Earthquake Predictions

Has anything ever been a more sure-fire way since the days of alchemy to be more reliably, predictably wrong?

We have here an article about how earthquakes hitting Southern California have been more frequent than previously suspected, and how the supposed 'Big One' could be right around the corner.

Okay, great. What is someone exactly supposed to do with this information? We aren't even able to predict an earthquake down to the precise month. There are early warning devices that can be used once an earthquake has begun, and aftershocks are common, but that doesn't give anyone a chance to leave or take cover before the earthquake actually starts. Which makes any prediction functionally useless. Hurricanes, tornadoes, thunderstorms, blizzards, you can see those coming with a reasonable amount of lead time: enough to say 'there's something coming, make yourself scarce', but not enough to where you're asking people to make themseves scarce for a ridiculously long time. 'Uh, yeah, can you take April off and hole up with friends in Nevada? We think it's April, at least. Maybe. We're not entirely sure. You know what, go ahead and take May off too. Perhaps June.'

When one can say 'There is an earthquake warning until Thursday at 8 PM for Kern, Los Angeles, Orange, Riverside, San Bernardino, San Diego, San Luis Obispo, Santa Barbara and Ventura Counties and the Baja Peninsula', then you've got something useful. Until then, what is the point of this?