I had not planned on attending O'Reilly's Gov2.0 conference, which is an exposition and dialog about new forms of government and information technology. But at last week's Foo Camp (another O'Reilly event) I met a number of people in the field, and I became pretty excited with what I heard.
For example, I attended a session on government and data and sat next to a deputy CTO from the White House. I was surprised by the sincere and urgent dialog that was taking place with information activists and coders. The White House and geeks? What's not to like?
As I write this, I'm sitting in a huge room in the third sub-basement of the Grand Hyatt in D.C. Microsoft's chief research and strategy officer is speaking, so it's a good chance for me to reflect on what I have seen so far.
Tim O'Reilly is pushing the idea of Government as operating system or platform. I interpret this to mean that if Government allows "OS calls" or "APIs," then civil society and commercial services will be able access government in new ways. Think of the strategy behind Google Maps: Google releases the API, offers some hosting, and as a result tens of thousands of developers create mapping applications. If an API was released for accessing a government process or service, users would be able to read (and perhaps write) back to it. It's a big idea, and I'll need some time to follow it through completely. Does the idea break down, and if so then where? What happens if the intention or interest of the government suddenly changes? What would writing look like, and would it be subject to misuse? Isn't writing already being done by lobbyists, and do we really want to make that process simpler?
Regardless of the details, something is clearly going on here, sixty feet under street level, in a vast but unreal ballroom. An unholy alliance of free/open enthusiasts, D.C. suits, information activists, spooks, soldiers, and entrepreneurial techies are thinking about government and the Internet, and moreover they are experimenting.
One of my favorite examples of a successful government "hack" is called Amver and it's from the Coast Guard. It was started as a decidedly pre-2.0 app -- or even .4beta! -- in 1958. Amver is a voluntary, opt-in system for ship captains, asking that they simply let the Coast Guard know their departure and destination, and to report their location every 48 hours. That's it. The Coast Guard then keeps track of this, and when any vessel is in danger, it extrapolates the nearest ships and knows whom to contact to ask for help. Amver was used in the recent Air France crash -- indeed, 190 rescues have relied on Amver so far this year. Recently, Coast Guard developers created a graphic-only Web input interface, a virtual spinning globe like the one in Google Earth. This single, banal new feature has significantly increased their capacity simply because so many users of the system speak so many languages.
Why do I like this project? It's a reminder of how much can be done with information, it's opt-in, and it is deathly serious. Also, it demonstrates how important it is to be language inclusive.
Another great moment at the conference was the announcement of the Sunlight Foundation's Apps for America contest winners. GovPulse, for example, won second place by making the Federal Record readable.
Other notable highlights:
- EPA functionary Kim Balassiano's MyEnvironment, a location-based portal for all the EPA data in your area. (Personally, I would have named it EveryBlockTox.)
- Mikel Maron's work with Open Street Map. This is the humanitarian wing of the incredible OSM project. It maps places like Gaza, Afghanistan, or the largest slum in Africa. An audience member next to me asked, "Isn't Google already doing that?" The answer is no, absolutely not. Check Google's map of Kabul versus the one offered by OSM; there is no contest.
- George Clack of the U.S. State Department, who ran a competition for short documentaries answering the question, "what is Democracy?" Using YouTube, they allowed the public to vote on the best video. he videos are kind of tragic, but it is laudable that State is engendering dialog about democracy in contrast to the old approach of "making the world safe" for it. Clack was one of the several people in government who argued that government needs to learn how to let go online. It's okay if users of government sites have opinions, if they contradict your message, if they give you feedback. O'Reilly had a great line in this spirit: "Government does not regulate your political or personal opinions on the roads that it provides and maintains, so why should it be so worried about its Websites?"
And of course there were some lowlights:
- The very same State Department project. They had to throw out two thirds (!) of their finalists because the submissions used music under copyright. The intellectual property lobby is hampering our ability to conduct diplomacy, and thus damaging our national security.
- A presentation on Second Life diplomacy. Like most Second Life presentations, I had a great deal of trouble understanding what was actually happening, or how the use of bad 3D graphics helped or hurt the dialog. It did not take long for the presenter to start echoing the worst cyber-utopian spiel that we have heard since virtual reality made its debut: that this new media would finally create a massive global consciousness... If this is the new face of diplomacy I am cleaning out my bomb shelter.
- Are You Safe. This is an iPhone app that uses crime report databases to show urban "danger zones" and help people navigate through a city. We must have seen at least five such projects throughout the conference, all of which walk a dangerous line between being useful, ethically complicated, and potentially disturbing. There are laws that prevent this sort of information being used by real estate agents because it leads to more segregation an less social mobility. I question the utility, though, as well as the ethics. Does such an application in any way reduce crime? Does the app differentiate gang-on-gang, domestic, or other relational violence from professional robberies against strangers? Moreover, would Wall Street have shown up bright red and glowing over the five or six years preceding the mortgage crisis, during one of the longest and largest crimes in American history?
In many cases, the successful applications demoed at Gov2.0 reflected collaborative, two-directional implementations. They were often done by real organizations (the state of Utah, BART, the Coast Guard) that already have a working model of needs and users. The worst applications were often done by coders who moved into a new topic and created naive, first-order projects.
Most developers are thinking from databases outwards: they are simply hoping to visualize (or, if they are a behind the curve, "mash up") government databases. These projects offer quick, high returns, but in many cases actual users and uses are not well conceived.
Finally, fewer of these developers are designing around actual mechanisms of social change that might improve the situation they are visualizing. Nor are they imagining how to impact government with the data they are making usable. They should listen to Philip Ashlock, who gave the simple, pithy advice: "The goal is read/write government."
I mentioned earlier how sincerely interested a number of government people, most of whom are new to their jobs, are in accelerating public access. But data is slow in coming; systemic change will take a long time, and it is crucial to remember that not all administrations are as invested in transparency as this one is right now. So I was happy to see that a few of the projects are aimed specifically at expanding the data that's available, with or without goodwill from D.C.
Among these folks, the hero in the room is Carl Malamud who runs http://public.resource.org. Carl is a radical who crowbars open the U.S. government. In the mid-90s, with some urging from information activism god Jamie Love, Carl put the SEC's EDGAR database online against all odds (read the full story at http://public.resource.org/sec.gov), and has been continuing this mission in one form or another ever since.
Recently, Carl strolled into libraries around the country and managed to collect about 20 percent of PACER, the Federal court record system. PACER costs money to search because the government uses a commercial "value added" (to themselves, apparently) provider. But the government also allowed free access at a few libraries, at least until they noticed that Carl was downloading in bulk. After they closed the Malamud loophole, Carl provided that 20 percent to Steve Schultze and other insurgents at Harvard's Berkman Center and Princeton. This second team developed a Firefox plugin that allows lawyers who need to do PACER searches to do so for the regular fee. But as these lawyers access data it is cached by the plugin and uploaded to the Internet Archive. The next time someone searches PACER for a previously accessed court case, the plugin will intervene and access the now public and free cached data, rather than the private, official-but-expensive data. Steve admitted that his goal in building the plugin was simple: he wanted the plugin to be useful for as short a time as possible. If it works, it will show the government that it should be providing the information for free.
This is great information activism: building applications that provide services but also imply and, through their use, coerce further openness with the goal of a better society. In a sense, these projects are "sticky" -- they function but also transform.
If anyone has a good name for these new sorts of applications, please suggest it in the comments!