You are viewing a single comment's thread from:

RE: 2 Problems Plaguing Steemit That Synereo Could Potentially Solve

in #synereo8 years ago

I personally think its not how many upvotes you get, but what power does the upvoter hold. Its a little unfair in my point of view upvote is an upvote right? Some people end up with $2000 in just 80ish upvotes while in your case 350 upvotes gives just a ~$150. System is generally unfair why not each of the upvotes gives a worth of $1 and the more u are upvoted the more that $1 worth so the last 350th vote would not give just a $1 ,but somewhere close to $20. Then again it would be totally unfair towards whales "invested millions and still have a voting power of $1". I don't really think there is a right solution to this yet.

Sort:  

"why not each of the upvotes gives a worth of $1 and the more u are upvoted the more that $1 worth"

It's simple: If people made hundreds or thousands of fake IDs, it would be easy to game the system. Then someone would complain on how is it possible for all these fake votes to game the system. This type of problem is solved through the invested vote.

Very thankful for the patience of those around here who continually explain advanced game theory dynamics and sybil attack realities to those who are uninformed. I've been working hard the last month to catch up, and I greatly appreciate this effort of education.

when I saw someone responded to this, I figured it would be you @lukestokes, even before I looked at the username. That's the whole reason there is a need to sign up through facebook or redditt (I believe that is still true). Steemit is looking for unique individual users who will engage and use the system, not total number of accounts.

I guess people could keep making email accounts and facebook accounts and sign up for the steem power and only vote on one users post and eventually power down and transfer to one account, but most trying to game the system are looking for a quick buck and not willing to get their money over years.

I know how to game systems pretty well. It's just how my mind works. It's gotten me in some trouble, but I can now use it to point out bugs or exploits to help a community instead of just myself.

Thanks @bendjmiller222. Even then though, 3 Steem Power isn't going to impact much of anything. That's the whole point of this design. It's those who are VESTed which get to influence the network. Having 100 accounts at 3 Steem each is still only 300 Steem Power. Not exactly a whale and certainly not an influential vote, even if combined. The white paper gets into this stuff in detail, but I guess many haven't read it or thought it through.

Regarding defences against this kind of gaming, such as with bots, for example, well, first of all, flagging works effectively to push them off the page with subzero reputations, but I think that some kind of minor PoW that cannot be accelerated would limit user post rates, same as what HashCash (the precursor to bitcoin, in fact) and Bitmessage do, precisely to rate-limit spamming. Bitcoin uses it to rate-limit new coin issuance. It works as an inflation controller, but as Steem proves, you can direct inflation to useful purposes. Dash also demonstrates this principle as well, by redirecting it to fund development and platform marketing. Steem doesn't need this because the users themselves are the biggest marketers, and right now, playtesters, and maybe eventually the idea of this being a governance system to regulate developers may become integrated too.

I have read the white paper multiple times actually and

when I saw someone responded to this, I figured it would be you @lukestokes, even before I looked at the username.

This was actually meant to be a compliment (rereading made me realize that it may not have sounded like one).

You are correct that one hundred accounts at 3 steem power is not a very big account, but you could easily outsource and pay x amount of dollars for people to make fake facebook accounts and slowly build up a following of a lot of micro accounts voting for them.

It may not make a lot of money (then again if the price were to boost up, the value on each account could add up to a nice payday for someone), but I believe is clearly not the way steemit was designed to be used.

Steemit has added a lot of features to prevent abuse, but you are absolutely right in your reply to mine that a bunch of small accounts most likely won't be able to do much unless they scale up their micro operations.

Content is really what should be making money and finding content. I think that less steem power could even be given (or none) and make peoples first post worth a little bit more in some way. Just an idea.

Content is really what should be making money and finding content. I think that less steem power could even be given (or none) and make peoples first post worth a little bit more in some way. Just an idea

Exactly, it will be the great and quality content which will bring in fresh writers. why not give more weight to writers who have pulled more people into steemit based on their content? also aka the referral.

I understood it as a compliment and attempted to thank you but accidentally typed "Then" instead of "Thanks" (which I just edited). Sorry for the confusion.

That's an interesting idea... your Steem onlocked on your first post. Unfortunately, I think Steem is required in order to use bandwidth on the network to post, but I'm not sure how any of that works yet.

excellent way to put it, no need to get made when some one gets 500 votes and makes only $100 while others get 50 votes and makes $4000 its just who votes you. Minnows have no weight in the vote so having 490 votes from minnows means you get close to nothing. its just how it works and its perfectly fine, of course we all wished our content can make more then $500 a day... I know I would but that's just not the case. Build your reputation by providing great content and you will get noticed. Unless you made your self known outside of steemit and come here... I always say "your success travels with you" - @carlidos

This isn't a problem at all, why not wait till its a problem before fixing it so you don't go too far in he wrong direction.

It's a problem right now, as some people control thousands of accounts:

https://bitcointalk.org/index.php?topic=1594979.msg16019000#msg16019000

1 person = 1 vote is doomed with such practices...

Well the bigger problem is active user count. Why should Steem power be squared ( I think that's right) maybe the the right formula is to the 1.5 power. This alone would help the minnows and dolpins have a bigger impact.

A vote per person with = weight would destroy the system. Steemit like @will-zewe has said is an investment. There could have been 0 monetary rewards. Look how excited people get when their IG selfie gets a lot of likes. By weighting votes you allow large investors to have more say in what they feel is the most valuable content. If all that were needed was a vote, i would go to every single family member and tell them id give them a quarter for every vote they make on just my content and keep investing in myself and not worry about others content one bit

You are assuming the high invested vote actually reads content. They actually use voting bots, and if your on that list, your in the in crowd. If not, sorry.

I am not assuming this. This is another type of investment. Whales, like everyone else have a limited voting power per day. They have made these auto votes to reward creators of good content and I have seen people have whale bots removed. Also if you look at @smooth or @berniesanders (both of who have voted for me and I am not on their bot list if they have one) you could see that they spend time engaging with others. Some whales have a bot and then at the end of the day read over the content. It's simply a different type of investment.

While I would hope that my content gets read, I would be happy to have created content that is deemed good enough to be put on the bot list. Check @sirwinchester. He didn't start with any bots on him and he made a big splash. I'm happy making the content I wish to and getting rewarded. I don't cater to any one audience and if that makes it so I am not on an auto bot list so be it.

I love the community more than the money here.

Yeah i'm a huge fan of smooth and bernie, they both engage and communicate, plus I'm pretty sure smooth is a huge video game fan which is totally right in congruence with myself. Smooth is probably my favorite whale not gonna lie.

Flag bots for capcha. Their vote doesn't count unless capcha is correct.

It's the design that is flawed. It's needs to be re-designed, obviously.

I don't know how to phrase this diplomatically, but I think your issue is that you were benefitting from the system, but now are not.

There is a fixed pot of steem that is generated by the blockchain every 24 hours - about $200,000 worth I think at the moment. If one person gets huge rewards, everyone else must get tiny rewards (as in $0.05 a post, especially if it's a growing platform and lots of people are posting. If it has gone from about 5000 users to 70,000 users it is not possible for a handful of people to continue to earn huge amounts and everyone get even less, because remember it is a fixed pot of steem. Something had to give.

The devs made a small change - they allowed the whales to choose on a sliding scale how much their vote would give. Before their vote was huge and they could only give 27 people a massive amount and lots of people nothing. Now they are trying to distribute the funds much better.

In addition, as the platform has expanded new writers have come in. The whale bots curate based on profit. If they've removed you from their list it is because they have found another more profitable writer to upvote. That is the nature of this platform.

The amount of Steem is fixed, the value of Steem is not. If Steem grows to 70k users, the value of 1 Steem will be higher and even so will the total reward pot of Steem in USD be higher.

Why dont start to pay a max. payout for each post, lets say 300$ for example.... more people would benefit at the end, guess it will be way more fair for everyone.

I didn't know about this modification, I had seen mention, but not a clear explanation. It may well be that I am wrong about curation rewards and they can become a net value add. Possibly some kind of ranking that modulates how the whale can win them would eliminate abuse potential, making it really a proper competition to see who can see the best stuff early on. The goal of developing with this AI algorithms to pick this then would make more sense.

Anyone who has more than 30,000 Steempower can now select how much to give - say 33% instead of the full value of their vote. That's why the really big single upvotes have stopped and you see a lot of smaller amounts given.

Yes it was added a few weeks ago to the Web GUI but was always a part of CLI based voting. When you use the CLI you give a percentage to indicate how much of your vote weight you want to give at the time. For the GUI it only applies to high value accounts from what I have ascertained (don't know the cut off) - they have a little slider like a volume control to decide how much they will give each post.

Ah, this explains why the multi-thousands rewards have stopped, and a lot more stuff is apppearing on the trending page.

If I may interject: why the discrimination, why do only "whales" get to use a slider?

There is a minimum vote weight accepted by the blockchain at all to prevent spam. If you have low SP and you you reduce your vote power further with a slider your vote won't be acccepted at all.

Very interesting take. If the price is right and the people want to get paid more, maybe some sort of affiliate marketing or advertising system could be set in place. Money will not appear out of nowhere and people want to invest in something that has the potential to have many eyes on it. I hope that steemit is never publicly traded because that will kill the community and instead focus solely on the bottom line.

People also need to realize that since steemit is based on user content, those with a large stake want steemit to succeed. If the whales decide to vote on content that others feel is not worth the price they may move away where content is rewarded more for substance and that will make them lose a lot of money as investors.

So grow the collective expected account values using creative development and marketing?

The amount of users on Steem has nothing to do with the value of each user on Steem. There is something more going on.

https://steemit.com/steem/@dana-edwards/are-we-over-valued-or-under-valued-on-steemit-a-steemit-blogger-is-currently-valued-at-usd25-878

The pot of steem is fixed in a 24 hour period. And in dollar terms the amount distributed in any 24 hour period depends on the price of steem.

At present about $200,000 worth of steem is being distributed to 70,000 users. Of course not all of them post, I think Ned said 21,000 posts are being produced per 24 hours. If it was distributed equally, each post would get about $9.50. Of course the distribution is not even, it is done with the Pareto principle.

A month ago, when the price of steem was four times what it is now, you had $800,000 worth of steem distributed in a 24 hour period amongst a lot fewer users. I think it was 30,000 users at the time.

It is simply arithmetic - if you are distributing a fixed pot amongst a lot of people, then everyone must get less. If you are distributing a shrinking pot amongst even more people, everyone must get even less still.

Stellabelle was moaning that she wasn't getting paid the way she was when steem's price was four times what it is now and it's users were half. She's effectively arguing she should earn the same as she was and everyone else should sink to fractions of a penny to pay for it - in other words she wants a higher % of a shrinking pot to maintain her dollar amount and by definition she wants every one else to take home zeros to pay for it. That's a dodgy attitude to take particularly as she's couching it in terms speaking for the majority, LOL.

If you want individual posts to earn as much as they were and an expanding user base, then the price of steem must rise proportionately to pay for it all.

You achieved #1 ranking, you've earned an enormous amount here, and now you're throwing a tantrum and declaring that actually you've been just hanging out at a party in a drug dealer's basement. I'm sorry for any pain that may have come your way, but this approach seems unnecessary.

I don't really think it matters whether she's made big money or no money. If there are safety issues, then they need to be addressed. It often takes someone who has standing to truly draw attention to a problem. So good on her for speaking up.

Fine, address safety issues.
Does the #1 ranked poster need to sh*t on everything on the way out?

She didn't say she was leaving.. she was pretty clear that she will be exploring the other system and writing about it here.

Nothing is flawed about number of votes being meaningless. If you have been on other sites with rampant purchased followers and sock puppet bot voters you should understand this. IMO the biggest flaw is that the count of votes is displayed at all. Voting power = SP, not vote count.

@smooth In a sense, you're correct, only invested users should have meaningful weight on content curation; however, keep in mind that the stake now may not accurately reflect user investment. In the long run, the rewards system is intended to align SP distribution with reputation (people who provide real value in the system outweigh trolls), but that self-balancing effect might be too slow. Eventually the number and quality of whales will be quite large, and they'll have sufficient bandwidth to upvote all kinds of good content and downvote the crap, but right now we're still working on attaining that level of decentralization in the content curation.

It appears that the decentralized social media competition is now on, and if Steem is going to survive, it needs to keep up. Part of that might be figuring out how to grow valuable users' relative SP faster. One obvious way to do this is to decrease the degree of dilution mitigation SP holders enjoy, which will require SP holders to create comparatively more value to maintain their slice of the pie, and instead allow the high-value minnows to grow their relative stake faster. I'm not saying this is the solution -- I haven't put enough thought into it -- it's just an idea off the top of my head.

@stellabelle is pointing out real problems. Let's not let our excitement about Steem lull us into complacency; we need to fully understand these problems, and act to resolve them at their roots.

Designing a content publication platform with an economically-driven curation model is still an open problem. Will the current algorithms move the SP distribution in a better direction? I don't know what an ideal distribution would look like, but certainly some distributions are better than others, and we would do well to figure out what makes for a better or worse distribution, and ensure we're moving in the right direction. Even if the current algorithms are moving in the right direction, are they moving fast enough? If someone else can create a competing platform that moves in the right direction faster, they should rightly eclipse Steem, as that would be a superior solution.

I like Steem, though, so let's try and beat 'em to it.

@modprobe, my comment was narrower than that. I'm not suggesting that there can't be improvements, only that raw vote count is meaningless and has been consistently and rampantly abused on every other such platform, and those are platforms that don't even award money. With direct monetary incentives the abuse would be even greater.

So I agree with the broader issues you mention being considered and explored, I just disagree with statements made on the basis of "how many votes".

1000 minnows probably couldn't move that post.... but then again 1000 views on youtube doesn't pay the creator either.

+1

As a redditor for 5+ years, most votes are bots and the number is meaningless. This is the same on Steemit and I completely agree with this. Engagement is another story though -- there could be some economics built into posts that drive engagement (right now there's not).

This is where my ideas about groups and integrated chat systems helps. Especially the latter, because it is a place in which like minds can interact in real time and spin off ideas, or even just to organise external activities that base their economics and accounting, and put their marketing, on the Steem blockchain.

This post is past the initial reward period, and I didn't actually specifically write it just for rewards, I wrote it because I want to help people understand Steem:

https://steemit.com/ascensionteam/@l0k1/steem-is-not-a-blogging-platform-steem-is-an-ad-hoc-public-and-transparent-corporation-let-me-explain

If these principles are refined, expanded, and integrated, I think the solution of engagement is solved. A lot of what drives engagement cannot be directly integrated into Witness nodes, or easily quantified, and the mechanisms of users using Steem as their Department banking system and investment pool gets around this issue, so you can market, bank and account your little enterprise's activities, and by the simplicity of also doing so, the integration of Steem into the enterprise's accounting and fund management system, will naturally anchor more vested and held funds in the platform, where it continues to maintain the ability to scale up and keep rewarding good ideas and good content.

The problem is attention scarcity and limited means to delegate the human attention resources (human computation) in an orderly fashion. It is in this area that Synereo will excel with their design while Steem can adopt delegated human computation but it's not currently designed for it.

So you can simply let whales delegate their voting power and the problem is solved. I outline it here: https://steemit.com/steemit/@dana-edwards/time-to-decentralized-curation

The reward system itself is fine, but there needs to be more human attention, from a greater variety of minds, to distribute the Steem Power via curation. Randomization can help provide for that variety of minds.

So you can simply let whales delegate their voting power and the problem is solved.

Not even close. ;)

Maybe the function rewarding sp is too steep. I believe the function is based on squaring, why not based on a smaller exponential. Bots are not a problem unless not armies are making worthlessness ostensibly of over. Dollar or two. I honestly don't think this happens . If it does just penalize by the percent of votes that go to the same author like the thirty minute rule. If 100% of votes are one author then voting power drops to 1%, if 50% then 50% and if 1%?or less go to same author you get 100percent voting power.

The squaring function of SP is largely a misconception. SP is linear-weighted between accounts. 5 accounts with 1 SP each voting on a post is exactly equal to one account with 5 SP voting on a post. The squaring comes into play when comparing posts against each other. A post with only a small amount of SP voting for it gets essentially no reward at all, and the reward rapidly climbs as the amount of SP voting for it (whether from one person or multiple people) increases. This is useful because otherwise you would have thousands of spam posts getting paid each just getting a few votes from each of those thousands of bulk accounts that @alexgr just posted about (vote limits per time does not work there).

So content is not judged by votes but by the SP (steem power) holders.

Correct and the answer is to decentralize curation itself. The owner of the Steem Power only wants the profits, they don't have to do the human computation if they are willing to accept half the profit.

Content is judged by votes, but 100 minnow votes will not be worth the same as one whale vote.

And you can build SP by writing good material. Which also qualifies you as a quality voter. I think the curation rewards system adds nothing but we will see whether I am right and this becomes commonly agreed.

We already have content arranged in order of number of votes. We only need to look at the Hot page. This page often shows posts with spam votes so of course the trending page ends up being better. But from what I can see at the moment it doesn't look like the posts that are getting the most votes are not getting enough payout.

I'm pretty sure the Hot page was reworked and no longer considers number of votes at all, only rshares (essentially SP) and recency. I could be wrong, but that was my understanding from what one of the devs told me.

It's not votes count only that needs to be hidden its everything, dollar amount and author included,

I disagree, I think that the biggest flaws on Steem remain in the interface code, not the blockchain. On the blockchain my personal bugbear is the curation reward system and flagging, I think both need to be scrapped, because the first encourages vote-gaming and the second encourages reputation-bombing by people with a lot of money to instantly upgrade their vote downgrade power. Take these two things out, add a group management system for moderating membership, then create a distributed chat system that functions like a lower latency version of bitmessage.

With all that in place, I think that my idea that Steem is a transparent, public corporation, it's somewhat different from being just a social network or blogging platform, it becomes a literal, functioning, living breathing, ad-hoc corporation in which everyone involved is incentivised to add value to the platform and make it bigger and better. Well, bigger impact, not necessarily complexity or excessive execution time for code...

On the blockchain my personal bugbear is the curation reward system and flagging....the first encourages vote-gaming and the second encourages reputation-bombing

Exactly these are big problems.

Take these two things out, add a group management system for moderating membership, then create a distributed chat system that functions like a lower latency version of bitmessage.

I don't know if it would definitely work but it is worth a try.

One other additional thing that I think might help (and I know it is controversial) is rank all posts according to the number of votes only and not the payouts - it may even help to hide the payouts as others have suggested.

Obviously the biggest problem with this would be using sybil accounts to upvote your own posts to get higher visibility. If someone could figure out a solution to that fatal flaw it would be fantastic.

This idea of ranking could be added as one of the members of the list 'hot', 'trending', etc. 'most-voted' would be a good name. Because it is vulnerable to sybil attacks, it would not be in the top of the list.

Regarding payouts, I think they should show more information, not less. The estimated current value, plus the curator share/author share, and perhaps even the SP and SBD that could be awarded.

This idea of ranking could be added as one of the members of the list 'hot', 'trending', etc. 'most-voted' would be a good name.

@l0k1 Actually that is a good idea - and using an appropriate name that explains that it is vote based would help.

Perhaps also having some kind of reputation ranking of the votes would be useful - this would need to be based on some kind of percentage based mean. It would need a suitable name - I was thinking "Voter Ranking Index" but that is terrible!

Regarding payouts, I think they should show more information, not less. The estimated current value, plus the curator share/author share, and perhaps even the SP and SBD that could be awarded.

That could work too. I don't think people like hiding information and the other thing is even if it was hidden people could just use a tool to add it back in from the blockchain - I'm sure someone could just create a browser extension to do it. I didn't think of that until now.

@l0k1 - there's a WHOLE lotta win in those two paragraphs.

I really don't think its flawed as I stated above, lets say you get 500 votes and 490 are from minnows those minnows have no weight value so you wont get much in $ but if you get 50 votes and they are all with much higher weight lets say over 5,000 steem power of course your going to get more $. That is the way i see it. the weight of the vote is what matters. not how many votes.

One vote per person wouldn't encourage people to buy steem. Steem has no value then. The more steem you have the more votes you have. This is totally fair.

People fail to understand that the voting system in Steem is like that of a publicly listed corporation. Vote power is based on stake in the platform.

If the 3000 content providers have no stake in the platform, then why would removing all of them cause the entire system to collapse. The system is dependent on them. How is that not a stake?

No, actually, that is a very common misunderstanding. The system depends on investors. Without them there is nothing to redistribute. The free stake you get when you start, gives you nearly no voting power. You have to either invest money, or quality capital, ie, articles that get votes, or you are a zero, and rightly so.

The issues that Steem faces at the moment is that investors can game the curation rewards system, and there is a remaining, serious means by which those with a big stake can harass smaller users, both potentially very serious problems, one we already are seeing, and the second given time will become a serious problem. Dan, one of the devs, hasn't looked at cases where this has happened, but I was directly involved in such a situation. Come to think of it, I haven't seen the whale that did that to me since.

My article, https://steemit.com/ascensionteam/@l0k1/steem-is-not-a-blogging-platform-steem-is-an-ad-hoc-public-and-transparent-corporation-let-me-explain lays it out pretty clearly, exactly what this platform is. This 3000 hypothetical users with no stake they are risking, risk nothing by leaving, and damage the system in no way by leaving either. If they have vote-worthy content, they already have raised their vested SP, and are no longer plankton.

While non-steem related posts are more marginal in their value, they don't entirely not matter either. People want to read good stuff, and people with big vests (whales, dolphins) vote up good stuff because when joe sixpack shows up to read, there's something interesting there. It is not everything, but the more plankon, the greater the chances there is minnows, and amongst the minnows, there is dolphins, and if the user sticks around and keeps up the good work, they could become a whale just backed mainly by their excellent writing skills and ability to predict (or luck, maybe) what subjects may be of the kind of vote-attracting interest that keeps them being able to work as a writer instead of getting a regular job.

But ultimately, the core elements are that people will, in some way, with their posts, induce people to vest assets in the system, and use the forum to discuss the projects that they are involved with, that are profitable, and will be a target location for part of their long term investment or short term banking solution, for their activities.

All of the above concerns are part of the whole picture. People who just watch, don't need to even sign up. The blockchain is public and so is the website, currently the main interface. Steem is not about the consumers, it's about the producers. It's up to those who want to participate, to find ways to turn their interests into something that either wins votes, engenders profitable business that invests in Steem as part of its business model, or more centrally, participates in the central building of the platform's facilities for users, which is another type of value-add.

Steem is not meant to just be a blogging platform, or just a cryptocurrency, but something that blends both together. Its success will entirely, ultimately, depend on who joins, and what they do, that brings more money into invest in the enterprises that users create with its help.

If you were to invest in a company like Apple and buy a share at $100 (rounded) you are important to apple. If just you left, yes there would not be a big goodbye party, but if everyone who had invested $100 left, now Apple has a giant problem. They have hundreds of thousands of dollars invested in them by people with a share or less. But you as a holder of one share is not that big in comparison to 100,000 shares. Unlike investing though, you have the ability to make an incredible amount on $0 invested. But yes the system would take a nosedive if many users decided to just leave. So yes every minnow to every whale is invested, but not equally and that would prevent any investment in the platform at all.

There are no investors @loki unless you think buying power ups in Second Life is investing in it.

@dana-edwards

There are no investors @loki unless you think buying power ups in Second Life is investing in it.

That's a little silly, second life has nothing to do with steem especially in the stake weighting system for votes.

Investors are investing in both the assets built into steem and also into the incentives for quality articles, and curation. And I think that it should be pretty clear by now that groups are on the agenda of the developers now, since I have given a clear illustration of how it helps defuse the problem of centralisation of attention in the trending and hot feeds, and would allow users to specialise in subject areas.

It's investing it's not fair. Invest some money if you want a higher upvote weight.

Last i checked @stellebella was worth over $200,000

I dont have a half million

In your mind, who is the arbiter of what is "fair" or "unfair"?