Monday, March 27, 2017

It is a long journey - Cognitive Science to Cognitive Computing - old theories are still very profound

In late 80's - more precisely 1988-91 -  I spent quite a bit time to read and (trying to) understand the "science of cognition' - chasing questions like - what is "knowledge"? ....  what is "reasoning"?...    the puzzles of language, learning, memory, inference, problem solving, philosophy of life, human mind, etc., etc. This was the time when I had a free hand to pursue research on topics of my interest and was bold enough to propose a "Center for Cognitive Science"  together with a thinker researcher colleague Dr.  B. N. Nair at Center for Development of Advanced Computing (C-DAC) in Trivandrum, India.

In fact, it is very amusing that I stumbled on a hand-written journal in which I was noting down important points from some of the books and articles I read at that time (photocopying was expensive then in India  !!!).  When I browsed through them again after 25 years - a time when cognitive computing is hot and fashionable - I see very valuable nuggets in the very old references starting from Aristotle. I may share some of those here intermittently.

Among different topics from my old journal, one thing that looked profound is a definition of knowledge from "Sankhya" Philosophy (https://en.wikipedia.org/wiki/Samkhya)

"Knowledge is finding association about things. It is pigeon-holing one experience with the already existing fund of experience and this is one the great proofs of the fact that you can not have any knowledge until you have already a fund of knowledge in existence. There must be a store at hand to which to refer a new impression. Suppose a child is born into this world without such a fund, it would be impossible for that child ever to get any knowledge. Therefore, the child must have been in a state in which he/she had a fund and so, knowledge is eternally increasing"

Very thought provoking!!!  Need to digest it more even after I am reading these notes from many years ago!

No comments:

Post a Comment