The discreteness of language seems to stem from brains seeking a single robust interpretation of the world. For sure humans would encode continuous variables in continuous outputs (e.g. pitch), if they could reliably interpret them, but the channel is just too noisy.
-
-
-
Our perception is stable but continuous. The discreteness of language may be primarily a learnability constraint for shared protocols.
- 11 more replies
New conversation -
-
-
How do we transition from one form of language to another? Or, Will machines eventually develop language and protocols for social interactions just like we humans did.
-
Language is used to manipulate concepts, which are in turn evoking compositional simulations, which can be re-translated into a different language. I don't think that AI needs to be social, but if it has to negotiate with other intelligences it will come up with language.
- 3 more replies
New conversation -
-
-
Closures, better relative pointers and loop constructs, better spatial, temporal and hierarchical references, distributions, mnemonics for numbers and paths, ...
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Yes, as a child, and later I looked at synthetic natural languages (Ithkuil etc.) and tried to find improvements. I still think that a designer language could improve our cognitive abilities, but eventually this is just the scripting layer and the hard work is done below that.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Parts-of-speech are the standard function arguments. A lot of the work in phrasing a thought comes in arranging, tagging & reformulating words to better serve their part-of-speech roles. I loved the cleanness of Esperanto (o-nouns, a-adj, e-adv..), like going from CXXIII to 123.
-
Esperanto is the Dvorak keyboard among languages. 20% better, but that is not enough to give sufficient reason for adoption except among nerds...
End of conversation
New conversation -
-
-
I think natural language reduces perceptual content to low dimensional and discrete perceivables, but math/programming language actually inflates low dimensional perceptual samples (like counting) to a high d intellectual 'perception'. Functions may not be ontological.
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.