Evaluating beyond human analytic capacity
An efficient manager has a high evaluation quotient. A machine often has a better one in an increasing number of fields. The problem for a human is to understand the evaluation machine intelligence has produced.
Sometimes a human will say "that's a good machine thinking result" or "that's a bad result," without being able to explain why or determine whether there is a better solution.
Evaluation is one of the major keys to efficient decision-making in all fields: from chess, production management, rocket launching, and self-driving cars to data center calibration, software development, and airport schedules.
We'll explore a chess scenario to illustrate the limits of human evaluation.
Chess engines are not high-level deep learning-based software. They rely heavily on evaluations and calculations. They evaluate much better than humans, and there is a lot to learn from them. The question...
