While writing software and endlessly refactoring to reach the ‘ultimate’ structure, one finds oneself often wondering that truly Dawkins’ ‘blind watchmaker’ metaphor must be correct, that competence without knowledge is a fact of life, and no amount of speculation on agency, mind and creation can cancel the simple truth that order can emerge from disordered systems via an algorithmic process which is only local in scope, has no global view of the domain it lives in.
The question appears to be the following about Deep Mind and General AI:
are those attempt to getting closer to a global optimization process? Comprehension (Verstehen) as global optimization?
Leave a Reply