The process of training AI with a human evaluating outputs to better change the weights during Deep Learning.