Memory Replacer: Remove old neurons containing useless information and replace them with new neurons

Hi there,
I read somewhere that that total memory capacity of the human brain is about 2.5 petabytes. Assuming this is true, I begin to wonder how much “space” is now left in my head after living for so many years and storing so much information during this period. I then realise that I also hold a lot of useless information in my head (The movie Daredevil or Ghostrider :stuck_out_tongue: or the number of tiles on my school bathroom floor :D)
Using brain scanning is it possible to locate portions of the brain which hold these kinds of useless information and remove them and in their place insert fresh neurons which are made using cell culture??
Thanks

Right now, no way is that possible. We have no way to track where each specific memory is stored let alone do the surgery required. Maybe in the future, maybe not.

Relevant Far Side

People with hyperthmesia can recall with clarity and detail most of the events they personally experienced over most of their lifespan.

You are unlikely to approach the level of storage they require, and they don’t appear to run out, so it is unlikely that you will either.

And no, storage areas in the brain cannot be mapped, and new neurons cannot be made to grow in areas of the brain where neuron loss has occurred (either by deliberate removal or by disease).

And the reason why memory can’t be mapped (for now) is because it’s all over the place. When people in imaging devices are asked to recall events, the brain lights up in lots of different places.

1 - We generate new neurons throughout our lives

2 - Stem cell motor neuron treatment in monkeys reduced Parkinson’s symptoms

Long term memory is not stored in individual neurons, but rather in the connections that are formed in between neurons. In other words, memory is encoded, so to speak, by the directed formation of proteins as directed by something called cAMP response element-binding (CREB) protein, which directs transcription of particular genes involved in memory formation. The way this works is, like all of cognition, still not very well understood, but while long-term memories are persistent, they are also highly plastic; that is, they change over time and with recall. It should also be understood that what you think of as a memory–like your first kiss or getting cake thrown at you at your eighth birthday party–is not one single monolithic experience but rather a collection of different stimuli (visual, auditory, tactile, olfactory) and associated emotional response. All of those individual components are stored in different areas of the brain that are associated in processing each of those particular stimuli, and are tied together in recall by a mechanism that is not at all well understood, except that it effectively alters your sense of time and location, i.e. when you recall a memory you are actually, on a certain level, reliving that memory (or at least your polluted recollection of it).

Regarding what is really known about cognitive memory, Nobel laureate Eric Kandel is one of the undisputed experts in this field, and I think his quasi-autobiography, In Search of Memory: The Emergence of a New Science of Mind is probably the best survey of the field for the layman. For all of that, even Kandel frankly admits that while we have a much better grasp on the mechanics of formation of individual memory pathways, we really have essentially no clue about the higher level cognition involved in memory formation and recall.

Stranger

Yes we do, but in the CNS, the response to damage tends to be scarring and damage isolation, not neuroregeneration. Research is ongoing into preventing glial scarring in the CNS, but it is not possible now (although there are some promising lines of enquiry).

It is one thing to induce dopamine producing neurons and transplant them into a brain where they continue to produce dopamine (which will reduce Parkinsons symptoms). It is quite another for those cells to be integrated into functional networks to enhance neural function - that has not been demonstrated.

Ignore the bit about replacing neurons … the basic bit is indeed how it works to some degree. Part of new memory formation involves pruning of past storage and suppression of old patterns, much of which occurs during sleep.

I don’t have the time to chase the cites right now but maybe later today I can dig some.

There’s a compelling theory though that these peoples’ memory feats have more to do being OCD types who keep detailed journals of their lives which they obsessively review and memorize, rather than having any special powers. There’s an interesting article about the phenomenon on Wired.

The memorization powers of idiot savanst like Kim Peek, on the other hand, are truly astonishing.

Not the cite I was thinking of (but ironically cannot remember exactly where I read it!) but this chapter discusses it some.

So from what I’ve come to understand thus far, I don’t really need to fret about memory space being lost because they keep getting recycled over time… yes?
So to extend this further… by some miracle of technology, I’m able to live a life spanning 500 years… what is going to happen?

  1. By the time I reach 100 years, memory is full and I keep forgetting things… almost like Alzheimer?
  2. Will forget most or ALL of my childhood / adolescent memories and keep replacing them with new memories?
  3. Forget EVERYTHING because by then age would’ve killed my cells and the’ve lost all of their memory potency.
    (The queries ultimately question the plasticity of the human memory and the age effects on them.)
    .
    .
    .
    .
    Also, (Another burning question), if I were to take one particular neuron storing one particular piece of datum and cloned it… Would the clone also contain the datum from the mother cell or would it be empty?
    Thanks

Assuming the current understanding is correct, the datum/data contained in one neuron is only valuable in it’s current physical relationship to everything around it. Clone it and move it somewhere else and that datum/data is basically meaningless.

Like if you saw the number 5 on a computer screen and you pulled out 1 of the pixels that make up the 5, that pixel by itself isn’t really valuable info. It’s only value is when it is considered in relation to and along with the other lighted pixels making up the 5.

Again, there is no data stored in an individual neuron any more than there is in a transistor. A neuron is fundamentally a switch or amplifier. Memory is a collection of protein structures and neurotransmitters whih facilitate storage and recall. So no, cloning an individual neuron would not contain any data. Even reproducing an entire complex of molecular structures would not accurately reproduce a memory in the high level cognitive sense.

In general, it is a mistake to take any analogy between how the brain functions and how an digital computer works very deeply. Computers are organized along very linear, highly rigid plans with a fundamental distinction between structure (hardware) and response to stimuli (firmware and software). The brain, on the other hand, is comlpexly networked and very plastic (often called “wetware”) which changes basic structure in response to stimuli. An analogy of the distinction can be made by comparing a model train set (which has a set number of routes along which the train can travel regardless of what combination of directions are given) and a climbing vine, which will change diection as it grows in complex and difficult to accurately predict fashion, often splitting off and going sideways rather than straight up.

Stranger

It does contain some data.

1 - The physical structure creating links to other cells which change over time due to learning
2 - Also, due to learning, the neuron’s DNA changes (gene expression) to maintain synaptic connections at the new level over the long term

Again, memory is the experience of particular patterns of neuronal assemblages firing in particular orders. One school of thought (heh, literally) is that a conscious thought occurs when a particular pattern of firing resonates in a feedback loop. (See both temporal binding/recurrent thalamocortical resonance and Stephen Grossberg’s Adaptive resonance Theory – “All Conscious States are Resonant States.”)

Do the math. Start with this:

How many potential different patterns are there? Gotta tell ya, the number is bigger than I can count even using my toes, and the toes of every person who has ever been alive I suspect. The issues is selecting between all these possible patterns and not having some get in the way of others.

Yes, but there’s not a one-to-one correspondence between a synapse and a piece of knowledge.

Even the smallest piece of data in the brain is “smeared” across hundreds or thousands of synapses. And many of those synapses will simultaneously be taking part in encoding other pieces of data.

Yes that is true (see my earlier post).

But it’s false that a neuron contains “no data”, which was the point I was correcting.

The SciAm take on the op:

FWIW.

The other bit is that there is little doubt that memories are compressed and then recreated filling in the gaps with stuff that we make up as we remember but are sure is real.

Yes, while it is in situ, but remove the neuron from its place in the brain and all that structure, the arrangement of axonic terminals and where they terminate, is gone. They are just blowing in the wind. The “information” their arrangement once embodies will not stay with the neuron when you remove it from its place, so in an important sense, the relevant sense, this information is not “in” the neuron.

Um, no. The DNA most certainly does not change. That is the whole point of DNA.

Gene expression is not a change in DNA. It is a (temporary, dynamic) change in RNA and protein, and (once again) there is no good reason to think that anything distinctive about the condition of a particular neuron’s RNA and protein will survive the cell being removed from its place in the brain. Again, it will not carry with or “in” it any information relevant to any memories it contributed toward encoding whilst it was in place.

The DNA in the cell does carry data, of course, but not data relevant to any particular memory.