Practically we have no control over earthquakes (except maybe with fracking) and no way to really forecast them.
Another idea I wish would work is getting energy from lightning strikes. I guess it's not practical, but if it could work, "powered by lightning" sure would sound awesome. :-)
The author has a reasonable idea in applying survival analysis to estimate the probability without making any assumptions about the underlying model.
If you roll a dice 10 times and don't get a six, the probability of rolling a six is still 1/6 for the next roll. However, as the time since the last earthquake increases, not only does the probability of one increase (unlike dice which has a constant probability) but also the likely magnitude of it increases as well.
The equivalent with dice would be you roll a non-six 10 times in a row, and instead of 1/6 probability of rolling a six, you now have a 1/3 probability of rolling a six and a 1/100 probability of setting off a nuclear explosion.
It wasn't so much the probability model (independent or otherwise) I intended to question (hence my very poor choice of the die rolling analogy), but the choice of describing an event as 'overdue'.
At the risk of introducing another down-voted analogy, presumably hurricanes have a probability of occurring based on things like ocean temperature, time of year etc., but I don't think I've ever heard the phrase 'the east coast of the US is overdue a hurricane'.