You are viewing a single comment's thread from:
RE: @Whales, @Witnesses, We have to talk!
Great detective work.
The trick is coming to consensus and agreement on what should be done about it. There are many different opinions on the motivational psychology and game theory aspects in play here, and it's not at all clear what the most ideal system is. Things will always be exploited, and those who care will always have to stay ahead of the arms race to do something about it.
There was once an idea (which got rejected) to allow some users to negate other users. Maybe we need something like this to stop abuse? Maybe a downvote bot we can easily use and more readily-accessible reports about payouts on posts with low view counts?
Yes, more open access for whales to countering bots would go a long way. It's one thing to be expensing SP to fight abuse it's another to spend so much time clicking little flags manually. @personz has done something great but it's still not convenient enough for non-tech savvy people to use.
So it turn out that @freedom might be more about money than freedom after all and thus something like @free-doom might have been more appropriate. Interestingly a lot of clues seems to point to Steemit Inc as being the possible owner of @freedom, not that these clues prove this but strongly suggest so.
I had noticed @freedom delegating his SP to some vote sellers earlier this week. I'd say this is a huge issue.
I was in favor of Dan's idea from the very beginning and I'm still am.
I also think that a none linear reward curve would benefit Steem. It might seems conterintuitive to think that none linear reward could help on this matter but in the long run I think it would.
None linear reward incentives people not to split their votes and SP. It also create a strong incentive for people not to sell their Steem. The price of Steem should in theory increase as the number of users increase and even more so if reward curve is greater than linear.
All of this incentivize people to empower one another more directly with votes creating more meaningful connections overtime as opposed to the current situation where people are incentivize to lease their SP as much as possible.
Dan has shared very valuable insights on the subject in his post "Evil Whales", more precisely under the title "The value of Consensus" and "Curation Reward".
Maybe a sigmoid function as suggested by @twinner might be interesting as it would make self-upvotes less attractive but at the same time would prevent the incredible high rewards for some trending posts which resulted from the n^2 curve.
Yeah I was looking into that @freedom account a week or so ago and I agree with you. It must be an in-house owned account. If not directly owned by the high ups it was given access to exist and feed from the beginning by the high ups.
SMTs have no hope of ever becoming mainstream when this crap exists. What do they think, that people are going to blindly throw money in and not exercise due diligence. Some might yeah... but a fool and their money are easily parted.
Anyone or any entity looking to invest in steem will not proceed when there are issues such as these. The players that do proceed will simply follow the same greed routine. Thus destroy the system even more. Ned needs a good rubber chicken slap to the face.. Wake the fuck up @ned.
First I advocate against the use of violence whenever possible even of the written form thus my opposition to any slap to anyone at this point.
I didn't say @freedom was owned by Steemit Inc but I tend to think so for the reasons mentions in @paulag's post.
The fact that the Steem are concentrated into the hands a small group of people is known and must be consider by all investors. Highly concentrated coins are a negative point when investing but other positive point can make up. Current investors who think this Steem is too risky are selling.
Come on man don't read so deep into that there. I do not advocate violence either, and what I said was more figuratively. Which is why I added the 'rubber chicken' as I was aware that just saying "ned needs a slap to the face" IS harsh in words, even if used figuratively.
It's baffling how they let these things slide. There is clear corruption. If the exploits only exist at the member level, it's bad and needs fixing. If there is a direct link to Steem employees and founders, then... something like that will bring the house down eventually.
I mostly agree to what you said.
It's hyperbole, dog.
HardF***k 19 , yeah that really did a lot of good right?
Interesting connections @teamsteem
a lot of people talk off the cuff and don't seem to realize that much like "The WIzard of Oz" there is a lot of bullshit behind the curtains
Hardfork 19 has made a huge mess in my opinion. It was better before it, but i wasn't as savvy back then, so I could be wrong. Sure seems like the "issues" have escalated to subscriptions since then though.
I'm thinking communicating better with our whales from Steemit, Inc. would be best.
Something like how when you get a comment reply, and it notifies you....
Once or twice a month, Steemit, Inc could put an alert on their GUI which needs to be read, which talks about "issues" such as this, and makes recommendations.
Socially, we're all confused on what is right and wrong. It's easy to find the obvious wrongs, but some of these services which are well intentioned do abuse the system.
For instance, maybe we need a hardfork which says, steem power delegation can't be used to upvote the same user more than once per 24 hour period. That would slow down some of the self-upvotes, and upvotes of your "other" accounts.
Do either of those resound as possible short term solutions?
The reason why this exists, is that it's a difficult problem. sigh.
Your proposed hardfork sounds interesting, but it sounds difficult to implement unfortunately.
I just made a comment over on the latest @steemitblog posting asking about the 2018 roadmap. If there is a way for every user to only self-upvote themselves or any account no more than 10 times per 7 day period that would work well.
...but I know exactly what you're saying. It would involve expanding the chain to constantly save and retrieve this data.
The only time it becomes important, is during the payout of a post or comment to do the calculation so you could nullify the votes automagically by the payout algorithm.
Perhaps this could be done by a huge account, like @ned with a bot which looks at posts "after no more votes are allowed" before payout and then does this research and adjustment.
Of course though, this is decentralized. A round table discussion by developers is necessary though to see if we could solve this in any number of different ways.
Thanks for your comment, you're right. What sounds easy, rarely is... when it comes to the decentralized and autonomous nature of the chain itself.
Banning or reducing self-upvotes sounds reasonable until you realise that an account can just create sockpuppets and upvote by proxy. Of course, with good auditing you can discover these voting rings, but it does take someone to do it in the first place. I'm keen on creating a bot that identifies this sort of stuff algorithmically and then acts on its findings.
You know about @Patrice and @steemcleaners, right? It has done exactly that for months and months. She busts her ass to find and defeat these voting rings.
I'm only vaguely aware of them. I'm looking to automate the whole process. The bot identifies vote rings via an algorithm and then down votes them.
There is a lot more too it than an algorythm can solve. Patrice has been fighting bot rings since the beginning, there is quite a process involved and human intervention often required. Bots are involved too, but its a big operation. This effort might could use your help, but its already widely known and being done pretty well by steemcleaners. They have a discord and you could go find the project witness @patrice and she might enlist your aide though.
Restrictions such as this will always be circumvent in some way or another. You can't get something out of someone by making roadblocks.
To change someone's behaviour you have to make the other legit options more attractive.
I'm sorry but I disagree. In the wider context maybe. But here we're talking about a small number of people whose behaviour probably won't change under any circumstances - more attractive alternatives or otherwise. We're stuck with them until they cash out, like it or not. But the least we can do is call them out continually for their counter-productive behaviour - a big thank you to @transisto for laying it out so clearly. I would like to think the whales, including Steemit.com, who are most vested in the platform, might take the lead here with those of us who believe this matters in support.
@freedom is effectively selling his votes, just like the dozens of bot vote selling scheme on steemit, he does it in a less conventional way which is why its more controversial but all the bots and people using them are doing the exact same thing so it's not a small number of people, it's a large percentage. Wait until we reach the mainstream, the situation with abusers is only going to get worse.
Have any ideas for other "legit options" ? I have no ideas.
The other legit option is to curate properly, unfortunately doing this will yield very little reward..increasing curation rewards would be a good start.
So essentially an abuser has to create 10 accounts (total 60SP or about $270 delegated SP and 2SP) then just spread votes amongst them.
If they have the knowledge to abuse the delegation system then they have this knowledge already.
A hard fork is not always the answer. At least not one that is trivially circumvented by the bad actors
It would definitely curb the "manual" abuser, who has 3 to 5 accounts and they manually click upvotes
What about a discord chat where users can post plagiarism/low quality posts/socs > mods will review and delete false reports and give a comment on the post > whales downvote the post > if mods made a mistake then user will appeal in the chat > higher mods will check it
What do you think? We can also make bots to route whale downvotes. I mean, discord bots.
What @transisto described above does not involve plagiarism. We have SteemCleaners handling that well already and they have channels where you can file reports. That hasn't been the main area of abuse for a long time now, largely because there is community consensus that something like SC is needed and worthy of support. What he's talking about in this post is spamming and sockpuppetry, which have been much harder to address. There's never been any clear consensus about whether/where to draw a line on the latter.
You nailed it exactly with this:
Without a clear consensus, it's not possible to program blockchain-level solutions or to even build bot-solutions. In many ways, we also have to consider cultural differences in opinion on what constitutes abuse, spam, and "valuable" comments and posts. There's a lot to figure out with regards to clear community-driven norms which stake holders and new users alike will support.
We don't have to define what is or isn't abuse to get a fix into the blockchain. We can have solutions that make it easier to counter abuse while still allowing it to be 100% down to subjective stakeholder opinion.
Consensus can't be built if we don't talk about it. Would join a discord chat for whales and senior users if I made one?
Nicely said. Maybe it's time to draw the line.
we need to have a community forum to discuss these matters, and we need consensus. all of the major players need to come to an agreement, not that we all need to agree about what we believe, but we all need to come to an agreement about how to proceed, as a community!
two thumbs up
I updated the comment. And all plagiarists can't be handled by SC. I recently outed one with 200$ payout, a lot of that from your SP and Neds SP.
Before that, I outed another with 70 per post. There is also a zerohedge copy pasta account that keeps on doing what it does best. It gets huge payouts.
It needs to be done on a protocol level, not bots. Bots just consume bandwidth, and it's an endless arms race with the advantage to abusers. Having negation at the protocol level reduces the cost of policing while still allowing accountability for those who engage in policing.
Another good suggestion!
Nail on the head right here.
What are your thoughts on the issue Dan created which ultimately got rejected?
This post isn't about countering flags, dumbass...
I was originally against the idea (well, to be fair, I don't think anyone had much time to fully think about the idea. It was proposed and on track to be implemented so quickly that I think the backlash against not knowing what the outcome would be, was what shot it down in the first place.)
I still think there needs to be a better way, fundamentally. Right now, regardless of implementing that design or not, it costs opportunity to fight abuse, and it leads to a tragedy of the commons (being more short term beneficial to upvote oneself rather than downvote abuse, or if that design is implemented, being more short term beneficial to keep your SP rather than lose it to cancel someone else).
The biggest problem with that design is that it assumes the targeted account only ever acts in bad faith -- you nullify ALL their actions even if some might be "good".
Sadly, I don't know a better solution. The only one I can think of to remove the opportunity cost is having separate voting power pools for upvotes and downvotes, so it doesn't "cost" to downvote (since you have, e.g., 10 downvotes and 10 upvotes per day to use).
Hi.......Help me.. your group my join me pl...
The first thing to do is getting the whales together in a discord chat where we can talk about it. If some whales do something bad, we can just do a harkfork and separate the spammy steem into a new better steem. We would need devs for that though. If we can convince enough devs to defect then bingo, freedom's SP will be worth 10x less.
Do you think that it will take such a drastic step to counter this? I agree it needs to be discussed but don't expect the whales in question to involve themselves in the discussion - at least with our knowledge.
Back to 50/50 curation/poster rewards. Make it more lucrative to skillfully curate than to self upvote/sell votes.
This is something that I recently posted about (creating a downvote bot to target these self-vote rings). It's on the backburner for the time being, but it's something I will eventually get going.
It's also important to keep some perspective, I think. The sums involved (i.e. proportions of the total reward pool) are tiny, as far as I can tell. It's why I shuffled my anti-vote-farming bot project down my to-do list.