OK everyone. Let’s get a little perspective!
There is no “new” AI in concept… only in implementation.
I wrote my first computer program in 1968 – and though I did not know it, it was in some sense a program in artificial intelligence. On a, by today’s standards, ancient card-reading, IBM 1440 computer. My father was a superintendent of schools in Michigan and he setup a training course for some high school students in this new computer thing… and when asked to create a new program about anything I wanted, I wrote a program to solve a puzzle I had at home.
Wherever you look back then, as AI was just beginning, the image always was to create computer systems that could mimic human intelligence. It was never a question of what but how.
As an academic for many years, working in the auto industry consulting while teaching AI, Expert Systems, Systems Engineering, Statistics, Quality Control and many other things, I can tell you one thing. Whenever a young professor wants to clarify their own thinking (and make a name for themselves) – they invent some new words to mean the same thing that others have always meant.
It doesn’t really matter. As generations go by, rephrasing things for the next round of thinkers can be refreshing. But just look at science fiction – the stories of what AI computers / robots / automatons can do has not changed. We anthropomorphize and want the computers to think (sense, behave, remember, integrate,…) like we do and hopefully even better.
So Zadeh created fuzzy logic and fuzzy systems like that would solve all problems. Gubrud changed AI to AGI (Artificial General Intelligence)… why?
It has never been a problem of what the ultimate goal is.
It has always been a problem of HOW to reach that goal.
And as a long time engineer in AI, what I want to address here in True AI are the realities and methods for achieving true artificial intelligence in the fuzzy, general sense that all researchers have always understood.