.. what skills/traits allow a person to make such predictions with high accuracy?One thing is that I think you need a pretty wide set of priors–breadth. Stuff like history, anthropology, economics, the history of art. Lots of knowledge about human behavior, politics, culture, stuff like how emotions guide behavior, etc.
When I look at a typical STEM education, we deliberately don’t prioritize this stuff. We know lots of things about how electrons behave and which sorts of functions grow the fastest and how cellular mitosis works. Not as much about why empires fall, the role of greed in political revolutions, or the changing role of women over the last 500 years. I think this puts HNers (I think STEM people are probably overrepresented here) at a significant disadvantage at making these kinds of broad predictions.
The thing we do have going for us is our ability to understand the course technology is going to take: what’s possible, what will and won’t work, and why.
I also wonder whether the people you’re around influence your ability to predict what’s next. On one hand, it’s a well-established fact in social science that many social trends, at least in the US (things like marriage and divorce rates, educational trends, changing attitudes around dating, purchasing behaviors), start in the upper-middle classes, as they have the numbers (population) to make real differences in buying habits, politics, etc., whereas the rich have more money but much smaller population. On the other hand, the lower classes in the US vastly outnumber what I’d consider a typical HN reader. Something like 70% of US adults don’t even have a bachelor’s degree, and the US median income for an individual is around $40K. Keep that in mind as you think about this stuff.
.. One of the things I find most striking when watching old movies is the general attitude of people toward tech.
If you look at movies from the 70s and 80s, conspicuous display of tech was common. Look at stereo systems of the time, and how people treated mobile phones (they were huge and conspicuously displayed). This partially echoes the “machine age”  of the early 20th century, a a time when tech was seen as “modern” and a force for progress.
Whereas these days, we want things to be light, invisible, and out of the way. That’s a major change in attitude.
I actually feel we might see fewer “screens” in the next few years if the combination of voice and AI becomes powerful enough that most things can be done by voice or thought. I think more and more decision-making (things like which plane to book/flight to take/etc) will be made by automated systems that know our preferences and we’ll be picking from fewer and fewer menus. Sort of like a human assistant, but available to the masses and more accurate. Google’s Duplex is a big step in this direction. The key is ceding more decision-making authority to software.
In any case, I wouldn’t be surprised if we all just have earphones, either over-the-ear, or implanted in our heads, in 15 years. The broader theme is that I think we’ll want things to be invisible rather than visible.
I also think you’re right that the rich will want less of this stuff. There’s already a huge socioeconomic difference in how people use tech. Look at how a rich family eats in the US today vs. a poor family. Rich families put their phones away, poor families spend the entire dinner posting stuff on Snap. Just walk into a burger king vs. a fine dining restaurant to see that trend in action.