This detection in itself doesn’t do much for us. Our by-far-best-accepted models all agreed that gravitational waves should exist, so learning that isn’t particularly big news. And the source of the waves detected here is presumably something of an unremarkable sort, most likely a merger of two neutron stars or black holes (objects that we already knew exist, and which must merge occasionally).
Now, as time goes on, we’ll be able to detect more and more of these anticipated sorts of events, and hence learn more about things like the abundance of neutron stars in the Galaxy. Which will be useful information, but unless they’re a heck of a lot more or less abundant than we expect, still won’t be revolutionary.
What’s really big here is that, now that we’re able to detect gravitational waves, sooner or later we’ll detect them from some sort of source which we didn’t know existed, or possibly even didn’t suspect existed. Once that happens, we’ll have a huge advancement in our understanding of physics.
The catch is, the hardest part of detecting gravitational waves is in picking out the signal from the noise. In fact, with current instruments, it’s so hard that the only way we can do it at all is via template matching: You take a bunch of expected possible events, calculate what a gravitational wave would look like from each of them, and then compare your data stream to each of them to see if it matches. This is sort of like how, at a crowded party, you can’t follow a conversation at the other end of the room (it’s too noisy to figure out what the words are), but you can still recognize when someone says your name across the room (because you have a template of what your name sounds like, and you’re always listening for it). This method works, but it has the drawback that you can’t find anything that you don’t already have a template for.
There are three ways past this problem: You can construct general-purpose templates that would match a wide variety of sources (but which give you less signal-to-noise benefit than a specific one), you can try to decrease the amount of noise you have, or you can try to find some other way of picking out signals from the noise. Personally, I favor the last approach, specifically by building more detectors: A real gravitational wave will show up in all detectors at the same time (or rather, delayed by an amount corresponding to the separation between them), but the noise sources at different detectors will be mostly unrelated to each other, so something that shows up in many detectors at once is more likely to be a real signal. The big advantage of this is that it doesn’t require development of any new technology: It just requires us to build more of what we already have.
