Yesterday, I spent the day at a meeting on "city innovation" at Harvard's Kennedy School, with 30 or so CIOs, CTOs, and other technology executives from around the country. I did a short presentation on predictive analytics and cities (slides here) -- thanks so much to everyone who sent in comments and who emailed me with suggestions. The "aha!" moment of the day came during a coffee break conversation with Boston CIO Bill Oates.  Bill was describing how frustrated he felt by the city's procurement process (this is widely known as a problem across government).  He said that he felt like he was "handcuffed" by having to prove -- up front, and before actually doing anything -- that he wasn't being dishonest, wasn't corrupt, and was serving the city's best interests.  What if, he asked, he could instead proceed ahead and prove -- after the fact -- that his actions were pure.  Using transparency, rather than bureaucracy, to establish accountability and ultimately trust. This strikes me as a big idea. What we have now -- in the era of increasing information liquidity -- is an opportunity to re-think the way we establish trust.  This idea has been proven out by web services (think Ebay, Airbnb, StackExchange), and I think it's time we start thinking about how this applies to public sector policy and regulation. After the conversation with Bill, I ran back to my seat and sketched out the idea, then quickly turned it into a slide for my presentation in the following session.  This is what I came up with:

` The idea that the purpose of bureaucracy and (certain forms of ) regulation is to establish trust is perhaps obvious. But something about it struck me as a new way of looking at things. It's an idea that superblogger David Alpert gets at in his coverage of the Uber / DC fight, which he describes as a conflict between the "permission model" and the "innovation model". I understand that it's hard to get past the permission-based way of thinking.  Before information was available in real-time, it was the best way to make sure bad things didn't happen.  But we have a new tool -- real-time information -- that makes a new approach possible.  Yesterday at Harvard, we were discussing this in the context of government procurement.  At USV, we've been talking about it a lot in the context of online privacy (I'm pushing Brad to write about his idea for this soon). Hopefully you'll find this helpful -- I think I'll be coming back to it w/ some frequency now as we continue to work on this stuff.