It was recommended reading in my undergraduate mathematics class in 1980, we had a lot of serious math texts on the core list and a wealth of periphiral material that pushed various edges.
The rationale was that "this is probably bullshit, but it's exactly the kind of BS that you need to read closely and argue for or against various aspects".
At the time, of course, fractals were vogue, AI was just around the corner, and we had exercises on robotic algorithms, speech recognition, distributed text searching and indexing, read on neural nets and so on.
It was a great read - I never bought the 'recent' evolutionary arguments but I enjoyed the discussions of what constituted 'intelligence', a section still worth reading today.