The number of active participants is a key parameter in the discussion and referendum phases. A certain number of participants must signal support for a proposal for it to pass to the next phase.
This is designed to prevent governance attacks where proposals are pushed through a low-participation environment simply because no-one noticed enough to stop it.
We use this parameter in the current governance processes via Discourse plugins. Actions such as creating a post, replying to a post and liking a post all count towards a participation threshold. Once passed, a user counts as an active participant.
This seems to work well. However, it probably canāt be migrated exactly to Community Trust governance, for several reasons:
The system is only semi-transparent. Weāve published the sorts of things that count towards participation, but never the full plugin. This is to prevent people from gaming it. This approach doesnāt fit a fully decentralized setup, however.
There is extra moderation on top. As the moderator, I have discretion to ignore posts which I consider spam but the system doesnāt flag. I think I do this even-handedly, but again this isnāt appropriate in a decentralized setup.
There are some interesting ideas in the space, including some pretty wacky AI-based solutions. Weāll certainly be exploring these, but for now we need to create something simple and workable.
The new token-gated participation proposal will help to combat spam, but we may need additional metrics on top of this.
Questions include:
What actions should qualify as participation?
How should we measure these?
How active should someone be to qualify here?
Should this parameter be global (i.e., at any given time thereās a single figure for number of active participants across the forum) or local (i.e., different sub-topics or governance areas have their own participation figures somehow)?
I think actions such as creating posts, replying to posts, and engaging in substantive discussions should qualify as participation, and regarding the scope, both options should be considered IMHO, since a global parameter gives an overall sense of community engagement, while local parameters can tailor thresholds to specific sub-topics or governance areas.
Local makes more sense as a sticky couple of threads might keep some proposal threads away from the top, and in that case should not be considered as āparticipated inā if all the attention is on other threads
any participation in the discussion should already be considered as involvement in the community. it doesnāt matter in which topic they were made and when. even better, if a person has been participating in discussions for a long time, it shows that he is constant and committed to the community. at the same time, I believe that all monosyllabic answers up to 10 words or emojis or likes should not be taken into account in useful activity. and the topics read and the time spent on the forum should be of half importance.
that is, the main indicator is the number of complex useful posts and the number of hearts received from other people.
Would participation on telegram and/or discord be considered or just forum? Also, hard to gauge the right participation. For several reasons someone might of already commented your thoughts and a simple I agree message might be sufficient. Reading comments and liking them shows engagement but its hard to decide if itās enough. Maybe have a tier of participation. Comments + replies + likes?
Would be interesting to see if we could also incorporate activity from telegram or discord somehow. I know the general goal is to have increased forum based activity though.
It would be interesting to have a very broad list of participation metrics, but rotating which ones qualify between weeks/months/seasons etc.
That way its not easily automated. IE oh I know i need to do 2 post 1 comment 3 likes etcā¦
Discord and Telegram are great suggestions, but is it possible actually to take it into account? I mean, a tool or something to measure participation. The forum is by far the easiest tool and the team has already internal metrics for that. I think we should focus to do things here.
Theoretically itās possible, but I do agree that it would be hard.
I think for any first version of this weād have to limit it to Discourse. Designing Discourse plugins + a Discord bot + a Telegram bot + something to link them all together would be a lot of work, even before we get into how to do the cross-platform identity linkage to protect privacy while preventing Sybils
A very good idea. I think this is the method that should be used to measure participation exactly. There is a user habit situation. Not all people are familiar with forums, but they use telegram extensively or friends who are relatively younger prefer discord. Some of our friends participate from all platforms. An online tool that can measure this engagement would be great.
we should add the reference votes in the telegram community, we wonāt necessarily get that result but it will be worth our reference, actually the telegram community is quite large. On the Hopr forum I saw less active members, so it will not convey all the information to everyone.
I think any metric should take into consideration the age of the account and the regularity of posts as well as the number of positive reactions, replies and likewise age or ātrustworthinessā score of the people interacting with the message.
However, ultimately Iām afraid any automated solution will eventually be gamed. Even posting quality content canāt be used as a reliable metric because ChatGPT can produce better contributions than most actual users. People can just create a few sock puppet accounts, space them over time and run a script to have them like, reply, etc. at random moments. You can quickly build a small set of seemingly reputable participants to push your proposal through.
So I would suggest either adding:
Human curation. This can be in the form of a popularity score, elections (fixed term) or co-optation by people currently in authority roles (mods, team, etc.) Maybe not the most decentralized or transparent but the safest
Proof of humanity types of scores. I believe Sismo, gitcoin and others offer something like that based on various metrics for a wallet (link to SNS profiles, on-chain activity, etc.) There could be a requirement for a PoH lowest score.
Here I think age of account could make sense as a parameter. Iām reluctant to modify token holdings by age, as tokens in HOPR are quite fluid, especially for node runners, but your forum pseudonym is a different matter.
ChatGPT is interesting. Iāve certainly noticed a lot of ChatGPT content in the past two discussions (some of it disclosed). I think itās a useful to help articulate peopleās thoughts (especially non-English speakers), but Iāve also seen people just dumping the proposal topic or the thread content and pasting the output.
I was an editor before crypto (hence my pseudonym) so I can quite easily spot these. But that might not be the same for other people, and I expect by this time next year things will have improved exponentially.
Iām interested in automated and AI-assisted forum moderation, but I canāt see a way for humans to be fully replaced here safely.
Proof of humanity is a fraught issue where HOPR is concerned. There are limits to how much metadata analysis we should condone. But Sismo is certainly interesting. I talked to the founders in Amsterdam and thereās lots to admire there.
For the activity part of this discussion, Iāve had issues staying active in the past governance discussions due to simply missing the announcements, not being able to keep track of the actual HOPR Announcements. I am subscribed to a ton of announcement channels, more than Iām sometimes able to read, and sometimes some slip through the cracks. But even if they werenāt I donāt think a Telegram announcement is always enough.
On that note, Iād propose a mobile app integration for HOPR Governance/Discussions iOS/Android. Would be much easier to keep track for users (at least for me), and easier to follow and participate in the discussions than it is through just this forum (I personally go through extended periods of no PC access)
As for participation, I donāt think you can go without human curation, as you can have participation in the form of simply stating support, or actively discussing and arguing something. Not sure thereās a foolproof way to quantify those, unless maybe some AI solution can help, but I have doubts thatād be foolproof either.
You could have different ways of interacting with proposals. Directly āvotingā at discussion phase Yes/No through some UI option, which would be the most basic level of activity, coupled with critical discussion which would elevate the level.
And I think the parameter for activity, if there is one, should be per discussion, not per forum. Donāt see how a forum wide one would help identifiying involvement in a particular discussion.
agree with this - posts such as āI support thisā arenāt really participation - especially if participation is incentivized buy tokens/nfts etc - āI support this because I agree with x point and y point that were raisedā isnāt much more to write but at least it take a bit of effort
good points re human curation. but especially with the number of update channels etc we all are part of - All project info should be accessible in one place , and that place should be somewhere we all go anyway so for me that is the network dashboard because we are generally node runners and generally pretty obsessive about checking our nodes performance
Whilst being a long term HOPR member, Iām no expert on DAOās or how best to add value to this discussion, more questioning how metrics and processes have been arrived at. I guess thats just the engineer coming out in me.
As other have said participation shouldnāt IMHO be restricted to doing certain things, as I have no idea what proposal I could put on this discussion unlike previous ones other than spamming the forum just to achieve a higher rank. We all have strengths and weaknesses so a more rounded review of anticipation which doesnāt affect contributors that donāt have a proposal shouldnāt be marked down.
also likes should count for zero! over on Torum one of the daily missions in to like 5 posts, so ppl just like the first 5 posts without reading them, which means they are then more likely to be in the 5 for the next person etc etc