Can an automated consensus process provide features that would give it some advantages over the in-person kind?
I’m not sure, but I think that there are options worth exploring. Right now lots of people are doing distributed consensus over new media including email, wiki, and video conferencing. For example, see notes on the Wikipedia consensus process (http://en.wikipedia.org/wiki/Wikipedia:Consensus).
My sense is that some of the consensus processes and methods that have been used extensively for group decision making are more mature than current distributed processes. In person processes rely on visual group mechanisms and other in-person group dynamics, some of which may not be replicable in cyberspace. Other processes and conventions that were developed in shared physical environments, if rethought, could be distributed and might be more scalable by automation. Once automated it should be easy to automate certain counting, recording and real-time display functions. This might help to compensate for the lack of visible body language or other in-person feedback mechanisms.
While dotmocracy (http://www.dotmocracy.org/), may also be a viable process to base automation on, this post concentrates on an examination of the mechanisms and automation options for “Fist-to-Five” consensus (http://www.freechild.org/Firestarter/Fist2Five.htm)process.
Fist-to-Five hand signals allow all participants to indicate visually without speaking the wish to:
- block consensus (fist),
- suggest changes (one finger),
- raise minor issues (two fingers),
- let the issue pass without further discussion (three fingers),
- affirm the decision as a good idea (four fingers), and finally
- volunteer to take a lead in implementing the decision (five fingers).
This can be done by having a facilitator ask everyone for a response, but can also happen spontaneously during discussion. Because it can be done immediately and non-verbally in response to a speaker – immediate feedback sometimes occurs, which in the best case can speed and create widespread comfort to the participants that the conversation/decisions are heading in the right direction. In other cases, it provides immediate feedback to a speaker that they are heading in a direction that may not have much support, allowing the group to collectively change course faster.
So in the distributed and automated context something based on ‘Fist-to-Five’ would support the same six indications. Either by originator or in summation indications could be automatically displayed to all participants. Derived values could also be automatically tallied and displayed (e.g. 84% consensus, 0% blocking, seems promising to continue working it). As within meat space, a facilitator or group member could request responses from all. Responses would also be generated from participants spontaneously and displayed in real-time. Text from responders including reasons for blocking, comments and suggested changes could also be displayed to all.
I’m thinking about this mainly as a real-time immediate response form of distributed consensus. To use the benefits of the in-person model, the main dialogue might be phone based dialogue, comments, blocking and suggested changes generated by individuals without necessarily interrupting the main flow of the dialogue. This is just one model. Non real-time models where a question could be left open and worked – as long as there is some cut off of ideas and a consensus process at the end might provide more value or be an alternative available process.
Clearly most of the technology toolkit components exist – the hard part is having groups begin to agree on, adopt or implementing an adequately configured tool kit. Then the next hard part – having groups begin gain experience and begin to adapt and change them.
The ideas above need work, however simply replicating our current voting system as an electronic process complete with the tyranny of the majority, seems short sighted as we continue to look for ways to distribute and scale democratic processes and allow them to be used and supported by a wider range of processes and entities.