Hey, from SourceCred so biased, but may be able to offer some ideas.
Are these new techniques? More of what 1Hive has seen in the past when HNY was mooning? Is recent abuse documented anywhere?
From the comments here, I’m guessing you’re seeing infractions that fall into a grey area? The often blurred line between collaboration and collusion? In such cases, Gravity and Celeste both seem promising.
Using Gravity to provide (hopefully) skillful dispute resolution could go a long way, especially if there’s a genuine disagreement over what constitutes farming. It also has a nice side effect of building trust in the community. If people in conflict can communicate honestly about something difficult, and come out the other end, they’re more likely to be respectful of each other and collaborate well in the future. It also shows other community members that it’s safe to voice concerns.
However, if there’s genuine disagreement over what constitutes farming, there will likely be cases where people just can’t reach an agreement. Celeste could be useful here. From my (admittedly crappy) memory, in the initial wave of cheating, as documented in the epic Pollen user report thread mega, the hive was pretty good at collecting evidence in varied and creative ways. And there was a general sense of who the offenders were. But there was hesitation to ban all but the most eggregious offender (he who shall not be named but has apparently recently re-appeared? ). Presumably because there weren’t systems in place to make those calls without the perception of bias by project leaders, who do take issues around free speech and censorship very seriously. Celeste could offer a good way to reach the nuanced, intersubjective judgements required here. In more obvious cases (e.g. clear upvote rings on spammy content), Celeste could provide a legitimate avenue for the community to punish offenders (opt of out payments, ban, etc.). In less obvious cases, it could help 1Hive build precedent and “case law” interpreting the covenant’s values. This could be necessary in some form, as defining hard rules in this area can be problematic; they can be gamed around or prohibit valid speech, or instance. This would have a nice side effect of further defining 1Hive’s values generally. It would also provide a regular stream of disputes, providing testing for the system and revenue (DAO drama is an inexhaustible natural resource ).
Another possible way to address this is tweaking the algorithm. The first idea that comes to mind is just lowering the amount of Cred entry-level roles can mint? Like others in the thread, wary of negative Cred. It generally goes against the positive-sum design philosophy and I can imagine it leading to new types of abuse.