(Originally posted here by @kottke, and sent via @dapostrophe)
What Feynman talks about here is a constant problem in science education - when do you 'black box' something to kids? That is, when do we stop explaining 'why' simply because it is just too complicated to give a really satisfactory answer.
We're all familiar with the situation where a kid asks "Why?" followed by "Why?" and an endless stream of "Why?s" until the parent says, "Because I said so."
People tend to want quick and definite answers. We don't deal with uncertainty very well; usually, it gives us a queasy feeling in our stomachs, or just outright frustrates us.
Ultimately, the questions of science are not fully answered. But that doesn't mean we don't know anything. A better answer to this looping "Why?" question is "I don't know. Nobody really knows. But people are trying to figure out. That's what makes us human."