Yes cortical columns, general representations for objects in the real world.
Just a representation space for a feature, a cat rear, side view, front view,
bottom view, and overview all light up 'cat node'. Just a representation space
for activation. A general robust feature node that will recognize cat, no
matter the angle, lighting, species, size, rotation, etc. Hence the
representation space. Of course the brain may simply work with any lighting
because it uses shape outlines instead! So it uses a certain representation
space that can be activated. Like Word2Vec / Glove, the text word 'cat' has an
ex. 1000 dimensional space, it is related to all other words some certain
amount of probability, and so to activate it you can see other words that are
very related instead, and so multiple different words will light it up ex. cat
overview/ rear view/ cat front view = cat, tiger, dog, kitten, animal, lurks at
night, eats, run. The point is certain things light up that node, and word2vec
could accomplish just that. Even thought it never seen cat=kitten it can
recognize they are in almost same dimensional space!
------------------------------------------
Artificial General Intelligence List: AGI
Permalink:
https://agi.topicbox.com/groups/agi/T1a1cab406375d46f-Mda115231c7495a8385b6128a
Delivery options: https://agi.topicbox.com/groups/agi/subscription
