I loved the Foundation series -- mostly because of how many sub-series were contained within it, and each one was seemingly independent from the rest, but how all of it was guided by Seldon's Psychohistory equation which even from its macroscopic perspective allowed for small irregularities that it predicted for.
Fantastic series and nothing has matched its scale for me since other than the Ancillary Series from Ann Leckie.
One thing I like about Asimov is that his books are genuinely timeless, they make sensible predictions about future (read: current) tech without delving into rigid implementation that would make a future reader (read: us) laugh at it's absurdity. E.g, the seamless blending of different virtual environments during a video call is something that I can totally see happening.