If the AI hype from 37 years ago is any indication (fuzzy logic on lisp machines, anyone?): Yeah. 99% of the buzzwords will be dropped because they're empty, and the remaining 1% useful ideas will vanish into the background and see continued use under new names, because they won't be sexy enough any more to be called AI.
I sat down a to a Sun workstation that was described as a "thin client". This was slicker than my Linux system at home (which I mostly ran as a terminal using Lynx or something to browse the web) and featured the Sun browser and blinding fast Internet. Since I was using dial up at home, this was pretty intense.
Yes, I believed the future was thin clients, web browsers for everything, and that hyper-connected devices were just around the corner. It seemed totally reasonable to me that every system would be like a Chromebook. You sit down, login, and your "system" is waiting for you just as you left it the last time you used it whether it was here, or across the world. I also imagined this experience working on your watch (I imagined a bigger watch - we had watches that you could watch TV on in the 80's - a kid in my school had one) or on what I imagined were inevitable table-top and wall sized monitors that I assumed would be ubiquitous in 10 years (surface hub comes to mind).
So this was a popular vision and one that is kind of realized today although not all the way and not quite in the way "we" (people I was reading, talking to, and hoping to become) imagined.
Frankly, idea of mainframe is much older, and never really appealed to me. I prefer a kind PCs where P is for 'personal'.
I wonder if that was a JavaStation?