Recovery 2.0: An agenda?

I have no idea what we should accomplish at the Recovery 2.0 meeting in San Francisco (which is on for 6p in the Olympic Room of the Argent Hotel, 50 3rd St. in San Francisco).

The starting point is simple: I thought it would be a good idea to use this opportunity to get into a room some of the people who want to find ways that the internet can be better at responding to needs in the next disaster, and in the recovery from recent disasters. What comes out of this, if anything, is up to everyone in the room. I’ll suggest a starting point and will, as Craig says, get out of the way. I will be the least qualified person in the room to lead anything; I’ll be eager too follow. We’ll get this on the wiki and I urge you to correct, add, delete.

The cause: What made me write my first Recovery 2.0 post was seeing a confusing though good-intentioned array of more than 50 boards and means to find the missing. We need to do better.

Two goals: We need to be better at swarming. That is, when we see a need, perhaps the best thing to do is to see whether someone is trying to meet that need and whether they’re doing it well. If so, perhaps the best thing to do is point people to that effort with the power of our links. If not, the choices are to offer to help or to do better. The distributed nature of the internet is its greatest strength but so is our ability to swarm and pool our efforts.

Thus we also need to be better at communicating. We need a means or a place to better share needs, solutions, resources, and calls and offers of help.

A review: Tom Evslin suggested that we should do some self-criticism of the internet’s response to recent disasters: What should and can we do better?

Needs: We need a place to communicate. Is a wiki enough? Do we need a blog and a forum?

David Weinberger suggests that we need tags or a microformat so the things people do in our distributed places can be discovered.

What else?

Field day: Jeff Pulver is suggesting that the web, like ham radio operators, hold a field day to test what it can do.

Meeting: Yes? No?

Names: Who’s doing what?

We’ll be together just an hour or two. Nothing will be accomplished. Much should be started.

I would love to have a volunteer at the meeting act as wikimaster to record ideas and issues. Thanks to Ross Mayfield at Socialtext and the great work of Greg Burton, we’ll soon have a new wiki at

  • Bostoniangirl

    Find a way to keep the process of recovery transparent. I’m really concerned that New Orleans could wind up turning into a back-door fix up, where all the poor people are kicked out.

  • Off topic but Jeff’s CNN bit is up over at the Political Teen

  • grog

    uptime and reliability of the internet in general will always be an issue until computers, network switching equipment, transport systems and the like are as interchangeable and interoperable as ham radio. the only advantage ham radio has is that of a common air interface, regardless of the features on any particular radio. so long as you have two radios meant for operating in the same frequency range, you’ll be able to communicate. no worries about brand-incompatibilities when it comes to basic voice or low-speed data communications. it’s time for some standards to be placed upon vendors in the computer arena as well.

    just my $0.02


  • Pingback: PRspeak()

  • Shalom Jeff,

    I helped to set up and run the emergency community computer center in the Cleveland Convention Center. We were prepared for as many as 500 refugees. We got about 150.

    The services we provided were:

    Internet access for FEMA registration.
    Aid in using Googled search engines to locate lost family and let friends and family know where you were.
    Help in checking existing email accounts and in setting up new accounts.
    Games access for the kids.

    Our No. 1 complaint? The FEMA system was overloaded. Forms that should have taken no more than 10 minutes to complete often took more than an hour with no guarantee that the host computer had actually received and processed the information. I felt like I was back in college and dealing with an IBM 3000 in batchmode.

    The system would have worked much better if we could have had the forms running on our own server that could then uplink to the FEMA computers once a day on a prearranged schedule.

    The FEMA people on the ground were great. But the system was not able to handle the load.


    Jeff Hess

  • Jeff Hess – do you mind if I put that in the wiki?

  • mm

    “I have no idea what we should accomplish at the Recovery 2.0 meeting in San Francisco”

    Then you should cancel the meeting, let everyone save $$ on flights & hotels, donate that money to the Red Cross, and figure out what the goals are before you move forward!

    Just like when you design survey questions: if you don’t know what you’re going to do with the result, then you don’t ask it. If you don’t know what you should accomplish with a meeting, don’t hold it.

  • Five of those goals are essentially communication objectives: share information, report and act on calls for help, connect the missing, provide connections for such necessities as housing and jobs, get people connected to these projects – and the world – sooner.

    One aspect that I haven’t seen commented on in your posts or elsewhere is language: a lot of the people that’ll have to have access to this/these ressource(s) aren’t fluent in English. If your disaster is in LA or in Florida, your Spanish-speaking users have to get access to the same info as everyone else in the same timely fashion.


  • James

    Looking forward to seeing you there. I was in Houston for 10 days.
    Took a couple of days off and then was in Washington for an additional 5 days working out of the American Red Cross (NOC).

    In Houston did and saw a number of different things. Such as:

    Network Troubleshooting
    Implemented bugzilla to facilitate coordination between 2 megacenters in their Public Announcements.
    People at work setup a metasearch that searched 50 or so different db sites where people were posting their information. It was / is a very productive tool.
    Helped role out and staff two computer centers.
    Numerous other people from work helped train and staff the “Computer Centers”. Which is were we saw the need for the meta search.
    The complete list is much longer…

    I’m very excited about this “meeting” and believe it can be productive.

    There seems to be another Recovery 2.0 wiki at .

    BTW — the linke “Recovery 2.0 Post” points to http:///

  • James – that was our old wiki at 4setup. We moved from MediaWiki to a hosted Socialtext workspace. Thanks for reminding me to change the main page on the old one.

  • Liz

    I was in Houston as well. I’m interested in talking about knowledge management. This is a huge problem still. I tried using flat text files, email, chat, blogs, and wikis to keep information current. Nothing quite worked, though I still would put most work into blogs and wikis for current info. At any moment during a disaster or in disaster relief, “what is true” has the potential to change, just as in a war zone. Who to call? Where to go? Who’s in charge? We passed around scratchy handwritten out-of-date xeroxes like they were pure gold. Giant pieces of ambitious bloatware were slow to respond, slow to be used: think of CAN…. or, as I have been calling it, CAN’t.

    Wikis were great, but it was hard to get people using them. If information flowed into them very quickly, then the organization of that info became a problem or was just timeconsuming. Blogs were good, but questions of authorship/authority stopped people from putting in their current information.

    What would work, I think, is some kind of tagging structure and… i guess i’d call it granularity… voting… and algorithms that help determine truth. Say I’m working in a huge shelter and I’ve just found out that the shelter manager is named Jane Smith, and her cell number. Hey! A True Fact! I jot it down. In the Astrodome I put all that stuff in a flat text file. But what if I was able to enter it as a “current truth” with the date & time and maybe my own identity and reputation attached? Then, other people who found it to be true could vote it up. (Of course, that information was obscured intentionally as the shelter manager didn’t want people like me calling her!) What if the “fluid truth” was something like, “the phone number you call to get section 8 housing” or “the best how-to document on peoplefinding”? Or a rumor like “FEMA is handing out cash cards!”

    The information coming from top-down was not reliable. Not at all. We treated that official story like just another rumor until it had been verified from several sources. By the time you get that verification, the situation may have changed completely!

    What you have is a rapidly changing environment. Figuring out what is true at any given moment is like creating a rumor processing algorithm.

    In addition to cell phones, wireless, and computers, I was thinking of little information devices. For example in the 90s there was a little handheld game called “POD” that never quite took off, but the idea was that you built virtual fighting robots, and when you happened to walk by somone else who had their POD device on them, maybe within 100 feet… then your robots would battle and you’d get a “piece” of the losing robot for your toolkit. What if instead, that concept was adapated to Knowledge Management? Information could be exchanged – it could be voteable, searchable, modifiable, between devices. (Which would not need cell coverage OR the net to work, though it would be great if all that info would then automatically slam up into some KMS system in the sky…)

    From here in California, I have been continuing my relief work and I find that during a day of calling and searching, in conjunction with my boots on the ground people who are in contact thru phone calls, I’ll amass better “fluid truth” than the head Red Cross people or shelter managers or hotline supervisors have. For example I told all the people I could in Austin and San Antonio all my most current Houston shelter info one day last week just after Rita hit. The top people in Austin and San Antonio were propagating hideously wrong out of date info, and they listened to me, a random person on the phone from another state (I told them how to verify the info.)

    I’m a complete amateur, but I wonder how, say, military intelligence handles its own knowledge or situation reports? I wonder if it is in such a linear old-fashioned way as I saw things working in these disasters? But also, I wonder if people like John McCarthy, who worked for years on AI knowledge management systems, common sense algorithms, would be able to help?

    I think these ideas are worth throwing into y’all’s discussion and I hope I can make it on Thursday, but I might not be able to.

  • Pingback: : The Jon Phillips Portal()

  • Pingback: TheBizCast » Blog Archive » Web 2.0 & Business Models()

  • I will attend, happily, excitedly, with bells on.

    Our blog, url above, has been wildly successful in helping donors and volunteers with direct contacts in the disaster zone. I’d like to develop our model into something that can be utilized in any sort of emergency situation. The recipe is simple and practical – what it takes are on site folks who will serve as ‘boots-on-the-ground’ volunteers, communicating by email and/or SMS to bloggers who update the site with what is needed where, by whom, with phone numbers, addresses, directions.

    I look forward to the discussion.

  • Pingback: Weblogs Work » What We Learned From Disaster Blogging()