AI is artificial intelligence. It is intelligence that is artificial. When one plays tic-tac-toe and the computer never loses, the computer is acting intelligently, but it was programmed to do that, so the intelligence is artificial. Asking for "true" artificial intelligence is like asking for "real" non-dairy creamer.

It is the job of the AI researcher to figure out how to tell a machine to act intelligently. Speech recognition, imaging, and simulations are all areas that benefit from this.

The gap that we haven't yet been able to leap is teaching a computer about understanding. The computer may know it won at tic-tac-toe, but it doesn't understand why or how. A computer may play chess really well, but it has no outside concept of the game. The computer lacks true intelligence (not "true artificial intelligence"). The computer needs to understand that the opponent is playing to win; that the words someone uses mean something and have a particular syntax; that the objects it is identifying in an image are real, definable things. So far, it can't, and that is where you see the failure in AI, when in reality it is the failure for humans to understand themselves and their own thought processes.

I do agree that a lot of AI stuff is hype. It's insanity to believe that humankind is getting closer to creating thinking machines when our own understanding of the thought process is ill-defined. I think the field really needs to establish realistic goals before continuing. But what do I know? I'm just an undergrad.