I think the best way to try to predict the future is to look for historical analogies in the past. Sure, sometimes something totally new and unprecedented happens. But usually there is SOME analogy -- like something else that was totally new and unprecedented in a similar way. :-)
I've heard lots of dire warnings lately about how AI is going to destroy jobs and create massive unemployment. I doubt it. I saw a video a few years ago that warned that "in the next 5 years 60% of workers will be unemployed". Their proof was to list a bunch of jobs that they said would be replaced by robots or computers in the next few years. Like they predicted that truck drivers and taxi drivers would all lose their jobs to self-driving cars. Etc. They added up all these jobs that would be destroyed and it came to 60% of workers.
Well, according to the US Census, in 1850 over half of Americans were farmers. The 1860 census was the first time that farmers were less than 50%. Today they're about 3%. So does that explain why today we are struggling with 50% unemployment, because half the people lost their jobs to improved agricultural technology and so are now permanently unemployed? Except that's not what happened, is it? Someone who loses his job to new technology is not permanently unemployed. Most find some other job sooner or later. And their children and grandchildren are certainly not doomed to lifetime unemployment just because granddad lost his job.
The people who made that video were wildly overoptimistic about new technology. I think I can very safely say that self-driving cars (to use that as an example) are not going to put all the professional drivers in the world out of work in another 2 or 3 years. (They said 5 years and that was 2 or 3 years ago.) In real life, it's rare for a new invention to come along and be in mass production and universal use within a couple of years. These things take time.
And that's not a trivial side point. The fact that it takes time means that people have time to adapt. In real life, if a 100% practical self-driving car was invented tomorrow, that was better in every conceivable way over a human-driven car, it would still take some time before the inventors could turn it from a working prototype into something that can actually be produced at reasonable cost. It would take time to get government regulators to approve it. It would take time for factories to be built or reconfigured to produce them. And it would take time for the newly-built vehicles to replace ALL the current vehicles. (I haven't read anything about self-driving cars in a while. Maybe there is a truly practical prototype available now. But if so, I haven't seen any of them on the road in my neighborhood.)
In practice, self-driving cars will probably never replace all human-driven cars. Just like cars today have not 100% replaced horses. There are some specialized jobs for which a horse is better than a car.
In real life, new technology does not destroy jobs. It enables a smaller number of people to do the same amount of work. And it frees people to do things that are more productive.
That same video I mentioned tried to brush off the obvious comparison to the industrial revolution by saying that then people were able to move from manual labor jobs to more creative jobs, but with the AI revolution, there will be nowhere left for people to go. They solemnly intoned, "We cannot have a poetry economy."
But, umm, yes we can. Even if someday robots, computers, and AI get so advanced that all but the most creative tasks can be done by machine, and the only thing left that machines cannot do is write poetry, paint paintings, write novels, and the like, then ... that's what most people will do. And realistically, it will be millennia before we get to that point, if we ever do. We'll always need human beings to manage the machines, to invent new machines, to maintain the machines, etc. I don't think that the fact that I have a computer makes my job border on the superfluous. Yes, the computer is a big help in writing this post, I'd much rather write it with full word-processing capabilities than be typing it on a typewriter, or chiseling it into stone. But the computer isn't going to write the post for me.
** Reply to SurpriseDog **
I am 100% certain that any technology that resembles current robots and computers is not creative.
Yes, I've seen plenty of golly-wow articles like the one you cite. To say that someone programmed a robot to paint a painting does not make the robot creative. I'm a software developer. I can easily program a computer to write a poem. In the simplest sense, I could type a poem into the computer and have the computer type it back out. I don't suppose that you would say that "that poem was written by a computer" in any real sense. In a broader sense, I could make up rules for how the computer puts words together. I've written a few programs like that just for fun. I mean, I've written programs that string words together that almost sort of sound like they mean something, but which are really a lot of gibberish. Years ago I read an article in a computer journal about a program that took words from published articles in scientific journals, rearranged and strung them together in various ways, and spit out a fake "scientific journal article". The programmers then submitted these article to a bunch of journals and several of them were actually accepted for publication! The programmers thought it was hysterical.
To the extent that such a program produces interesting results, the creativity is again in the programmer who wrote the rules. The computer is not really inventing anything. Just like, if I type up a poem using an old-fashioned typewriter, you wouldn't say that the typewriter wrote the poem. You'd say that I wrote the poem using a typewriter. Indeed it's just like, if a carpenter uses a power saw and a drill to make a bookcase, you wouldn't say that the saw and the drill made the bookcase. You'd say that the carpenter made a bookcase using a saw and a drill. The computer is just a tool that the writer uses to help him express his creativity.
I don't know just what the programming is of this robot described in the article. Programming a computer to create abstract images would be trivial. I could do that easily. Just throw some random colors at random places on a canvas. I presume you'd want to put some constraints and rules on it so there's some pattern to it, but that would be easy. Programming a computer to take a photograph and make semi-random changes to colors and patterns would be more difficult but not fundamentally ground-breaking. I'm sure I could write something that would give amusing results with such a process if I worked on it for a few weeks or maybe months.
Building a robot capable of holding a brush and painting on canvas no doubt presented some interesting technical challenges, but that has nothing to do with the robot being creative. Just like programming a self-driving car to stay in its lane presents interesting technical challenges. Getting the robot to handle the brush and actually transfer the image from an internal JPEG file to paint on canvas would be the challenging part of this project. This would probably be useful to advance the technology of robots in general, but of itself of little value: a laser printer accomplishes the same thing with much simpler technology.
At some point could one say that, while the computer was built and programmed by human beings, it is now an independent intelligence exercising creativity? Maybe some day people will invent some true artificial intelligence, some machine that could legitimately be said to be a living thing or an independent mind, or at least that would raise serious questions about that. But if so, that would not be some extrapolation of present technology. It would be a very different thing.