Lots of people are excited about the video-calling capabilities that Android will be inheriting through Google's recent acquisition of Global IP Sound. What's more interesting - to me, at least - is the ability that Google has to turn the operators into just another IP pipe.
Today, your cellular voice calls are just that: voice calls. They're billed as voice calls. Sure, you can use Skype or Fring on your phone, but there are alot of tradeoffs to doing that (it's not the same phone number you typically use, you need to have a fast 3G connection, dropped calls more often when in motion, it's yet-antother-app instead of your "phone", etc.). Some vendors like Research in Motion have enabled Wi-Fi calling (aka UMA) on some of their devices. When you have a UMA-enabled device, and a UMA-enabled carrier (like T-Mobile USA), and your phone is connected to a WiFi access point, it tries to connect to your carrier's servers, and if the conection is good enough, it'll route all your voice and data over WiFi... so it's like you're using your phone back at home, even though you're in a hotel room in Phnom Penh. Pretty cool.
Well, sort of cool. For this to work, you need to be an old-school telco, with old-school gear from old-school vendors like Ericsson and NokiaSiemens.
But imagine a world where you had a fast IP connection on your phone wherever you went... a 3G, 4G or WiFi connection. Imagine that you'd already chosen a "virtual" telco like Google Voice to be the keeper of your phone number. All of a sudden, you can start making phone calls and just treating the data network as your main bearer network. What happens if you're in a spot with crummy (EDGE or GPRS) data coverage? Not a problem: the phone "app" on your device will just use the native phone network to route your call (like Google Voice does if you use the iPhone web app, for example).
So what piece of this puzzle does Global IP Sound give Google? The pieces that let you deliver high-performance voice (and, of course, video) across a number of different data networks (some fast, some not, some consistent, some jittery) and different devices (smartphones, not-so-smart phones, desktop clients, etc.).
By making the acquisition, Google gets to own what could be a valuable piece of infrastructure as voice (and video) moves from the telco-centric model to a more Internet-centric model. The biggest surprise is that the existing guys (Alcatel-Lucent, NortelSiemens, Ericsson) didn't pull the trigger sooner.
Actually, I guess that's not such a big surprise.