RE: 100 DAYS OF STEEM : Day 33 - Tackling Abuse on Steem - Part I - What is Abuse?
I think we should probably eliminate self-votes. It's usually considered somewhat rude to "like" your own posts on other platforms, and the fact that money is involved with it here makes it seem especially weird. I think that the fuzzy rules around when it is or isn't appropriate are a turn-off to a lot of people. (It's true that people could get around this by creating alternate accounts, but the existence of a counter-strategy doesn't mean we should avoid picking the low-hanging fruit).
I don't have any solid definitions of abuse off the top of my head, but I think one thing that could potentially be a factor is scale. In the offline world most people think it's fine when the owner of a small company gives a family member a job there, but it's a much trickier question if the CEO of a fortune 500 company does it. So giving your friend a max-strength upvote on everything they post regardless of quality may be no big deal if the average user does it, but might not be cool if you're powerful enough to direct a substantial fraction of the reward pool with this kind of behavior.
Some more points I thought of: One of the "milking" methods appears to be focused on upvoting comments rather than main posts to be more "under the radar". It might be worthwhile in the long term to change the reward algorithm so that rewards on comments are capped to a reward that's some fraction of the rewards going to the main post they're attached to, to limit this sort of behavior. In the short or medium term a downvoting bot might be able to enforce a policy like this.
And I don't think this is really an urgent matter, but something to think about long term is how voting can potentially be used as a tool of harassment. Right now we have some persistent downvoters that seem focused on arguments or differences that have happened on-chain, but if we think about mass adoption then we're likely going to need a story for how to deal with situations like people angrily downvoting their exes after a bad breakup, for political reasons, for group-based hatred, etc. Harassment is generally "persistent unwanted contact", so people could argue that having a downvote bot sicced on them is a form of harassment, but putting a limit on who you can up- or downvote would cut against the philosophy behind the steem economy. So there are some philosophical issues that may need to be worked out there.