The last invention.

  • 388 Replies
  • 55820 Views
*

Korrelan

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1130
  • Look into my eyes! WOAH!
    • YouTube
Re: The last invention.
« Reply #330 on: June 27, 2018, 09:40:13 am »
Quote
But is sigmoid what humans do?

There has obviously been a lot of research into the human neuron over the years. 

Researchers look at the structure and electrochemical properties and try to derive function; the weighted/ sigmoid neuron is just one interpretation of the human neurons schema.

Some people will argue that the human neuron must use something a kin to the sigmoid function because the whole weighted/ sigmoid schema ‘kinda’ works.

I personally don’t think the human neuron uses anything like a sigmoid function.  It’s just a tool, algorithm that was employed to make the classical neuron mathematical model work.

I stopped using non-linear activation functions/ back-propagation many, many years ago lol.

https://en.wikipedia.org/wiki/Neural_coding#Temporal_coding

The human neuron is not just about ‘what’… but ‘when’.

 :)
« Last Edit: June 27, 2018, 10:12:48 pm by korrelan »
It thunk... therefore it is!... my project page.

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #331 on: July 12, 2018, 11:09:58 am »
Korrelan, can you explain how I am able to link together 2 memories like "buildings" and "snakes bit" almost instantly when the neurons can't connect that fast? I thought neuron's little feeler ends take eons to connect, little own to the right one....what is dis even idek.

#explaindatstuff
Emergent

*

Korrelan

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1130
  • Look into my eyes! WOAH!
    • YouTube
Re: The last invention.
« Reply #332 on: July 12, 2018, 01:02:26 pm »
Quote
Korrelan, can you explain how I am able to link together 2 memories like "buildings" and "snakes bit" almost instantly when the neurons can't connect that fast?

You are still thinking of information/ memories in a classic format… in a human language/ information exchange/ scientific notation/ serial format.  As words, nodes and links… Your brain does not store/ use information in this format.

I will try to explain… high dimensional memory/ space.

Let’s say you have a serial database list of all the people on the planet and you wanted to search for/ select just one person.  Using just a few search parameters like sex, age, nationality, etc would yield too many results, but the more search parameters you can apply the narrower the criteria and smaller the result list/ table. 

So you could imagine creating a set of search criteria that would always find just the person you are looking for.  What this method is leveraging is high dimensional space, the higher the number of dimensions/ fields the more focused the search results.

The problem is that the above example still uses a serial schema; items are checked one by one in the list until a record is found that matches all the criteria, and although the search can accurately find the person every time, the time it takes is linearly proportional to the amount of information being searched… this is bad lol.

Your brain also uses extremely high dimension space to store knowledge; in fact the number of dimensions is so high that any object/ concept can be encoded and the proportional time problem is avoided by accessing the information in parallel, so all the search parameters are applied at the same time, not in sequence/ serially.

A good analogy would be the difference between a picture and paragraph describing the picture.  When you read the paragraph you are absorbing information serially, the words slowly build up an image in your ‘minds eye’.  When you look at the picture ‘BOOM’ you are absorbing the information in parallel, all of the information is arriving at the same time and you instantly understand the picture… hence the idiom… a picture is worth a thousand words.

Ok so… the high dimensional space means that the memory recall is very accurate, and the parallelism means its extremely fast… so technically your brain never has to search for information… not in the classic sense of search anyway.  All your knowledge relevant to your current state of mind/ task is instantly available because the knowledge/ experience IS your current state of mind.

Also consider that we are not like classical computers, Information is not stored inside your brains memory; your brains connectome structure IS the memory/ information.  Your house wiring doesn’t need a memory/ logic to know what to do when a light switch is flipped; its operation is encoded in its structure. If you want an extra function from your house, if you want your house to ‘remember’ what this new switch does… you have to wire it accordingly.

Even the hinges on a door are providing a kind of function and memory, they allow the door to swing freely in two dimensions and not fall off, without the hinges the door could not function.
 
Remember my short video of the binary logic example, there is no program or traditional memory required for the neuromorphic neurons/ synapse/ etc to simulate a binary counter, the program and the logic are encoded in the structure.

https://www.youtube.com/watch?v=nqPKOrhZlAw

Quote
I thought neuron's little feeler ends take eons to connect, little own to the right one....what is dis even idek.

Synapses are only created when encoding new information, not when retrieving/ using current information.

I hope that made some sense lol.

 :)
It thunk... therefore it is!... my project page.

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #333 on: July 12, 2018, 01:19:10 pm »
A table with chairs, people, and lots of food.

Buffet.

A picture in 1 word lol.

Ok korr...that no answer question.......I know how we have a re-use hierarchy, but when I select a mouth and a nose and link em ex. "UK and teddy bears", the 2 things are linked now for life if I repeat them for a few minutes. How are they linked so fast? Does a connection poop out of a neuron soma and look for the other feature?
Emergent

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #334 on: July 12, 2018, 01:22:57 pm »
maybe all the layer nodes are already all connected to every, single, other node?
Emergent

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #335 on: July 12, 2018, 01:32:59 pm »
furthermore, there must be a new node in a deeper layer made that lnks to the two, else no new larger feature is made by the 2 smaller sentences!! And that's important to my work !
Emergent

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #336 on: July 12, 2018, 01:47:41 pm »
btw we scan images too, like

girl...holding turtle......eating..........wait for ittttt,,,,starts raining.........paper drawing gets ruined

It's a video sentence
Emergent

*

Korrelan

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1130
  • Look into my eyes! WOAH!
    • YouTube
Re: The last invention.
« Reply #337 on: July 12, 2018, 02:18:46 pm »
Quote
Ok korr...that no answer question

Yes I did... you asked...

Quote
Korrelan, can you explain how I am able to link together 2 memories like "buildings" and "snakes bit" almost instantly when the neurons can't connect that fast?

I explained high dimension memory, or the fact that neurons don't have to connect that fast... there already connected through high dimensional space.

Quote
Does a connection poop out of a neuron soma and look for the other feature?

Yes lock... why not... it's all done with pooping neurons.

For you to pick those two concepts ‘UK and teddy bears’ there must already be something subconsciously connecting them.  You are not able to think randomly, I know it feels like you can… but you can’t.   Something, some were in that noggin of yours linked the two together before they even popped into your consciousness.  It feels random because you can’t access your subconscious and see the connections.

Because information is stored in high dimensional space, and because all ‘memories’ use the exactly the same facets to provide the ‘details’, this combined with the various combinations through episodic memories etc, the connection could be very convoluted.

So in this instance you are just reinforcing an existing connection by mentally repeating it, which is further reinforced by the memory of the fact that you made a conscious effort to repeat and remember it.

And don’t say that you personally can think randomly… no you can’t.

 :)
It thunk... therefore it is!... my project page.

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #338 on: July 12, 2018, 02:24:05 pm »
ok

korrelan and murderer

:))))

there MUST be something linking them!!! :)))))
Emergent

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #339 on: July 12, 2018, 02:37:47 pm »
I could be the next physic.



Wait, before we reply anymore, let me just re-inforce that above in my brain.

:)))))))

korrelan isA murderer

korrelan isA murderer

korrelan isA murderer



Also, new update: The pink unicorns can indeed fly through the moon, where we live. I seen them with my own eyes. My leg actually jumped a little bit, I got to meet aliens, finally! I made some new friends, just last night, named sonica blue 4728 and vivian majasker 899.
Emergent

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #340 on: July 12, 2018, 02:49:12 pm »
Quote
For you to pick those two concepts ‘UK and teddy bears’ there must already be something subconsciously connecting them.

back to the drawing board ye go !
Emergent

*

Korrelan

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1130
  • Look into my eyes! WOAH!
    • YouTube
Re: The last invention.
« Reply #341 on: July 12, 2018, 03:02:14 pm »
Quote
korrelan and murderer

You have experienced the concept of both.
They are probably both human.
A human can be a murderer.
They both exist in your reality.
They are both sentient.
They both probably would have a body, arms, legs, etc.
Etc.. Etc…

You probably thought of korrelan, a human (honest) that fired all your underlying knowledge of humans and primed your brain for the next phase. What could I pick that would prove my point… got to make it seem random…ah I know… murderer.  There are so many deep sub conscious connections between those two concepts lol.  There could be any number of reasons/ explanations why you picked those two concepts.

You can’t include or use any knowledge within a thought that you have not experienced, and if you have experienced/ learned it then it has been integrated into your connectome/ high dimensional memory… and ‘linked’ to every other thought or memory that uses the same shared facets.  Linked/ recalled by facet… as well as complete/ partial memory engram.

BTW… I took the time/ effort to answer your questions, I would appreciate it if you could at least extend to me the same levels of courtesy and stop being a di*k, grow up.

 :)
It thunk... therefore it is!... my project page.

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #342 on: July 12, 2018, 03:22:42 pm »
I read all of your replies. I was simply asking questions/pointing out flaws, and with a touch of funny, I really lmao btw...

Well, true, any selected feature/memory is only selected if it self-ignites, or, is matched, or linked to one of those events. But maybe we can also do a random pick too?

I could pick korrelan, then read the morning news story and choose a word at random and link the 2.
Emergent

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • **************
  • Dot Matrix
  • *
  • 2306
  • First it wiggles, then it is rewarded.
Re: The last invention.
« Reply #343 on: July 12, 2018, 03:27:33 pm »
(we often do linking-throughs like animal>food>lunch>school>work>pay>job>cook>fun>etc)
Emergent

*

Korrelan

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1130
  • Look into my eyes! WOAH!
    • YouTube
Re: The last invention.
« Reply #344 on: July 12, 2018, 03:31:17 pm »
There is a big difference between thinking of a word, and choosing one from a newspaper.

One requires an internal 'thought' process, the latter uses an external source of data.

 :)

It thunk... therefore it is!... my project page.