HaroldDavis3
commented
3 years ago @jennymith Thanks for organizing this. Seems both Discord & Forum would be best. All for keeping up Gh project board updates too.
Update as of Feb 4
Here is the latest code that shows our ability to combine Loomio, SourceCred & Policykit for our decision making testing needs.
Latest thinking on (ranked choice) testing: After following some testing, discussion & experiments in other communities using SC (like the SC community themselves) I have been thinking of the distinctions in protecting against gaming the mechanisms —and protecting against silos of top down power.
These seem to be fundamentally distinct approaches. When designing for protections against gaming, things seem to get choppy when sussing out which contributions are deemed valuable in broader gov’t models. This has interesting impact on strategies implemented for SC. Gov’t models seemed to get infinitely more complex. This is concerning as it seems to broaden surface area for both top down & bottom up attacks in long term.
When designing for protections for top down power silos, the same means of protection from top down structures are also useful for protecting us from gaming from the bottom.
Example: I’ve confirmed it is possible for us to track cred flow via Policykit-- of selected posts in a given SC plugin.
Cred flow checking w/PK can give us bottom up quorums of what is most relevant for review before decision making actions.
Kinda like what this can do for human to human trust in context of a trust-less technical space in long run—but would stress the importance of bottom up mechanisms to establish protections —not engineering against gaming per se —but for broader gov’t models that can hold individuals & groups accountable to voices of whole collective. This designs beyond gaming protections and goes into bottom up systems design.
It also allows us to keep gov’t models in decision making fairly simple at the core with complexity in community interactions growing infinitely at the edges.
Bottom Up Role Selection via heavily weighted reaction model:
- Initiators
- Reaction specific indicator of new proposals for decision making etc. This role is responsible for gathering needed community feedback for decision making process & starting the process as needed - Setting quorum periods via Policykit & SC instance handling etc.
- Deliberators
- Reaction specific indicator of ongoing discussion on a given decision. This role takes on the responsibility of closely engaging in the decision making process. As well as bringing new, unfamiliar people into the guidelines of respective processes(may vary drastically from decision to decision of course).
- Observers
- Reaction specific indicator of ongoing review. This role takes on the responsibility of deep understanding of decision at hand & closely observing deliberations, with specific period for deliberation feedback triggered by PK. This period can be critical for ensuring consensus - rinse and repeat of this period - until consensus within respective process frameworks is met might be ideal.
First glance, this looks like a unique chance to track value flows, this can help ensure flows stay unalienated long term. Standards of checking that gaming has not occurred should be fairly straightforward to create as we go, which adds complexity to overall gov’t model, but with important constraints for bottom up voices.
This model dovetails nicely with our Two Row Wampum Working Group efforts, as it is largely inspired by indigenous bottom up consensus protocols. This is also exciting path toward protected spaces for decision making by women in web3, as this model should do well in any desired R+D in this area 🙂 Demo on Feb 11! 🤞.
Current Update of Jan 21
-
Latest testing will include using PK as a "driver" for SC & Loomio in the discord bridge mentioned.
-
Loomio gives nice polished modular decision making options.
-
Simple test could be asking PK to check if cred has flowed where the community has determined is needed for respective decision making. PK can then make an action based this notify users or even enforce predetermined actions like making new channel.
-
Folks in metagov have gotten loomio to release some API endpoints for voting & discussions!
-
Here is very nice work being done simultaneously in metagov prototyping.
(First Week of Dec)
-
[x] Currently reaching out to Amy Zhang of PolicyKit. Hoping she'll have some insights for us on a likely timeline and more.
-
[x] Tinkering w/ Friedgers OI chat & PolicyKit as is. They have templates that will require limited technical know how, which conveniently overlap with Jude's original ideas of quorums, date, time triggers etc. If possible with OI chat's Discord bridge, as opposed to Discord alone, we can even pull off Gaia hub & Auth integrations he mentioned too. If we are pressed for time later on, this tinkering will pay off if we want to use policy kit w/out OI chat.
(Second Week of Dec)
- [x] Will update project boards with results of said tinkering.
Was able to speak with both Friedger & Amy! Friedger is unfamiliar with ins & outs of the Matrix client bridging but sees the possibility of combining his OI chat with Policykit for the benefits of Stacks Auth & Gaia storage for our voting mechanism efforts. Amy Xzhang of Policykit has been kind enough to offer help on any possible blockers & this should be huge help to figuring realistic deadline if it works outs.
Tinkering with OI chat & discord bridge showed a straightforward path to how Policykit should be able to work for our community voting needs. Tinkering with Policykit showed its ready to test it usage with OI chat; as I was able to get through their README tutorial. Will report more here if we have any good news!
-
[x] Begin Discussion on other testing approaches promised in grant app.
-
[x] Publishing discussion results here for further input. @jennymith, Sean & myself will meet to discuss best paths forward for possible testing options. Some approaches may need an extension of the respective milestone others may be doable within the original time frame --extension may be best for quality testing. Sean mentioned liquid democracy prototype overlap possibilities. I mentioned to Jenny possible use case of the community mod project for our vote testing needs. She articulated the current limits, as its designed largely for moderation. But we will discuss further in context of vote testing possibilities. We briefly discussed Jude's mention of ranked voting as means for timely testing needs too.
Will update with our discussion points after meeting with Jenny & Sean. Looking for feedback here as well of course.
(Third Week of Dec)
- [x] Update with Amy X Zhang tinkering & timeline help. Amy offered to add OI chat integration to PolicyKit's list of integration efforts alongside discourse! Will be attending Metagov seminars in Jan to start this effort.
(Fourth Week)
-
[x] Tinkering with SourceCred for testing option & contribution based approach to Jude's ranked choice vote suggestion.
-
[x] Stacksgov/pm GH repo, Stacks Discourse Forum & test Discord server/bot data successfully loaded into local SourceCred instance for cred calculations/grain distribution tinkering.
-
StackCred GH pages - testing template Instance for local tinkering. Cred & Grain config can be pushed here once testing begins in Jan.
(First Week of January)
- [ ] Meeting with Jenny & Sean for last 2 testing options.
- [ ] Proposing community decision making testing via SC weights configuration in GitHub & Discord (ranked choice via grain)
- [ ] 1 week of contributions & cred earned in discussion, w/grain distribution on final day -> 1 week of decision making on community contribution weight config w/grain earned in wk1.
joberding
whoabuddy
cc: @joberding @HaroldDavis3