#neural prosthethic

[genidma]

Memory : The basic building block of cognition | And other meditations

The topic of how memories are encoded, accessed, enriched, modified (plus) is a super interesting topic. 

A memory or memories are at the root of everything that an agent would do. Whether biological or non-biological. But for the sake of this blogpost, I am mostly going to focus on the cognitive systems that are powered atop the biological wetware. 

It appears that evolution has enabled patterns. These patterns have emerged over a longer time period and different patterns perform different functions. There are many patterns driving the visual system. Same for the other systems. Then, there are patterns that work in conjunction with each other. 

There are a collection of patterns which determine why we’d want to stop by and look at a flower. Although, I am not really sure as to what the evolutionary reason for such an activity would be. Meaning, what evolutionary function does this activity perform. 

There is another collection of patterns that takes a cue from the visual or auditory system and would then trigger a mechanism. In-turn that mechanism would pump a specific chemical through a gland, which would increase the probability of the human engaging in some kind of a response. 

The schema determines how to process and route the stimuli and what set of a response ought to be generated. Specific patterns are being engaged for these set of purposes. Again, the base of all of this is comprised of memories. 

On this topic, one of my favourite quotes is from a book titled ‘On Intelligence’ which has been written by Jeff Hawkins. The quote being:

To me, consciousness is this super loaded word. There is a different definition to the word, contingent on who you decide to talk to. 

Some equate it to a soul or a spirit. 

Others hypothesize that the roots of consciousness are intertwined with quantum effects.  

I personally do not see a need to invoke quantum physics or a phenomenon that has never been quantified. This, at a time, when we continue reverse engineering the different sub-functions of the brain. 

Where I am going with this is, that the various functions performed by the brain/mind can, quite reasonably be encapsulated within the framework of the word cognition. 

And going back to what I started with, the roots of cognition are grounded in memories. Memories govern pretty much everything that we do. Including but not limited to:

How our behaviour is regulated (Memories stored, mostly in the subconscious impacting schema) 

The kind of stimuli we let in through our senses. 

Learning.  

As well, memories govern how we conduct our day to day lives. Memories have a direct hold over our schema. This is tied in with how we process stimuli and how we choose to make objective sense out of that stimuli. A process whereby an existing mental model is either affirmed/updated or something else occurs. 

As a species we are taking the very first steps towards demystifying how memories work. Here’s a thin slice and this certainly isn’t an exhaustive list:

Optogenetics: Using tools from genetic engineering, as well light pulses in order to map, activate, de-activate specific neurons. Note: I personally find that the optogenetics page on Encyclopedia Brittanica’s website does a much better job of explaining the procedure vs Wikipedia. 

Sonogenetics: Activating neurons using ultrasound. What I am not sure about is if sonogenetics is the non-invasive equivalent to optogenetics. 

There is clinical and also entrepreneurial activity in the domain of brain machine interfaces. But that’s a different set of layers. Besides, I am not at all familiar with the work occurring in these domains. Not even to the extent of reading up on the wikipedia page/company profile and/or news articles. 

What we do know about memories, right now, is that there is a translation/transfer/(insert a function that has yet to be identified) of sorts. One of the functions that the hippocampus performs is to consolidate short-term memories into long-term memories.

How all of this works, with respect to their being this overall cognitive system and the individual memories influencing/impacting the various sub-systems of the schema. We’ve yet to figure out all of this and more. 

As an uneducated observer, it is my guess that breakthroughs in the domain of memory formation and manipulation could be very advantageous for the human species. As breakthroughs in these domains may lead us towards understanding: 

How the process of inserting/deleting/swapping memories takes place.

Manipulating memories: Changing association of/changing the actual memory itself, linking and merging it with other memories.

Core breakthroughs in these areas alone could completely alter how we provision our healthcare systems today, how learning occurs (specially if knowledge is encoded in logic, or most of it anyways). As well, we may also get to a reality whereby we are truly able to ascertain/quantify intent. 

Now it could be that there is a universal (universal in a terrestrial sense; not panspermia sense) mechanism for writing memories. Realizing the fact that this (writing memories) itself is not a single step process. Meaning:

First some stimuli is observed in a conscious state. This stimuli is observed because some pattern or set of patterns kick in. 

A short term memory is generated. 

I am not sure if that short term memory is routed via the hippocampus right away or if there is a delay of some sort.

The neocortex processes the information. So that’s another set of patterns kicking in. 

Then a decision is made to engage another part of the brain. 

So, contingent on the type/kind of the stimuli, a different set of routes (not unlike the route identified above) would have been engaged. 

What I am interested in: What are the cellular (neuronal) correlates that drive the process of memory formation. Because there must be a series of processes kicking in, in order for the memory to be written to that cell. 

Just understanding this process alone could usher so many breakthroughs for our species and could potentially lead to many s-curves. We’d then have to think about how to time these breakthroughs and pick and choose the ideal set of realities that we’d want to live in. 

Here is what I mean by this:

We create institutions whereby the exchange of knowledge occurs. We train other humans via a variety of methods. 

Now there probably isn’t a schema on this planet (or the wider multi-verse at that) that is truly unique. But, again, the knowledge is being transferred with actual information being written to individual neurons. 

If there was a mechanism via which we could read specific information from a brain (individual cells) in a completely safe and reliable way (As well a transparent and ethical sense). And then replicate this knowledge in another brain. Then the means to be able to replicate skills would only be constrained by transfer rate  input/output (I/O).

What I am not sure about is if neurons have classes/library/functions and sub-functions. Or some means to be able to relate a block of information stored in the brain to a set of variables. Because, and as it relates to the scenario that has been described above: We’d be looking at transferring a selective block of knowledge from one brain to another. Not replicating the entire brain. (That’s another conversation regarding mind-files). 

But some kind of linkage must exist. Because if I am coding or thinking about coding, then the series of thoughts that emerge in my mind are linked. Sure, the mind may experience some distractions. But, overall, the set of memories invoked would be linked. But during their formation the memories were powered atop repetition. Which, again makes the process of extracting this information a bit tricky. 

Unless there was some prosthetic, that would precisely:

Capture/classify the set of stimuli being received

Marry step #1 above with enriching the set of memories above. Taking a page from (existing state) basics of cognitive computing today. 

Then measuring where the long-term memory ought to reside. 

And time-stamping all of these interactions (Smaller timescales) and sub-interactions.

So this would be a (one) mechanism via which a sytematic means to be able to encode and decode memories for recall and replication could be devised. To recap, you come up with a mechanism to direct how information ought to be enriched when it is written to cells. So that, retrieving this very information can be useful. 

This is a very very rudimentary framework (as theoretical as it may be) for transferring targeted portions of information (related to a subject or topic) from one brain to another. 

Even an uneducated individual like myself can appreciate that there are many problems to solve and as it relates to this vision to come to reality. (In addition to the problem of encoding and decoding memories itself). If the prosthetic ought to exists, then how would it regulate where to write this information and how best to write it in a redundant fashion. How would the prosthesis know which cells to write the information to in the first place, because it could be that there is already some pre-existing information stored at that location (Unless an agent is born with this prosthethic? and that all information from birth has been routed through the prosthesis)

Also, this may or may not be on topic, but there must be a series of patterns that link information in our minds. I was thinking about this when I was writing down the paragraph above about linking information and started thinking about the Jennifer Anniston neuron. Also referred to as the grandmother cell. These memories are connected to each other. Because, in our minds we equate Jennifer Anniston to terms (memory modules) like:

Movie star/celebrity.

Female. 

Blonde hair. 

Still looks super young at age 50. 

And other attributes like that. 

So the brain has made all these associations and all of this information is linked. And even though one set of neurons are actuated upon being presented a type of stimuli (name/image e.t.c). The fact remains that this information is linked to other concepts in the mind. 

I may come back and refine this blogpost as more information comes my way. But, to date, all of my experiences have led me to believe that memories play a pivotal role and as it relates to how our psychological schema is architected/sustained/enriched. 

There is also a miniaturizing angle to this conversation. A cyborg reality.

As with all technological breakthroughs, we will have many doors in front of us. Through each decision we will have the capacity to do enormous good. Boost scientific and technological growth and leverage that growth in order to help reduce human and also animal suffering. Eliminate most mental health and addiction related conditions. Step through other doors and in the wrong hands these breakthroughs (as hypothetical as they may be) could enable the worst kind of tyranny. 

The conversation about advanced computation engines that compute the probability of intents manifesting in certain ways. Those probabilities are predicated upon leveraging intents that are based on who’s system/values we are looking through. But, that, again, is another conversation.