Consider the billion-dollar debacle in the Nevada thermal solar power experiment. Now I say experiment because it did fail. Could it have been foreseen that complexity in so many areas foretold failure? So many of the technologies incorporated were unproven and pure experimentation with such a colossal investment that even the financing was experimental.
This is what humans do. They run headlong into catastrophe with little forethought and after-the-fact just say oops, it probably was a bad idea.
Artificial intelligence is going to be as dumb as its creators. Perihelion Engineering is that with each pass through the perihelion you have a new day and new vision having learned from your mistakes.
Stupid is defined as could learn but refuses. To foresee is virtuous. Solar power will show that maintenance costs will trump any short-term energy gains and will fail in the long term.
Prophecy, perhaps. I hope I live long enough to see the result.