I found myself using it for encoding game videos of EVE Online, and had noticed that it got massively better quality at the same bitrate as compared to the more popular WMV9--at least on those videos, it could get 2-3x better compression at the same quality. So I promoted and used it, and eventually I decided to learn how it worked and see if I could make it better.

Since then, I've moved on from EVE Online; in fact, I would say the primary cause is I found that x264 development was more interesting (if college itself wasn't enough of a reason!) I still use it for all sorts of video compression, of course, and every once in a while I find myself quite literally improving it because it gave suboptimal results on some video I wanted to post online.

By the way, I didn't note it in my above post, but I'm a sophomore in college (CS major), which puts it quite into perspective, I think.

I always tell people that they don't necessarily need lots of background knowledge or experience to get involved in something as complex, as, say, video encoder development. Almost none of my prior knowledge was at all useful in learning how x264 worked. At most my linear algebra and Fourier knowledge from my education went to some use on a theoretical level, but that's about it, and it was only "nice to have"; I could have done as well without. It seems to be this way with most jobs I've taken, too; 90% of what you need to know ends up being learned on the job. What really matters is your ability to learn and adapt, not your prior knowledge base.