Every time I take a class, I semi-consciously pick up its jargon and use it for all kinds of unrelated things. I'm aware that this is very common among nerds/hackers -- after all, I hang out with lots of them. I, too, speak of "pinging" people in real life, and of the "failure modes" of couches and suchlike. But because I'm a biologist hanging out with mostly non-biologists, it stands out a lot more because everyone else isn't using jargon from their biochemistry classes. I don't know many other people who use words like "inhibit" and "saturation" and "depletion" and "steady state" and "modularity" on a daily-to-hourly basis. (OK, maybe the last two are more widespread than I think, and I just need to hang out with more MechE or EE people.)
The most recent one is "timescale", or "on the timescale of". I picked this one up from my biomechanics class, which analyzes bio-materials of all different sizes from single molecules to whole organs. In order to keep ourselves sane, we have to take into account the size of the object in question when choosing an analysis method. Should we speak of the stresses and strains in a rod made of continuous material, or of the entropy-driven behavior of a randomly meandering chain? Can we ignore thermal motion of molecules, or the transient behavior when you begin applying force? It all depends on the length scale.
I find the word "timescale" very useful in my daily life. It's much easier to say exactly what I mean if I say "on the timescale of weeks" rather than "in the medium-term". I would love to say that it helps other people understand, as well, but unfortunately no one else seems to have picked up on it yet, so I will have to wait and see.
Strangely enough, the jargon-adapting habit seems to be largely involuntary. However, the success or failure of a given word is definitely related to its usefulness, to the usefulness of the metaphor. There's not that much difference between a feedback system in a cafeteria and a piece of complex software, so it makes sense to speak of both of them as having "failure modes". I guess this is what the "seeing-as" theory of intelligence is all about. (Something I read in one of Hofstadter's books... I don't remember which it was, and I don't know what this is all about.)
Addendum: jargon-adapting is also not particularly widespread among people who aren't part of hacker or twinkie social circles. Or, at least, I don't encounter it very much, and I often get laughed at (in a kind way) when I'm hanging out with my friends from Bioengineering and I speak of rainwater "saturating" a drain, thus forming a puddle.