The disaster exacerbated the existing crisis. Issue of diabetes at 70% in some housing areas, stuck on 40th floor with foot amputations. Have diabetes because they're in a food desert. Also they didn't get power for 3 months, whereas wall street was back on in 3 days.
72 hours is how long it takes the gov/formal response to get to you. You're responsible for yourself in the meantime.
Occupy Sandy group
- Mismatch of needs and what was sent.
Wake of need, announcement, response. How do you close loops?
- Transparency. Facebook and Twitter, people managing those channels and the corresponding coordination docs.
- Redundancy and scales - one smaller Twitter account says "we have enough diapers" then the larger one RTs, then the other smaller ones RT.
Had main hubs, distribution from those to relief centers by making internal requests. Acted as a buffer, needed because of bottleneck of cars.
Feedback loops and delays in OS
were based more on communication (shorter delays), whereas larger organizations have slower feedback cycles because the communication is delayed (individuals have a delay in GOING to the distro site, saying what they need, returning to say what isn't working).
- delays on communicating with distro
- delays on distro requested goods
- delays in stocking distro
Could swap between "I'm just a person here to help" and "I'm the Chief Whatever-The-Fuck for Occupy Sandy"
Tech in use (or not)
The tools we're using are corporately owned, and can be taken out by governments. How much reliance can/should be on these tech groups?
- status.net was a federated Twitter alternative. Alternative instances of social network sites. Wordpress/buddypress instance.
- usability issues
- people expect tools to work. fooled into "the mall" where you know what you’re getting and how much it costs. Versus the bazaar, where you contribute, and it's nuts, etc etc.
- CiviCRM as a system set up during OWS which wasn't done in time, but was ready to go for OS.
Less Orwellian, more Huxtabee. It's not "you can't send political messages" -- if it were, we'd be further ahead in encryption. Pushing forward federated decentralized tools where people are managing small, trusted networks which will talk with other networks is the way out of this mess -- in disasters fast and slow.
Mesh networks... not quite ready in time.
Interfacing with the formal through informal interactions. "I can't tell you this, but..."
Connecting to local resources? School, hospitals, etc? Yes.
Burning Man listed as a terrorist orga because it meant more funding for SF official response groups. But that was problematic as soon as it became clear how amazing Burners are at response. Couldn't interact with them.
Co-Ownership of addresses - you can have one identity which multiple people represent. This allowed interfacing with formal organizations as well as clarity for those under duress.
- Want 15 people on an account, 5 need to verify something should go live.
- Social media apparatus is for listening and RTing. Personal connection to those amplifications. Lots of implicit structure. "I knew who to talk to through personal relationships"
- With many accessing, how do you deal with verification and accuracy? You accept the 10% fraud. 10% of the information is going to be wrong. But the overhead to be precise would cost you 25% of the information.
Having a pre-existing community group allowed this quick spin-up to occur.
Moving from response to recovery... we set up a distro in a non-affected area. Gov agendies for ongoing needs. POlitical process to point out that disaster isn't just floods. "Who here needs something? We got stuff." We found XX people who passed through Union Square need food or councelling or etc.
Grew the membership at the distro hub church. To interact with the government agencies, the burros needed to come up with the long term recovery groups. Staten Island still sends emails, and we set up their tech infrastructure.