[cmiles74]

I don't disagree. But I don't think we can judge the intent of people who are discriminating in a racist way, aside from simply asking them what the think their intent might have been. At that point it's very likely we'll be dealing with a rationalization or a half-truth because not only is there a stigma attached to racist behavior but in many cases it is illegal as well and there are other punishments to contend with.

If we need to accurately gauge their intent, that's not really possible. In the case of an algorithm we've divorced the process from the source of intent (the author of the algorithm), there is no intent to evaluate.