Long ago, I forget where, I saw a blog post that applied this to writing. It pointed out that if we model the quality of your writing as having a mean X and variance Y, then the only way to hit those unlikely exceptionally good texts is to write a lot. Yes, while doing so you might also come up with the same number of exceptionally bad texts, but nobody forces you to show those to anyone. Plus writing a lot will give you practice, gradually pushing up the mean.
From personal experience, I'd also err on the side of publishing even texts you're not personally all that impressed by. I've noticed that I'm relatively bad at estimating what's going to be popular. Some of my biggest hits have been blog posts I'd never have thought would be popular.
Hmm, also, writing can be an excuse for generating ideas! The obvious thing to do would be to wait for good ideas to come into your head, then write them up for the world to see. But in my experience, writing and having my writing read boosts my ego, which somehow encourages my subconscious to throw up ideas which can be written up to derive yet more ego boosts. It's a virtuous cycle. Which makes not writing because you can't think up any good ideas a vicious one.
Followup to: Don't Fear Failure
In the same theme as the last article, I think that failure is actually pretty important in learning. Rationality needs data, and trying is a good source of it.
When you're trying to do something new, you probably won't be able to do it right the first time. Even if you obsess over it. Jeff Atwood is a programmer who says Quantity Always Trumps Quality
The people who tried more did better, even though they failed more too. Of course you shouldn't try to fail, but you shouldn't let the fear of it stop you from tyring.
I wouldn't go as far as to say that quantity always trumps quality, but where the cost of failure is low lots of failures that you pay attention to is a pretty good way of learning. You should hold off on proposing solutions, but you also need to get around to actually trying the proposed solution.
I'm normed such that I'll spend more time talking about if something will work than trying it out to see if it works. The problem is that if you don't know about something already, your thoughts about what will work aren't going to be particularly accurate. Trying something will very conclusively demonstrate if something works or not.
Note:
I originally had this as part of Don't Fear Failure, but that post got too long.