RE: Your First Impressions on UserAuthority? Contest Inside!
First impressions:
- UA-ranking: Some interesting concepts, but no teeth
- Witnesses might not be the best spring for reputation
- Algorithmic curation with UA: Hmm, trying to half-heartedly fix one problem by making another one bigger.
- Why not UA based flagging instead?
UA for ranking
The current reputation system has massive problems. Creators of poor content who rent SP or use bid-bots can crank up their reputation in zero time and once beyond the range of those that might deal out an occasional flag, there is no incentive left to even keep up the guise of half decent content. While the UA ranking would be amazing if used by the platform itself, most importantly in our ability to flag those who have artificially inflated their reputation. The fact that for now, it doesn't work that way seems to mean that basically, UA has no teeth.
Witnesses as initial source of reputation.
It seems a nice idea to have witnesses that are the initial source of the reputation as "trusted" users. Given though that a few witnesses don't exactly have a clean rap sheet when it comes to beening upstanding steemians makes me think they might do better. I think the source of trust should rest with established real curation initiatives like @curie, @utopian-io and @steemstem.
Algorithmic curation
Fake curation is the problem, not the solution. Bid bots messed up the existing rep system because basically, they do false curation. Thinking that creating another false curation system seems, well, rather optimistic.
Why not flagging instead?
The big problem with the current rep system isn't going anywhere if delegated SP gets used to reward. It might though if it was used to punish. What if the UA delegations would get used to actually undo or at least attenuate a bit of the damage done by the problems with the old rep system? The idea would be simple: Follow flags by high-UA users. Turn high-UA into flaging leverage so that high-UA users can help attenuate the high-rep poor-content problem the bid-bots are causing.
UA seems very resource intensive. Do you think it should be included at the blockchain level? As far as I understand things, that would mean all nodes which would have UA enabled, would need to have enough resources to run it, and that might be an issue.
The current solution however is not the best either. What if UA becomes highly used on Steem, but the guys behind it cannot support the system anymore, or something goes wrong at their end?
Thank you for your ample response. I would say a little more about some of your other points, but I'd rather see if we get more answers, and I'm trying not to influence them.
I think I'll also wait to see if you get more responses first. I just wrote some other thoughts on the subject, two ways in what I think @steem-ua could be that much better (bringing in @freezepeach and replacing algorithmic curation with algorithmic vote-follow).
Yes, I'll wait until the week-end to see if there are more comments. Then, the prize will be awarded! :)
I gave my two cents on your post.