The future is sometimes surprising, sometimes very much so, and that is one big source of prediction error. It would take a transcendent genius to have imagined

The future is sometimes surprising, sometimes very much so, and that is one big source of prediction error. It would take a transcendent genius to have imagined the full potential of computers back in 1900, for example, let alone everything else. Additionally, predictions of the future often lean either toward optimistic or pessimistic, rather than in between. But this is more complicated than it seems because a big part of the underlying mechanism is comfort. People make optimistic predictions that seem comfortable to them, and they make pessimistic predictions based on what seems uncomfortable. But comfort is not immutable, it changes depending on familiarity. The internet is uncomfortable to some and comfortable to others, as was the automobile and the telephone. When trains and automobiles first came about they were scary to a lot of people, partly for legitimate reasons and partly for not. Eventually most people became comfortable with such things. But even so, there's no new technology that doesn't come with down-sides, and it can often be difficult to imagine people of the future accepting new kinds of problems, sometimes even very severe problems. Nothing is perfect though, and people often have a blind spot for the problems of the things they are currently familiar and comfortable with and they don't see how getting rid of those problems makes the new problems more than worth the trade. Imagining multiple cycles of trade-offs and trade-ups from technology to technology is a very difficult thing to do. And too, as you say, the impact of technology on society and then society on technology is even more difficult to grasp. This goes along with pg's essay about "what you can't say". If you can't say such things now you certainly can't talk about them in the context of the future.