Artificial Intelligence? I think the phenomenon is not only very much overblown and it is also not new. What is new is the “Everybody Knows” fear-mongering that AI is taking over.
All computers can do is choose from among the choices a human or group of humans have provided. They may do so more quickly, sorting and choosing from a large number of options, but they don’t create, except under our direction.
Two quick examples: Does your car have an automatic transmission? It switches gears so the vehicle can speed up or slow down, climb hills or descend. You could do it yourself but possibly not as quickly or as smoothy. I think that’s an example of “artificial intelligence,” but it’s not new. Automatics began to appear in mass-produced automobiles in 1939.
How many minutes has it been since you cursed out the spellchecker on your phone or computer? These programs think they know what you meant to write or should have written and change your text, often resulting in embarrassment or gibberish.
Can AI create artwork of a pastoral day in the country? Yes, if someone programmed it that the sky should be blue, the grass and other vegetation green, and so on. Can it write text? Again, yes, if someone programmed it properly. Otherwise you might end up with a sterling endorsement, like the one that runs regularly on my Kindle: “Try to enjoy reading the poems.” Sorry, if I have to try to enjoy reading something that is for recreation, I’m going to skip it.
When “AI” beats you at chess, that isn’t artificial intelligence, it is good programming: Someone put in all the possible moves for all situations and the chances that the human opponent would do such-and-such as a result of a given move. College papers written by AI? Most professors probably would welcome the improvement! And what’s the difference between a heavily “photo-shopped” photograph and one created from scratch using an AI program? Where is the line?
One of my favorite Science Fiction authors, Glynn Stewart, writes about what he calls “artificial stupids:” Robots or holograms that make choices within severely limited parameters. I don’t think we have progressed beyond that: Humans set up the choices, even for sophisticated AI programs. I don’t see that changing any time soon.