Computer generated music wasn't that hard at the time the author's writing about. To be honest the process the author describes is just about the simplest there is in computer music: choose a sampler instrument, pick a sequence of notes.

Consider that Aphex's Windowlicker was a created at around the same time as the sound the author made, and it was an entirely computer sequenced track (with lots of effects programmed in the SuperCollider audio programming language):

http://www.youtube.com/watch?v=iZ8sZXFN6jc

That should give you some idea of the audio effects possible on computers at the time. The Tri-Tone is child's play by comparison.