Since I'm organizing ALL my work for good I DID come across it in my notes and
YES, I actually do, have a winning node terminology that exists. When input
"cat" in text enters your brain it activates multiple nodes some amount ex.
cat, cats, case, dog. For prediction. But the prediction you hear, of all
candidates, you only hear one. So although many nodes get lighten up, only one
is heard and pools there. You feel only 1 heard because it pools there, which
obviously will affect prediction lots. But then again, words heard light up
related words for 2 reasons, not just syntactical words. Though maybe the
future candidates syntactically don't stay activated because imagine me
predicting "i went to the river and the doctor..." simply because the node
hospital was activated when i reached "i went to the ?hospital?". So maybe
those shouldn't stay activated? Of course speaking predictions only says the
one it pools to, so, if you predict cat>ate/ran 70% and 30% likely next word,
and 30% of the time say using mouth motors "ran", why, ate is 70%, for the
times it does say 'ran' it must get more energy sometimes. So, it changes where
it pools. But this would be same for similar words, cat sometimes picks dog/
zebra/ etc. This will need to be understood better hmm.
------------------------------------------
Artificial General Intelligence List: AGI
Permalink:
https://agi.topicbox.com/groups/agi/Tcf8fbceae6bd0acd-M51b289a2cb4588561b848458
Delivery options: https://agi.topicbox.com/groups/agi/subscription
