We don't learn to manipulate data because it's an innate structure of minds. The "learning" is just pattern acquisition (not much difference between using a fork and navigating office politics, except for type of feedback loop).
Great questions. Would love to see folks like @NandoDF @egrefen @_rockt @AlanMackworth & many more weigh in. My position: capacity for acquiring a rule (operation over variables) has to be innate; some rules are innate, most are learned. Differentiability helps but doesn’t solve.https://twitter.com/beausievers/status/1071460348312469505 …
-
-
-
I think both these activities like on the same continuum and obviously the same "hardware" drives skill acquisition, but unquestionably social (professional) behaviour required acquiring and applying significantly higher order abstractions than those used to manipulate cutlery.
- 2 more replies
New conversation -
-
-
I found this paper by Bloch et al. to be fascinating. One could imagine that the "last consistent erosion" operator is implemented in wetware as a way to minimize complexity of representation, and the "rationality postulates" (section 6.3) emerge. https://arxiv.org/abs/1802.05142
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.