We have learned to use a lot of automatic processes to help real humans behind the screen see what is happening, be warned, and react. An example of this could be our anti-fraud systems, which are kind of like an AI system which reads a lot of data and (after a lot of trial and error) can understand what is happening and warn a person of our team to take a look at something that may be a fraud, kind of like the advanced credit card fraud prevention systems.