> You think of n/2 of them in the first 5 minutes.
The word that leaps to mind here is "half". :)
The word that leaps to mind here is "half". :)
It could be (cumulative) exponential. Ideas would then be independent and random. That doesn't feel right, because ideas are very much related, but a better model doesn't immediately spring to mind.
Had he said half, I would never consider the distribution of ideas over time. And I bet pg considered the distribution while writing the essay, and he probably also realized it wasn't really all that interesting, and consequently left it out.
The first half of the items is actually an article of n/2 items - so you think of n/4 in the first 5 minutes.
So you never actually think of an item.
On a related note, here's how to abuse the concept of URL shorteners in order to produce messages in which embedding a URL costs zero additional characters:
http://www.byrnehobart.com/blog/steganographic-typo-based-ur...
Maybe use a magic pattern in the first few characters to indicate a url ID is present.
"It was so clearly a choice of doing good work xor being an insider that I was forced to see the distinction."