Philip Tetlock is a professor of psychology and political science who set up a 20-year experiment that involved 300 experts in politics. He asked them to predict the future. And they didn't do very well. Those who assigned probabilities to rosey scenarios found out that those came true only 15% of the time. But the ones who assigned doomsday scenarios were only right 12% of the time.
For some reason I find some delight in learning that the doomsday-ers did worse, if only by a few points. It's comforting to know that, odds say, the negative and the dramatic are often incorrect. So it's probably good to use such doomsday speak as only a general guide for re-evaluation and prompting of further thinking.
In any event, the book Future Babble looks good.