[hn_throwaway_99]

Hah, such a great way to put it.

This is relevant to HN because I'm probably paraphrasing this incorrectly but pg has said the following about why it's hard to launch a startup: the vast majority of ideas that sound stupid are, in fact, stupid. The ones that sound like great ideas have most likely already been done. Thus, the startups that have huge growth potential tend to be the ones that sound stupid given the conventional wisdom (so unlikely to have been tried yet) but are, contrary to the norm, actually great ideas in disguise. These ideas are basically by definition very rare.