. If the brain does this to not become overloaded, then it demonstrates that its storage capacity is indeed finite.

Actually, I think you guys are looking at this wrong. While the brain is easily explained in terms of computers, the analogy is just a little off. Our brains don't work in terms of electrical binary, but rather chemical shades of gray. I'm no expert, but as I understand it, thoughts aren't completely forgotten or deleted, they are rearranged. Granted there are outstanding circumstances that would cause this to be different.

I believe a more accurate number is closer to 6Tb.
And Lomion, the brain is Not easily explained in terms of computers. A computer can sufficiently model a human brain, but cannot even come close to it's raw 'processing' power, to say nothing of its 'seek time'. The analogy is weak. Not useless, but weak.

That's a lot of information for a species whose processing capabilities runs something like a fractional megahertz four-bit chip.

The human brains actual mathematics power is about 10 flops when considering only maths, but taking into account all the other stimuli bumps that figure to 10 petaflops.

When I said the brain is easily explained in terms of computers, I meant that it's easily explained in layman's terms. They are metaphorically similar. I totally agree that the computers are a terrible example for in-depth explanations. Sorry to sidetrack...

"Seek time." Now there's another blurry analogy.

It's very true how you can't exactly compare the brain with computers right off the bat, because brains are still vastly more complex than personal computers, which perform relatively predictable functions with hardware that performs just several orders of magnitude below the processing power of the brain.

Yet still it is somewhat ironic that comparisons continue to be made despite the lack of similarity. This is probably because computers are the best man-made analogies that we can make, other than those which are natural such as the brains of other species like dolphins and pigs.

Suppose this happens: You grow a small brain in a test tube, fully alive but vegetated in the sense that you cut off all five senses. The only input would be something like electrical pulses not unlike that of computer input. Would the brain be able to interpret, store, or retrieve this information?

That's an interesting question. It's rather hard to say what exactly would happen (for me at least), but I'm pretty sure the brain would adapt, and be able to interpret, store, and retrieve the information. I don't know how it would use that information, or if it would interpret it correctly, but I'd be willing to bet it would adapt to it somehow.

I think that depends on if the brain can be programmed? I don't think anyone here can answer that question, but it is fact that the brain is functional enough that what we call "anomolies" exist with relative Extraordinary abilities that make us wonder why most are incapable. i guessl you can take from that the human brain works difirently when trained or deprived some how through natural and unatural causes alike. So i guess that if it were possible to program a fully functional brain with sort of an algorithm based union seeing how no eyes are attached to percieve anything, maybe it is possible?

Well as far as why comparisons are made because computers are needed and routinely used to study Neurological data like simple brain waves to the more complex EMP’s. Bottom line I think the reason is because the human mind these days works hand in hand with Computers to solve the world’s mysteries. Someday as we all fear that the Human brain will become obsolete.

Just my simple theory, good thread BTW
http://brain.web-us.com/40hz/default.htm

I don't really see how you could ever say that the human brain has some many "bytes" of memory. What does that even mean? A byte is a unit of binary digits, and as far as I know the brain doesn't store information with binary digits. Are they talking about how much information the brain can hold? If so, how would you ever even quantify that?

What about the brain's emotions?

Would it have any? Emotions are certainly another piece of information stored within the brain, so there is no reason a "test-tube brain" wouldn't have any emotions (assuming you "uploaded the required information"). BUT, with computers using binary digits as information and humans using something else entirely, would the brain even be able to understand anything being fed into it?

The brain would have to be fed information from some other source, but of course the other source would have to be compatible. I think we also need to understand how the brain stores information more completely to make the assumption that we could quantify it, and then subsequently convert that quantity into "kilobytes" and things of that nature.

How did you have encountered this numbers ?

There's a guy who's memorised thousands of books, but you could probably fit all of them onto one or two DVDs. Difference is, the human brain uses completely different 'data structures' to a computer. Not mere sequential storage.

I have encountered these numbers across the Internet.

I highly doubt von Neumann's estimate, however, although maybe the figure is probably closer to somewhere in the terabyte range, assuming a comparison can be made.

Someone said, human mind uses a type of super efficient fractal math technology, or fractal geometry. There was a company who sold such a database analysis product for storage and fast analysis - I forget the name. I almost used the product in a datawarehouse application for Bellsouth. They could do 1000:1 compression. The human mind perhaps could do a million to one compression.

Because the retrival process usues an algorithm, any problems in that algorithm could have issues in remembering stuff.

My thought is that because some savant people who had brain damage can recall perfectly, I think, there are other modifiers and filters in normal people that cloud the recall. Perhaps this is due to the survival algorithm which may be separate but connected to the recall mechanism.

that could be a good guess.. who the fuck knows..

you think of something that reminds you of something else. the speed of the brains 'database queries' is pretty impressive. usually pretty small queries, but found quickly in a large database. sometimes with help from related queries and of course, cache.

Because doctors have found different regions of the brain contain different memories, it is possible the brain divides the database into multiple smaller databases and shuffles the old ones for a higher compression rate except the ones that are so profound that it stays in a quick access format. If we can figure out exactly how that works, we can speed up relational query very fast.

I did not have a need for such complex queries but then thought of anti-terrorism needs where specifc dots need to be connected rather than querying the whole planet...I have some ideas...we shall see ...

Not too seriously:
100 Billion Neurons. Each neuron has various 'weighted' inputs. Randomly giving 2 bytes to model the accuracy of each weighting.... each neuron can have upto around 10,000 connections. Let's mess about and say (guess) more like 100 on average.

10,000,000,000,000 bytes = 10TB, give or take a few spliffs.

Though, a lot of that will be more or less fixed pathways.