Re: Super-connectors for robust network
Subject: Re: Super-connectors for robust network
From: Scott Mintz <scott.w.mintz@gmail.com>
Date: 7/23/10, 22:40
To: Clark Robinson <robinsonchicago@gmail.com>
CC: Barrett Brown <barriticus@gmail.com>, mano.singham@case.edu, tessig@me.com

I would like a clarification. Is the concern that we could end up with one big network rather than many small ones as everyone is connected? If this is the case, then I think to some degree overlap will be helpful to act as a redundancy in instances where certain participants become less participatory. However, I would passionately want to avoid what I view as a "homogeneous" web.

A random thought I had was to enable to network to grow at an orderly pace by instituting regulators, such that as a person's ability to add to their network is governed by their time on the network AND/OR related to the percentage of posts pushed aka reviewed, so the more active they are on the network, the larger theirs could become. While this could also play a role in creating some super-connectors (If I understand the definition correctly), if made public, could alter the incentives relating to pushing works.

On Fri, Jul 23, 2010 at 10:08 PM, Clark Robinson <robinsonchicago@gmail.com> wrote:
It will be useful to discuss (by e-mail) what adjustments we may want to make to the network structure in light of Todd's comment, and adjust the widget requirements statements accordingly.


TE: A robust network needs some but not too many "super-connectors"? You don't want a pyramid, of course. But you also don't want a flat homogenous web. I don't understand the incentives from the current description well enough to believe this will work. 

BB: Indeed, we want neither, and we're willing to make any changes to the schematic in order to prevent any homogenous web that might emerge under the proposed system. My thinking has been that individual participants would find it cumbersome to make too many connections due to the heavy resulting stream of information that would entail. Let me know what you think.

 Clark