In the "old days", you would devote yourself to the study of a single system - a process, an organism or even a gene. You'd publish 20-30 papers entitled "Gene A in organism B encodes protein X involved with process Y" and eventually, be in the running for tenure. Bioinformatics means that we don't have to work that way anymore. I can look at archaea one day, humans the next. I can apply whatever computational methods I see fit and am able to learn to any dataset that takes my fancy. In short, I don't fit well into academia at all :)
Yeah, that sums up the situation pretty well.