It was suggested to me, as someone who is interested in going into the field of AGI (artificial general intelligence) that I read up on epistemology (a field in which I am rather unfamiliar). Are there any good books on epistemology that would be most beneficial to me?
-
Pasquelli (I think?) has a new text about "Augmented Intelligence and its Traumas" that might be interesting here, though from the way you write about "epistemological" concerns I don't wonder if material on ontologies/knowledge representation for deep learning might be constructive in terms of technical background– Joseph Weissman ♦Jun 27, 2015 at 16:28
2 Answers
Because you are aiming at AI, I would like to recommend sources which combine classical epistemology with results from neuroscience.
Introduction:
- Paul Thagard: Mind. Introduction to Cognitive Science.
Fundamental textbook, challenging:
- Jaegwon Kim: Philosophy of mind.
Reader:
- William G. Lycan (Ed.): Mind and cognition. An anthology.
Classics from 1990:
- Margaret A. Boden (Ed.): The philosophy of artificial intelligence.
Knowledge is created by evolutionary processes: processes that involve producing variations on current knowledge and selecting among those variations. A lot of epistemology is implicitly or explicitly concerned with justification: proving ideas true or probably true, or more probably true than alternatives. This is a dead end: every argument makes assumptions and uses rules of inference that haven't been proven or shown probably true, so the same holds for the conclusions of those arguments. Our ideas are all just guesses that have survived criticism. This was explained by Karl Popper, see "Objective Knowledge" and "Realism and the Aim of Science."
Some of the implications of these ideas for issues like AGI have been considered in "The Beginning of Infinity" by David Deutsch, see especially chapters 4,6,9,13,15,16.