This is a good combination of Science, engineering (computing) and potentially industry here.
I started watching this video by Jeff Hawkins on May 11, 2013. I was _so impressed_ by it, that I STOPPED the video about 1/3 of the way through and I started researching on my own.
I wanted to understand COMPLETELY what he had talked about so far and, well, the research took on a life of its own.
It was about 1.5 years later that I finally got around to finishing the video.
In the inbetween time, I stayed up until 5am, changing my sleep/work schedule and just following wherever the research led me, mostly in cognitive science, computational theory of mind, and then beyond that into alternate theories of self and such until I finally hit upon some decent conclusions I could stand behind.
Now here, he talks about using sparse data sets to model the brain and mimics similar facilities, although I finally hit a point where I could see some potential flaws in it becoming a perfect brain model.
Someone asked my conclusions on my commentary on the video.
While Id likely explain it a little differently now (my head was still in circuitry at the time for analogies), the ‘gist’ would be the same at this point which is… um… a year since I wrote this. I’d like to know your thoughts on “year ago Ken’s” conclusions:
=== cut here ===
a) Embodied Cognition explains best my beliefs. The brain is not a computer in the head/nor a telephone switchboard/ nor a database : We model our idea of what the brain is based on the latest technology.
And these are all good, effective models, but not “enough”
I think the brains functions can be somewhat mimicked using sparse data sets as representation of switches flipped on and off representing aspects of a “concept” or “idea” or “function”… that our short term memory is NOR and our long term is NAND – and short term pulls analogies in from long term (NAND is a serial rather than parallel circuit – hence pulling seemingly unrelated things when we try to remember stuff)
NOR is parallel; why we can MAKE so many diverse connections so quickly and trim selectively on the fly while decsion-making.
But again; these are computer models of the brain; our ACTUAL thinking appears to be trinary rather than binary; I don’t think there is a need to invoke quantum stuff to mirror the brain; trinary is enough.
But anyway; embodied cognition says that we are brains + bodies + environments together; none are separate from the others. They may interrelate to different degrees; but they’re all one thing.
I believe we have an error in thinking that goes WAY back to Aristotle; including the law of excluded middle; that contradictions are “bad things” and should be eliminated. But contradiction is a part of our thinking processes; so is perspective and subjectvitiy and emotion; Ignoring those things will get us pretty far but someday we’ll have to address subjectivity to have a complete “science of what it is to be human”.
b) We build our knowledge based on physical analogy; it happens so young that we don’t remember; but it’s the best way to explain how we learn so quickly; “this is a that, except these are the ways it is different”.
=== cut here ===
As you can see, I relied heavily on the circuitry model here. I value it because it’s a mimic-able system unlike purely theoretical models which sometimes have a little bit of a barrier for application.
I’d like to know your thoughts.