https://overcast.fm/+Ic2hwsH2U/1:10:49 … #AI
“You could build a mind that thought that 51 was a prime number but otherwise had no defect of its intelligence – if you knew what you were doing” —@ESYudkowsky
Is it possible build a mind able to learn but incapable of correcting this error? (Why?)
-
-
Do you mean perfect in the sense that the AI's mind will be able to create explanations about anything (sane universal explainer) but that it will always be liable to have flaws in their theories (nuts)?
-
Not quite — if I understand, the idea in the original podcast was: when we know much more than we do now, it will be possible to create an artificial mind that has *specific* arbitrary beliefs/values, including false ones, which will not be possible to get rid of by thinking.
End of conversation
New conversation -
-
-
Sorry, I did not read carefully enough. Eliezer, are you not that specifically perfected mind?
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
So it is sort of warning, is not it?
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.