Unless you're coding or stress-testing benchmarks, the "latest and greatest" usually won't change how you use AI.

These implementations are for demonstration purposes. They are less efficient than the implementations in the Python standard library.