I'm at the Where 2.0 conference in San Jose. Unfortunately, the Loki location finding software on my laptop, which I raved about in a previous blog post, thinks I'm in Toronto. Probably the conference team picked up its WiFi access points from an office or event in Toronto and shipped them down here. At any rate, it's ironic, given the topic of the conference, but more importantly than that, for a few moments Google thought I was in Canada, and sent me to the Canadian version of the site (www.google.ca) when I tried to search. It was no big deal, but it shows you how location data applies to things you don't always think of as location-related. And the potential downsides to poor location data can be serious. Imagine if I had some emergency-response product that thought I was in Toronto instead of San Jose -- or if I was on a VOIP phone that was registered to a different location, and then I dialed 911.
Bad geo data tends to self-correct as users notice their applications are reporting the wrong location and send in updates, and an interesting project, OpenStreetMap, will help geolocation data correct even more quickly. This project lets its GPS-equipped users record their location traces (geek alert: It's like AttentionTrust, but for the physical world). There are, of course, privacy issues with this (just like there are with tools that track where you go on the Web), but overall, the more people are reporting where they are and where they've been, the more accurate maps and location-finding data will be for everybody. This consumer-provided geographic data could become an important part of the growing economy for location-based products and services.