I used a Cray Y-MP back then but you're still right... My iPhone is faster than that hardware now, which is pretty neat.
It wasn't so much "thumb twiddling" though, there was a lot of work being done on systems which were more focused on knowledge representation (like Cyc [1], which still exists). Also a lot of work was being done from a more Psychological direction (mental models, scripts etc) and from a physical/neuro-science direction (brainz!).
These were all happening simultaneously and it wasn't clear (partly because of the MIPS issue you mention) that ML was the winning pony (for now) and I still appreciate the broad spectrum of knowledge covered in my particular Cognitive Science program.
It wasn't so much "thumb twiddling" though, there was a lot of work being done on systems which were more focused on knowledge representation (like Cyc [1], which still exists). Also a lot of work was being done from a more Psychological direction (mental models, scripts etc) and from a physical/neuro-science direction (brainz!).
These were all happening simultaneously and it wasn't clear (partly because of the MIPS issue you mention) that ML was the winning pony (for now) and I still appreciate the broad spectrum of knowledge covered in my particular Cognitive Science program.
[1] https://en.wikipedia.org/wiki/Cyc