AI is a forever-in-the-future technology. When I was in school, fuzzy logic controllers were an active area of “AI” research. Now they are everywhere and you’d be laughed at for calling them AI.
The thing is, as soon as AI researchers solve a problem, that solution no longer counts as AI. Somehow it’s suddenly statistics or “just if-then statements”, as though using those techniques makes something not artificial intelligence.
For context, I’m of the opinion that my washing machine - which uses sensors and fuzzy logic to determine when to shut off - is a robot containing AI. It contains sensors, makes judgements based on its understanding of “the world” and then takes actions to achieve its goals. Insofar as it can “want” anything, it wants to separate the small masses from the large masses inside itself and does its best to make that happen. As tech goes, it’s not sexy, it’s very single purpose and I’m not really worried that it’s gonna go rogue.
We are surrounded by (boring) robots all day long. Robots that help us control our cars and do our laundry. Not to mention all the intelligent, disembodied agents that do things like organize our email, play games with us, and make trillions of little decisions that affect our lives in ways large and small.
Somehow, though, once the mystery has yielded to math, society doesn’t believe these decision-making machines are AI any longer.