I wish I could share your enthusiasm in automated spam bot detection but our experience trying to eliminate spam is that its not that easy. Having some automated spam bot determine if your vote is eligible sounds a terrible idea to me and one that is easily manipulated by the bot algorithm. Hive is actually much more widely distributed then almost anything else out there; so I am not sure its hurting outside investment (plenty of centralized chains getting lots of investment).
You are viewing a single comment's thread from:
It's not just a spam detection bot you'd put in place. Metrics like transfers, comments, claims, etc etc all need monitored and cross referenced to detect patterns or possible automation of the metrics being used to determine if it's real people or not.
Not saying it's something that would be an easy implementation nor perfect by any means. It's merely a thought experiment on a possible solution. Probably a terrible idea as you state. lol
So far my favourite would be if top 20 couldn't vote for top 20.. We'd see an incredible difference in witnesses and stake distribution if the creation of voting pacts within the top 20 was made invalid. still gameable but easier to detect.
Anything centralized that is getting investment is a whole generational step above our current functionality for the most part. And we're effectively centralized at this point regardless of the hundred or more witnesses. One company hosts over half the witness servers, and the central decision making ability resides in the top 10 active accounts rather than the greater community. As for actual distribution you mentioned.. Eh, I wouldn't consider it that, although lots of accounts do hold it.
No system is perfect, DPOS included. Appreciate the reply and share of thought. :D