In 2023, Douglas Hofstadter (sometimes referred to as the patron saint of GOFAI), did this interview. Near the end, he describes what it’s like when some of your core beliefs begin to collapse. However, you need to learn, embrace, and internalize those beliefs before the collapse can be appreciated or even noticed.
One of the main knocks against modern A.I. is its crazy power and water consumption. Vast GPU farms suck energy at a rate that compares with some entire small countries. GOFAI is attractive for this issue alone. Example 1
In 2011, I wrote an article on Future Psychohistory. I based it on post-GOFAI technology (mainly Neural Networks). I later realized that a better fit was Agent Based Models. These in turn, were better suited to GOFAI languages such as PROLOG. I’ve since been working on a hybrid model, where the individual agents are written in PROLOG and the emergent ‘world’ they inhabit is realized in FORTH. This brings me full-circle back to 1980s GOFAI.
Many are predicting the collapse of Generative LLM style A.I. which has been over-hyped and under-delivered. GOFAI may yet see a second spring.
That book captures the wonder of that era. I didn’t work at Data General, but I am certainly familiar with those times. Tracy Kidder really did a nice job as an embedded reporter.