[p1esk]

I think he was complaining about the algos' decisions being hard to understand, not them being unfair. From your examples, it seems like we more or less can understand human decisions.