On the other hand, flooding yourself with raw experiences as machine learning programs do can create theories of totally new shapes and sizes and types; what we may not have is the continuos path and bridge from old to new theories
These comments by Yann LeCun and Ali Rahimi are extremely thought provoking. LeCun's remarks remind me of the Tycho Brahe or Sloan Sky Survey approach - explore and observe, to help enable understanding.https://twitter.com/haldaume3/status/938478425777377280 …
-
-
-
If our goal is performance and understanding not explanation and interpretation, we need not seek theory ; not just theory that is learned but the theory of how to architect/train networks and other engines; networks that master complexity through search may not be explainable
- 1 more reply
New conversation -
-
-
Yep, but that is what debates are for. Talking a lill bit, provoking and changing opinions. I like that kind of debates especially from intellectuals, bc they often have a broad knowledge and thus interesting arguments.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Learning is understanding of experiences enough to be able to predict what will happen in similar but novel situations; while theory (current explanation) helps devise precise experiments to systematically add to understanding, experiments see(k) the world through colored glasses
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.