The thing with pointing out "AI can't do X!" is that, if you keep refining X into something narrow and precise enough, you'll eventually cross a threshold where a realistic amount of engineering and training data make X possible.
-
-
Explains why so many chatbot projects end up implementing IF/THEN statements: super narrow requirements, no need for training an AI (on often non existent data) and trivial for humans to describe
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
-
Where is it from?
- Show replies
New conversation -
-
-
Agreed. And at least with respect to many professionals' judgments, the thing about "refining X into something narrow and precise" is that you'll probably find you've made so many compromises and simplifications that you've redefined (rather than automated) X.
-
In the end AI works with proxies and a proxy is not the same as whatever it proxies (and if it is, your AI may be memorising instead of ‘understanding’)
- Show replies
New conversation -
-
-
(biological organisms are inherently incapable of having 'little experience', subsisting as they do in an unbroken continuum of experience across all life of common origin and back to that origination.)
-
Well, yes, but it's severely compressed.
- Show replies
New conversation -
-
-
I suppose, then, the question becomes: How long until AI algorithms have proliferated and iterated to the point that they have been "taught to the test" for virtually every conceivable situation? A functional facsimile of general intelligence, even if technically it isn't?
-
i.e., the world is its chess match?
End of conversation
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.