Abstract (from WI 2020 paper)
We examine the history of Artificial Intelligence, from its audacious beginnings to the current day. We argue that constraint programming (a) is the rightful heir and modern-day descendent of that early work and (b) offers a more stable and reliable platform for AI than deep machine learning.
We offer a tutorial on constraint programming solvers that should be accessible to most software developers. We show how constraint programming works, how to implement constraint programming in Python, and how to integrate a Python constraint-programming solver with other Python code.