So having a simplified explanation for dreary, low-information noise (“they’re all NPCs”) is actually the only way to combine curiosity and disciplinehttps://twitter.com/ashwinning/status/1057775541477089280 …
-
-
(so same info, different eval) - the AI agent isn’t using any innate heuristics so he wouldn’t miss hidden info in a screen but neither would he know to avoid this
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
that's a good point, and it feels like a cop-out to just say "everything is noise without imposing a representation on it"
-
in what sense (maybe it sounds better in german - “content without concepts are blind, concepts without content are empty”)
-
if i head over to oeis and pick a random integer sequence to show someone they might see it as just picking numbers out of a hat
-
and depending on context that might be effectively correct or it might be totally wrong
-
yeah, which might be where i'm going with what representations are useful to put on things? i'm not quite sure, i'm just thinking
-
the problem is that a lot of important work has been done on this separately in philosophy, in info theory, and in ev bio, and integrating it all is a bitch and a half
-
for example millikan’s work on biosemantics and “Normal” contexts might really interest you but I can’t promise it wouldn’t seem impossible to translate into language of AI
-
i'll check it out, normal contexts rings a bell
- 3 more replies
New conversation -
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.