“The next 50 years: A personal view” [of AI] by Patrick Henry Winston

Finding Judea Pearl's text "Causality" sitting atop the reading stack, I began looking for some way to avoid reading it. So I was delighted to pick up this older (2012) but extremely timely paper by Patrick Henry Winston, which gives his take on several topics:

- the retreat of AI to statistics,

- Brooks subsumption architecture and some pitfalls,

- the limited efficacy of neural networks and

- even a brief discussion of Pearl's previous book("Probabilistic Reasoning in Intelligent Systems") and work, which provided just enough information to send me in a new direction!