Well, I hope you see (on rereading) your previous post sort of forgot to make a point...
Beside the exact role of serendipitous error and the art of guessing, there are a few other things AI has problems with:
summarising (read a text and think of a good title etc.)
'gestalt'
innuendo (try sarcasm or irony on a bot and 99% of it will pass unnoticed)
On the other hand, quite a few things we thought humans had an exclusive to, such as musical creativity, turns out to be more mechanical than we thought previously, see cellular automata (IBM, Hofstadter)
the "guessing" part we have at some level solved in our AI designs, related to various Chaos theories and the last 2000 years of progress in math or so
But indeed you are right, making it summarize a text seems almost impossible.
Detecting sarcasm, yeah, i guess that it could be done but it would take tremendous time and testing so maybe we will see it in 10 years from now.
Todays AI is not intelligent, especially not the chatbots. Intelligence must come from within, not from other sources, because if that is the case then it is Knowledge (true or false, still knowledge)
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.