So, let’s say you get wrapped up in some argument, maybe one about religion, or politics, or historical interpretation. You know that your input can’t be exactly the same as another person’s; you know that you will process the input at least a bit differently, and you know that your output will probably differ from that of another. That’s the nature of black boxes. We can’t see how they work.
In every argument black boxes operate beyond the knowledge of the opponents. We don’t fully know what’s going on inside the boxes. The output is obvious, however. So, maybe the way to tone down contention is to put a filter over the output side of the boxes. But that means an “after-the-fact” filter. Whatever goes on inside the black boxes has already gone on by the time of its emergence.
All right, then maybe the way to obtain some control is to filter the input side of the boxes. Because we don’t know exactly what goes on inside the boxes—we don’t fully understand and we haven’t even fully identified the mechanisms at work—we know that something unknown inside will somehow change the input. We might not be able to filter fully the output, but by filtering the input, we get a bit more control before it disappears inside the boxes.