Solving the Problem of Bias in Artificial Intelligence