Brain Memory Limit?

[hauss]

Is there limit, as in hard drives, to how much info a human brain can hold? I don't mean problem solving or other cognitive abilities, I just mean holding info like a database?

Are those people on Jeopardy close to their capacity?!

I know there is no solid answer for this, but what I am asking for is the best expert guess we have at this point out in the medical community.

[Blake]

http://www.newscientist.com/lastword...e.jsp?id=lw340

Many studies have established that the human brain stores information in ways that are quite different to most computers. Sequential, flawless access is a task still best suited to PCs. But there are exceptions.

Antonio di Marco Magliabechi, born in 1633, became the library for the Grand Duke of Tuscany. In his case, he could read and memorise, permanently, entire tomes with only a cursory glance at each page, and did so until his death at the age of 81. He first demonstrated this ability at a young age, when he was still an illiterate fruit-dealer's apprentice.

Given that the Encyclopaedia Britannica has only just been released on CD-ROM (capacity approximately 600 megabytes, and highly compressed at that), it is fair to say that, for some people at least, no limit can even be theorised as to the storage space we possess.

[Trigonal Planar]

I've often read that our memories are virtually limitless but the problem comes in accessing this massive, disorganized database quickly and reliably.

[skaterboarder87]

Yes, our storage capabilities far exceed how long we will live. In other words, it's nothing you will have to worry about.

[county]

You have to be careful; after passing 50 years of age (assuming you have been learning all your life) each new thing you learn pushes out something you have previously learned.

In other words when you are introduced to someone new - you give up someone previously met.

Now that I am past 50 I try to be careful as I don't want to lose anything or anyone important.

[Joe Random]

Originally posted by county
after passing 50 years of age (assuming you have been learning all your life) each new thing you learn pushes out something you have previously learned.

Sounds like an old wives' tale to me. Do you have a cite for that?

[SPOOFE]

I think that he was being facetious. At least, I hope he was...

[Shalmanese]

Some estimates I have heard put human storage capacity at around 200GB which is less than some people have in their computers. In fact, its less than at least one person I knows porn collection.

[Zeldar]

It may depend on which brain is in question.

I had a friend in high school who wasn't overly bright to begin with, but by no means "slow."

I started to tell him something I had heard, and he said something like, "Don't tell me that. I have too much to remember now."

That made a big impression on me and I began trying to figure out how much info he just blocked out. It was hard to tell.

[Already in Use]

Originally posted by Shalmanese
Some estimates I have heard put human storage capacity at around 200GB

What would this even mean? How do you analyze human memory in those terms? It's not ones and zeroes we're talking about here, so I'm wary of comparisons to computer storage space.

[eburacum45]

The brain can have perhaps 10¹⁵ different states, based on data from Roger Penrose; although I don't agree with all his ideas about quantum consiousness, it is a useful estimate.

If this represents the maximum data capacity of the brain it seems likely that the brain has a capacity of ~10, 000 gigabytes.
This is the absolute maximum- the reality is probably a lot less.

I doubt that all of this capacity can represent memory,but in a separate back of the envelope calculation it seems that you would have no room for any more data storage after a thousand years of life, and your brain would probably start overwriting vital data well before that.
__________________
SF worldbuilding at
http://www.orionsarm.com/main.html

[eburacum45]

On the other side of the coin, active learning might be concentrated in the hippocampus.

This link
is a story about studies showing that remarkably few cells are activated during the process of aquiring new data, apparently these cells are located in the hippocampus;

while in other news
this link describes an artificial implant in the hippocampus which may aid memory. Eventually I imagine most of a person's memory will be located in other media, accessed by chips like this.
__________________
SF worldbuilding at
http://www.orionsarm.com/main.html

[SPOOFE]

Some estimates I have heard put human storage capacity at around 200GB

But human memory must be compressed something fierce, which probably raises that number by several orders of magnitude.

[Bill H.]

Interesting reading by the very cool Ralph Merkle: http://www.merkle.com/humanMemory.html

[Planet of the Shapes]

Originally posted by Shalmanese
Some estimates I have heard put human storage capacity at around 200GB which is less than some people have in their computers. In fact, its less than at least one person I knows porn collection.

In a way, i almost hope the capacity of the human brain is only 200GB. Then that "person you know" will have more information in his porn collection then in his brain. That'd be a scary thought...

On a slightly more serious note:

Eburacum45 - is that 10^15 based on the number of different states for each individual electron etc in your brain? Or just based on states of the neurons themselves?

[bryanmcc]

The human brain is made of a finite amount of matter, and is therefore limited in the number of discrete states it can have, and thus the amount of information it can contain. The Bekenstein Bound gives how much information this could maximally be:

The Bekenstein Bound is an upper bound of the amount of information inside a spherical region with a given energy. Information in this context is to be understood as distinguishable (quantum) states. Due to the uncertainty relations it is possible to derive a bound of the form

I <= (2 Pi E R)/(hbar c ln2)

where I is the information, E is the energy, R is the radius, hbar Plank's constant, c the speed of light. It can also be written as

I <= k M R

Where M the mass in the region and k a constant having the value ~2.57686*10^43 bits/(m kg). This bound was derived by J.D. Bekenstein in another but equivalent form, relating the entropy of black holes to their area (S = A/(4 hbar G), where A is the area of the event horizon).

Assuming a brain mass of 1.5kg and a radius of .1m, the maximal possible amount of information that could be contained within such a region would be

I <= (2.57686*10^43bits/m kg)(1.5kg)(.1m) ~ 4*10^42 bits

Of course, this is the most information that could possibly be stored within such a volume, and the human brain is likely to be much less efficient than that.

[Chronos]

The problem is that the human brain stores information so much differently than a computer. As an example: You know what your Aunt Mildred looks like, right? So that's a piece of information you have. How many bits in that piece of information? Well, we could scan a picture of Aunt Mildred into the computer at maybe a few hundred kilobytes. But that would only show her from one particular point of view, in a particular pose. Really, we'd need to constuct a full 3-d model of Aunt Mildred, together with physical information about where that model can bend and turn. I imagine that this would bring us up to the vicinity of 1 to 10 megs.

So, does this mean that for each person you can identify, you're storing 10 megabytes of information? No. The way a person remembers what Aunt Mildred looks like is associative. Maybe you remember Aunt Mildred as "She looks like Mom, but with dimples and longer hair", which is very little information indeed (once you already know what Mom looks like). Or maybe "She looks like an adult female human, with grey hair, glasses, dimples, and a nose that turns up just so". OK, so now we've still got to store some sort of model for that distinctive nose, but the rest of the memory is built from off-the-shelf components.

The point is, all of our memories are interconnected. Most of the description of Aunt Mildred, you got out of the way by saying she's human. You just need one model of a human stored in your brain, and you can use that (with variations) for all of the people you meet.

[eburacum45]

Originally posted by Planet of the Shapes


Eburacum45 - is that 10^15 based on the number of different states for each individual electron etc in your brain? Or just based on states of the neurons themselves?

This figure is supposed to be based on the states of the neurons and synapses, 10¹¹ neurons each with 10⁴ synapses
but it is unlikely that each synapse holds a discrete bit of memory-
some discussion of processing and consciousness here, not all of which I agree with;
http://www.kurzweilai.net/articles/a...ml?printable=1

[Planet of the Shapes]

Thanks for the link eburacum45 - its very interesting.

I've heard about Penrose and his quantum brain stuff before when i read his book Shadows of the Mind. I'm not sure i agree with it either though.

Another interesting theory i heard is that memory may be stored in a similar way to the way a hologram is stored. This means memories could be stored "on top of each other", which could also boost brain capacity.

[Epimetheus]

Originally posted by Already in Use
What would this even mean? How do you analyze human memory in those terms? It's not ones and zeroes we're talking about here, so I'm wary of comparisons to computer storage space.

Sort of. There are only really two types of Neurons as far as we know. (there are millions of types, what I mean is they only send two types of signals) Excitation Neurons and Inhibition Neurons. So you could technically assign a 1 or a 0 to either of those.

[Epimetheus]

Originally posted by eburacum45
some discussion of processing and consciousness here, not all of which I agree with;
http://www.kurzweilai.net/articles/a...ml?printable=1

I understand your reservations on this. Looks a bit hoaky to me in some regards. How valid is this site? It doesnt' seem quite like quantum physics, and it doesn't seem quite like neuroscience, so IMO it seems more like new age psuedoscience. Or just plain steps out of the bounds of science and falls flat on it's face across the grounds of philosophy. (I will have to read it when I get more time, so I am going to bookmark it- however from just reading a few paragraphs here and there it's veracity seems quite strained)

[Shalmanese]

Kurweil is a relatively respected if sometimes far out future writer. Hes written some fairly asine stuff in hindsight but he has also done a fair bit of good work. Hes certainly not a quack.

[toadspittle]

Does anyone have any idea how ANY info. is stored in our brain? I've never actually seen an explanation. Do certain proteins accumulate in the neurons, representing various data? Do the neurons have to fire continuously to keep information flowing (like RAM, which only stores memory while current goes through it ... turn it off and it vanishes)?

[Trigonal Planar]

The most I've heard is that memories are stored within the interconnected pathways of the neurons themselves. Stimulating certain patterns "triggers" a given memory.

[Trigonal Planar]

The most I've heard is that memories are stored within the interconnected pathways of the neurons themselves. Stimulating certain patterns "triggers" a given memory.

[toadspittle]

IOW, when someone forms a memory, neuron A may form new connections to B and D and sever connections with C and K, for instance, and this new "circuit" (A, B,D, plus whichever neurons they're connected to) = a memory?

[LiveOctopus]

Originally posted by county
You have to be careful; after passing 50 years of age (assuming you have been learning all your life) each new thing you learn pushes out something you have previously learned.

Like that time I took a home wine-making course and forgot how to drive?

[county]

exactly