Is important to point out that we utilise AI and genetic algorithms that simulate the natural selection for optimising the bots. Means that the risk reward is not something in our control and is identified by AI itself.
That is obviously nonsense.
First of all, "genetic algorithms that simulate the natural selection for optimising the bots" doesn't actually mean anything. This is just a bunch of fancy-sounding words that is meant to impress someone who doesn't know much about machine learning.
More importantly, you absolutely control any output from any model. When you are training a model, you are choosing the training data, the loss function, and any additional penalty loss that your model will attempt to minimize.