Our AI-Lab at that time proposed the university to buy the CM-1 for frontier computing research.

Sadly central computing budgets back then were (and maybe still are) dominated by the physics department which decided on a boring old Cray X-MP instead.

that's cool. I am reading a Brian Eno biography now, he learned 1970ish to concentrate on processes and systems, not on end results- same vibe here, just make it work and the applications will follow?