What is there to learn from an old book? A book that is little more than an old technology review at that. A book that is a snapshot of the state of the art of Artificial Intelligence in 1986 (or rather a book published in 1986 about the state of the art circa 1984 - the publishing cycle was much longer to get to print then)? Technology was still pre-Internet. A "US based" state of the art, so Europe (where I studied in 1986-1991, BSc and MSc) in effect would have an information lag of two to three years to catch up. It was a time when the US was running scared of the Japanese Fifth Generation Computing initiative and the US Congress acted to mobilises "a whole of" US academic and industry sector response to what was perceived as an existential threat. Does that sound familiar? So familiar as what happened next was an AI Winter (the second) because of the stinging nettles of over inflated hype and unrealistic market expectations. That was a winter I was caught up in as a graduate student, but took it on the chin and just bumbled off into mainstream commercial computing.
The book is fascinating as it illuminates characters and personalities but shows the still unfulfilled dreams of Artificial Intelligence, from the mouths of the original pioneers who were at, or were taught by the delegates of the infamous 1956 Dartmouth Conference. Marvin Minsky's goal of that summer camp being to crack significant fields of AI (which still as to this day remain uncracked) within six weeks .. the craziness and the optimism of it. The field of Artificial Intelligence (although nobody was readily recognising it as a "field" in its own right at the time) was all things to all men, half the delegates did not even like the title.
Marvin Minsky passed away in 2017. Late on in life, when asked of what he thought of today's take on AI, he remarked, "We seemed to be closer to success in the 1970's", which I think is a reflection on how much we have learned, that there is still so much yet to be learned, including many more "unknown unknowns". The book is called "Machinery of the Mind, Inside the new science of artificial intelligence" and I wish I could have somehow read it thirty four years ago (see below, written by George Jonson, who was given the chance to mingle with the AI scientists through means of an Alicia Patterson Foundation journalism fellowship in 1984):
Note: I have literally destroyed the book in act of reading it [sins of my marker pen highlighting and scribbling quick annotations (filling in with hindsight developments to know to have happened next) and there is also the age of the binding - the glue cracking as I turned its pages].
It is a good book, as it tells you as much about the "people" as the "things" they were trying to create and I learned a lot of stuff I had never dreamed of was actually happening, cool stuff. The funny, the bizarre and the quite, quite sad. There were many winners and losers, famous names galore from the annals of AI. There was also a tie into one of my long running science fiction wargaming projects, that of replicating Traveller's "The Trillion Credit Squadron" scenario, with Eurisko and Douglas Lenat's original fleet (Douglas Lenat sadly passed away in 2023). I came across the book because of an oblique reference to Douglas Lenat and TCS.
