Bots for Administrative Work
Opened this issue · 14 comments
As this group begins to move, it should not be fully expected for all issues and codes to be reviewed by members or people on-boarding. It would be helpful to create a presiding role, a non-voting role, for a person to monitor the slack channels and provide guidance to members of the collective making mistakes.
The responsibilities might include: taking notes of meeting, ensuring and tracking mistakes made by members for training and/or remediation of rights, and other administrative roles, role modeling golden behaviors. #22
Standard behaviors or best behaviors are terms to be avoided in this section of future governance as that they imply homogeneous approaches, which might impede the diversity within the group.
This team should consist of more than one person to manage the expected work loads of the community.
Supporting evidence:
@iceLearn @markwhiting might be two interested in this topic.
I expect this system is going to lead to a lot of notifications. Luckily they can be turned off on a per thread basis :-)
@qwertyone the needs: consensus label is used when consensus is broken on a proposal. We assume consensus unless its broken explicitly.
Ok.
I'm confused about what concretely this is proposing. Just trying to piece together meaning:
As this group begins to move, it should not be fully expected for all issues and codes to be reviewed by members or people on-boarding.
I somewhat disagree with this as it is written right now. Isn't the central intention of our governance mechanism that members are expected to participate in topics that they care about?
It would be helpful to create a presiding role, a non-voting role, for a person to monitor the slack channels and provide guidance to members of the collective making mistakes.
I don't understand yet: what about our existing governance mechanism disallows this? Why can't someone do this now? Why do we have to take away their votes?
It would be helpful to me if you could provide specific problems/scenarios you foresee, and what this group would be empowered to do in response that they couldn't do otherwise in each scenario.
Problem
Information overload created by the platforms that makes it hard to track changes for members (some members leave for a month at a time and return). No core team dedicated to make certain the larger group is up to date with changes and accurate interpretations. Inherent assumption that as it is now that the members of the collective will maintain a current and accurate mental model of the state of Daemo.
Proposal
A team of self selected members who will read the slack channels, github issues, and wiki. They will serve as a liaison between the living Daemo to members who do not maintain an accurate understanding of where the group is. Their job is reach out to individuals who have publicly expressed 'needing an update' or ask questions from regarding the platform.
Potential Alternatives ranging from fully human to automated
A team of people without voting rights (the nature of the role gives them an artificial influence compared to DRIs), a chat bot that looks for some of the things mentioned before and presents answers dialectical, or content specific static web sites that contain linked to meeting notes, a daemo hosted iframe of past meeting videos, a news bulletin.
A combination of them might be possible. A potential project for the group might be a chatbot dedicated to maintaining the mental model of the group.
Short Term Implications
The recruiting of a group of individuals might not be necessary for the group, as that people are already serving this function when mistakes are made.
Long Term Implications
At minimum, the team would require dedicated people to maintain this effort. However, the initial run through could serve as a springboard to record conversational dynamics and topics that users engage most naturally in as well as identify errors that are created by the work environment. The collected data might be used to slowly transfer knowledge into a bot that can handle most common queries and eventually take up the role of this proposed team.
We basically do the page with meeting notes and videos now (here), although we could do it for other meetings, e.g. things outside of our standard Wednesday meetings.
My feeling is that problem you are aiming to fix with this is important, but that doing it with a non-voting roll seems convoluted and problematic in other ways.
I really like the idea of a simple, low effort solution, e.g. formalizing practices around note taking and note sharing. Do you have more thinking on that, perhaps thats something we could flesh out here or in an upcoming chat?
I just revised this proposal. I view this as building upon the behaviors I already see within the group. How might we formalize and potentially automate the pro-social helping behaviors demonstrated by @markwhiting and @mbernst ?
If voting is a non-starter, then clearly the proposal needs revised.
This needs more information and details. I’m really not sure how is this OGov. What is OGov according to you?
@qwertyone said
A team of people without voting rights (the nature of the role gives them an artificial influence compared to DRIs), a chat bot that looks for some of the things mentioned before and presents answers dialectical, or content specific static web sites that contain linked to meeting notes, a daemo hosted iframe of past meeting videos, a news bulletin.
Yes, I agree with you! We don’t need more human involvement in this. Would you like to write the bot? The bot should follow the prosocial behavior; how do you envision the bot will show the such behavior? What should be the prosocial behavior according to you for an open community project?
A combination of them might be possible. A potential project for the group might be a chatbot dedicated to maintaining the mental model of the group.
I’m not sure what do you mean by combination here. Can you specify the entities?
I would certainly support the creation of some sort of opgroup or bot whose goal is to make sure people are aware of the comings and goings with proposals. So far it's been ad-hoc labor.
@neilthemathguy Open governance relies heavily upon the appropriate sharing of information. The aggregation of news, posting calls to participation, etc. are governance functions to this effect. We try to automate it as well as create HMI policies that determine information assess over time.
And yes, as an aside to platform, I would fine scripting the bot, though it would require additional learning upon my part. I have written one in Python/R before. The description of prosocial behavior would be well designed gentle reminders. I would like to think we might take the highly repetitive behaviors captured in the Daemo related forums and move them to code.
@neilthemathguy @mbernst @markwhiting
I guess the group's mission is to: 'is to make sure people are aware of the comings and goings with proposals'
What if we had an opgroup whose method is to create a database of interactions and incrementally seek ways to automate (create features) or informate (i.e. create interfaces, static web pages) to off load interactions that occur n>=3?