This is kind of close to a relatively new openai paper that uses annealing/genetic algorithms for making a great reinforcement learning algorithm. I do love th

This is kind of close to a relatively new openai paper that uses annealing/genetic algorithms for making a great reinforcement learning algorithm. I do love the idea of machines figuring out the simplest model with maximum accuracy. When we can do that across different domains using same algo then we can say we have figured out secrets of intelligence.