RE: Live with complete freedom, the way to make it big on Steemit!
If nothing else, scams like this invalidate one of the central axioms of the steem protocol:
Proof-of-brain.
As has been described many times, steem is supposed to replace the trade of electricity-and-cooling-for-commodity (found in proof-of-work systems) with the trade of interesting-content-for-commodity, the idea that those who create content and share it on the blockchain would be rewarded in line with how much "everyone" found the content "interesting."
Scams like this expose a very painful truth. "Interesting" is not a systemically emergent property from focus of activity.
All of these accounts focus quite a lot of system activity on some really terrible content. As you've discovered, vapid commentary – not even blog posts – are perfectly valid loci of activity, which the system as a whole treats as interest. Interest equals more likely to receive part of the random payout. More likely to receive part of the random payout equals a steady income.
The problem is the assumption that steem activity actually denotes interesting content. I'm not sure that the system as it's conceived can work in the way it is generally described because of the way that creation content is decoupled from any sort of real valuation.
Because the value is assumed to be global and all actors are assumed to be seeking the same definition of "good," all activity is assumed to be equally good.
Clearly not the case. Bots, which never sleep and are much more reactive to new information, make it easy to vote up (or down) posts based on any sort of arbitrary metadata with the actual content being of no interest at all. Feedback loops of delegators make use of the easily observable economies of scale when it comes to maximizing your potential of catching some of the random rain of dollars from on high.
I don't actually have a solution for this problem. I'm not even sure I have a really good suggestion – and that part bothers me. It bothers me a lot.
Playing whack-a-mole with these scammers is always going to be a case of trying to follow the trail and knowing that you're only catching a small percentage. I'm sure there are probably hundreds, if not thousands, which aren't making enough ripple to attract attention.
I wish I had an answer.
This is the most important part of what you said. I feel like I am in the same spot. There are many suggestions I have but in the end, it's like fighting spam and security threats. It's an endless game and you have to win 100% of the time, they just need to win once to be successful.
The blockchain will evolve much slower than bad actors.
All systems evolve more slowly than the parasites which live upon. If we've learned nothing else from evolutionary biology and simulation, we've learned that. Put a pile of self replicating bits into a primordial soup, turn on the sun, and let the system have a bit of mutation and within the first 15,000 generations you have a fully flourishing parasitic and meta-parasitic population.
We knew that just from playing with Ray's Tierra simulation back in the early 90s.
I'm coming at this as someone who used to be, once upon a time, in the software/security threats business but who moved on to something that suited him better: game theory and game design. And from that very different perspective, I can only say with aggressive emphasis:
You get what you reward.
The steem protocol rewards activity at loci on the blockchain. And that's all it does. The assumption of content value at that locus is completely assumed and arbitrary. The designers assume that there is interesting and useful content there because "people" are active there – but that has nothing to do with the actual content.
In my gut, I feel like the problem is the generalized assumption that everyone values the same content.
If the system differentiated between locations on the blockchain that I am active in and locations on the blockchain that you are active in and locations on the blockchain that the bots are active in, then we could make reasonable decisions about the value that we ascribe to each of those potential states. You might think that locations where I'm active are likely to be interesting and that locations where the bot is active are less likely to be interesting. I might think that locations where your active are likely to be interesting and where the bots are active are likely to be interesting (and let's be clear, this is just for the sake of illustration – because that would be insanity). What the bots think doesn't matter to either of us.
It's the lack of individuation that seems to be the attack vector on this system. I'm not sure if individuation is a thing that can be brought to the network as a whole in a way that retains the current set of expectations that a lot of users (not necessarily ones that we would like to associate with, but a lot of users) currently hold. Many of those users have significant stakes in comparison to either of us.
I really would like to solve this problem. It seems soluble from a game theoretic point of view. Viewed purely as a game which involves a series of exchanges between players, this is a thing that we can make interesting for everybody.
I'm just not sure if people want that.
Funny how people will always find a way around the rules...he has great pics, though....
Any set of rules has an exploitable architecture. Even the simplest. And in a situation where min/max is a value that can be optimized? Someone will do so, and they will profit by doing so.
That's an inevitability of systems. Orders of magnitude more if we are talking about "money," even money in the abstract. Just the idea of money will find people working very hard to get some of it. (Chinese gold farmers and World of Warcraft come to mind.)
We'll have to see how things shake out.
Im leaving them for after the article ;)
I think part of the rewarding pool should go to the most viewed and/or replied posts. Because it reveals how interesting to the platform it was.
The problem with that particular rule – and I would generally support it under other circumstances – is that there is no way to determine the difference between a human looking at a post and I bought looking at a post. We already know that bots can post new content and comments; we see them doing it all the time. Driving the reward pool to "most viewed" and/or "most replied" would drive the bot swarms to sell views contentless commentary, which is already happening to some degree thanks to mutualism groups who vote up each other's comments, either via automation or manual.
Again, I think the core problem is trying to figure out "how interesting to the platform" a thing is, where the real question is "how interesting to an individual user is it/was it/could it be?" We have all the signals we need to start developing an answer to that question.
I'm still not sure it's a question that the platform wants an interesting answer to.
I agree. Obviously the drive-by upvoters who just use the list to try to earn rewards aren't proving anything. Oh well, I'm a newbie minnow, so what do I know?
I don't get why we can't just flag their accounts off the platform?
You have to sacrifice rewards to flag, and most of us don't have enough stake to make a big difference, which means either a big whale or tons and tons of minnows have to do it. Plus voting power is a big limitation unless you have enough SP to be able to use small votes to accomplish it.
Me and a few Steemians discussed this yesterday on Steemit.chat. Is there plans to reward down voting? We think there needs to be, if down voting is not incentivized very few will ever flag.
There are talks about it, but currently no.
I think down vote rewarding and account creation are the two problems plaguing Steem's growth. Hopefully they'll soon be solved.
Would there be a way to limit the withdrawal, not only in the power down...but in how much could be sent to individual offsite wallets or accounts within a time frame?
I am not sure if I am saying it correctly, but if there was a way, then these multiple transfers to the same bittrex wallet could have been seen, and halted?
I know that would stink for people who just needed to shift funds for legit reasons...but if it could be done for a while maybe word would get out that steemit is not so easy to scam.
Or, to have things set up so that users could not transfer any SD they earned on steemit for x amount of time. Transfers off site in the first 6 months limited to only funds they personally invested into the platform. This way if they want out, they don't loose what they invested...but it would sure halt the pool draining scams.