"What’s wrong with Steem/Steemit?"
This is a question that I’ve seen asked many times over the past two years, but it has essentially been the topic du jour in recent weeks. This question comes from new and old users, large and small stakeholders alike. It’s becoming increasingly undeniable that there are real, observable problems with the entire social media aspect of the Steem ecosystem.
Something is broken. People have noticed. It can’t simply be wished away.
Many users have offered their opinions about what exactly is wrong and how the many issues can potentially be fixed. I know of several people who tried to get out in front of all of this and have been critical of the decisions that have led to the current mess, who offered alternative solutions to some of the perceived problems that many felt needed to be resolved, and who warned about removing/changing protocols that had a specific and coherent rationale behind them. These people have been vilified for consistently identifying the root causes of the degradation of the social element of this platform and the allocation of the shared reward pool. These people have been vilified for trying to hold others accountable for their poor decisions that have affected every user of this platform. These people have been vilified for daring to speak out against the failed leadership.
Yet here we are today, witnessing the very same criticisms that these people had made a year ago and since then...but this time, those criticisms are coming from other people – some of the same people who were contributing to the prior vilification – and now they are being handsomely rewarded for it in many cases.
Ignoring the hypocrisy and general douchebaggery of many of these now-critical users, let us look once again at the problems most of us are seeing and let me offer my opinions about the causes, then get into how we can potentially resolve them. I expect that this will likely be ignored again and that any solutions will likely never happen, but at least nobody can say that I “never offer solutions” (which is ridiculously ignorant in the first place).
"Why does Steemit suck?"
I’m sure we’ve all seen this question asked – or a question similar to it that tries desperately not to offend the sensibilities of the resident fluffers and cheerleaders – but what exactly does it mean? Are users not happy with the rewards? Not happy with their audience? Not happy with the design and function of the website? Not happy with visibility?
The truth of the matter is – all of the above are a problem. The flagship website (whether we want Steemit.com to be the flagship or not, it is) is horrid for a social media site in 2018. This directly impacts visibility and engagement. It directly impacts the user experience. It turns people off as soon as they arrive, without even getting into the economics of the platform and the relatively high cognitive load of the system in general.
But the website isn’t the main problem with why things generally suck. We know STINC (Steemit, Inc. – yes, it’s a play on “stink,” and no...I didn’t come up with it originally) isn’t interested in delivering a usable social media product. Expecting them to actually improve the website and bring it into the 21st century is absurd. You’re only going to be disappointed. And the website isn’t the root of the problems around here. It’s merely a consequence of one of the many bad decisions made during the life of the Steem blockchain.
The initial distribution – what is referred to as the “ninja-mine” – cannot be undone. The only potential remedy to this would involve burning a large amount of the accumulated supply by the ninja-miners, which we all know isn’t going to happen. But there are still other solutions – other protocol changes that can be made that would directly impact blockchain investment and social media behavior.
(For the record, when I use the word “investment,” I am referring only to speculative investment of STEEM and other crypto tokens.)
So, if Steemit sucks, then what are the causes of it? Initial distribution can’t/won’t be resolved. The flagship social media product won’t be improved. What’s left?
Bots? Account creation exploits? Vote buying and selling? Spamming? Various other forms of vote scheming and scamming?
Again, all of the above are potential issues that we see on a daily basis. Over the past year, we have seen a massive increase in pretty much all of them. And for the past year, the sources for this influx of blockchain and social media refuse have been pointed out. Very few have cared enough to pay attention. Many have actually defended the decay of the social aspects of Steem and Steemit.
Now that it’s the cool thing to do, let us examine the causes and see if we can’t find solutions. And no, I’m not going to sugarcoat anything.
How do blockchain protocols affect user behavior?
Economic incentives are what drive economic behavior. Steem is an economic platform above all else. It’s known as the social media blockchain, but the underlying mechanics are token creation and reward allocation based on how much one is invested into the platform. The incentives offered to us in the form of STEEM and Steem Dollars are what drive us to participate in the ecosystem. If our desire is a vibrant social media atmosphere, then the incentives must be aligned properly with that goal in mind. If our desire is to bleed the rewards dry with mundane, plagiarized, or spammed content, then the incentives must be aligned properly with that goal in mind as well.
The first question is, what do we want?
The way things look today, it would appear that the blockchain protocols have been aligned with the latter goals. There are no economic incentives to be responsible, accountable, or engaging. There’s no economic incentive for having integrity or for creating engaging or entertaining content. Users can spam and plagiarize, upvote their own spam and plagiarism, and then collect rewards without ever receiving a corroborating vote from another user. It’s practically consensus-free rewards allocation from the shared pool and it can be done without even investing a single penny, thanks to delegation and the account creation process currently managed by STINC, which has been exploited time and again.
Even without the account exploits, this is a problem for a consensus-based blockchain and a social media platform. Popularity – whether deserved or not – is the foundation of social media. The entire concept of “viral” and “trending” content is rooted in it.
Reward allocation is supposed to be at least partially based on what is popular on Steem/Steemit. How can something that nobody sees be popular? Why are rewards allocated to it when only the user publishing the content has ever voted on it? Where is the corroboration from the other stakeholders? Why should a single user be able to allocate an essentially guaranteed sum of rewards to themselves or another user with a single vote? What does that have to do with social media and popularity?
Prior to hard fork 19 last year, this type of reward allocation was somewhat mitigated by a superlinear algorithm for post rewards. The general idea behind the superlinear curve was that it would have to take a larger number of and/or more invested stakeholders to corroborate rewards allocation on a specific post, which would then increase the rewards more rapidly as more users upvoted the content, making it “popular” and sending it to the trending page. The larger the stakeholders voting on the post, the easier it was to trend – large stakeholders being the ones with the most to gain or lose by pushing good or horrible content onto the trending page.
Yes, the trending page has pretty much always been “broken.” And yes, the superlinear curve was too “super.” There had been many discussions about the reward curve. Many nights were spent in chats where some of us discussed for hours at a time possible solutions that were less than superlinear but could help achieve the same goals of popularity and still mitigate exploits. In none of these discussions did we ever think that a completely linear “curve” was the solution.
Weeks later, this became the recommendation from a small handful of users and then made its way into the proposed hard fork from STINC. And sure enough, this protocol change was rubberstamped by the top witnesses at the time.
The rest, as they say, is history.
No coherent rationale was offered as a replacement for the previous rationale of the superlinear reward algorithm and the abuse mitigation that it provided. It was as if abuse mitigation wasn’t even an afterthought. Protocol coherence went right out the window with linear rewards and many other changes made over a series of hard forks.
It’s no surprise that shortly after implementation, user behavior changed drastically.
Economic incentives were now aligned with less voting and, in particular, voting mostly for yourself. Linear reward allocation and the 10-vote daily target (making a 100% vote worth 4x as much as prior to hard fork 19) were the culprits for that incentive. This was predicted.
With the implementation of hard fork 18, delegation was possible. And when hard fork 19 came along shortly thereafter, delegation was further and massively incentivized via vote-selling bots and Steem Power leasing schemes. This was about the same time that accountability and engagement took a swan dive into the spam and shitpost abyss. Large stakeholders were able to disengage, lease their SP to anyone willing to pay, and then collect liquid payments for the lease. Those leasing the SP had little incentive to be good stewards of the power they rented, since they weren’t actually invested in the system.
It’s no surprise then that the number of vote-selling bots exploded, being operated by scammers, by sock puppet circles, and by those generally uninterested in the actual quality of the content or the effects of their vote-selling schemes on the system in general. Remove the purpose of stakeholding/vesting in the Steem blockchain and what you get is a system of irresponsible and unaccountable users – the exact opposite of what DPoS is supposed to guard against, in theory.
Delegation provides the means to easily circumvent such accountability and impacts the integrity of the system. Linear rewards make it more lucrative for those participating in the circumvention, thus greatly incentivizing those who can best exploit it. This leaves no other option but to merely hope that the lessors and lessees will behave according to our desired goals, despite no economic incentive for them to do so. And if anyone else refuses to participate in these new schemes, they will likely be left behind as the exploiters take advantage of the misaligned incentives and increase their collective influence.
It mostly becomes a race to the bottom.
I could be wrong.
But am I? I’m simply providing my opinion here on what I’ve seen over the past year and how things have changed from the year prior. These aren’t my complete thoughts, but they generally describe what has happened since hard forks 18 and 19. I’m sure some people will disagree with the causes of the issues here. Others will even pretend that there’s nothing at all wrong and will continue to shill and cheerlead and call people names for having the nerve to be critical about anything at all, ever.
Let them blow smoke and give their handies to anyone in their echo chambers and bubbles. I’m not interested in that. I like to see improvement and progress, not staleness, repeated failures, and empty promises.
Regardless of what you may think about me or my opinions of what and where we went wrong, I know that my critiques are well-intentioned, even if they are blunt or harsh. If you have alternative opinions about the root of the problems here, feel free to share them. But just know that you’re probably wrong. (Calm down. That was a joke. Mostly.)
What are my solutions?
Last year, I discussed many of the problems we were seeing back then and offered solutions for some of the issues – solutions that didn’t involve making huge changes like delegation and linear rewards. Some of these proposals were quite popular among many witnesses and large stakeholders at the time, but ultimately fell on the deaf ears of STINC and any other coders out there.
At this point, there are so many problems with the blockchain protocols that it would be a miracle to unravel the mess and fix the code to realign incentives that would encourage both investment into the blockchain and revitalize the social element of its interfaces. Nevertheless, I’m willing to offer my ideas, fully understanding that they will most likely never happen because there’s a seemingly vast amount of ignorance around here when it comes to economics/investment and what exactly social media is.
And then there’s the fact that those in “leadership” positions have no real incentive to put in any effort whatsoever into actually creating a functional system and/or interface.
Also – these may not be popular suggestions due to the fact that there are so many lucrative exploits/schemes these days that reward doing almost nothing at all, or worse, doing things that are observably detrimental to the ecosystem.
So, with that in mind...
Change the reward curve from linear to literally anything other than linear.
Linear rewards are a mistake. There was explicit rationale behind not having linear rewards in the original white paper. That rationale has not yet been refuted, as far as I have seen. In fact, the readily available data and observable behavior of users demonstrates that it is a failure in practice for the existing blockchain and social media platform. It wasn’t just theoretical. Despite what a “certain asset” has claimed, superlinear was not actually “evil.” It was a somewhat necessary mitigation protocol against exploits and behavior that are now commonplace.
The n2 reward curve was too much. I think almost everyone who was a user for the period of time that it was in place agrees that it needed to be changed. But just as n2 was considered too steep for rewards allocation, completely linear (n) is absolutely not enough to mitigate any exploits. There is a middle ground that can be found and at least tried. Whether that middle ground is n(log n), a sigmoid curve, or nany number between 1.1 and 1.9 isn’t that important, so long as the curve is properly coded and works without compromising the integrity of the Steem blockchain.
More testing should be in order here.
When we see that something isn’t working as intended, it’s incumbent upon us to at least attempt to fix the problem, not ignore it and plow ahead with even more protocol changes. Linearity is contributing to a massive disparity between popularity/“consensus” and reward allocation and it’s making exploits and unpopular behavior more lucrative. It’s about time we recognize this point of failure and actually attempt to correct it.
As stated previously, delegation provides the means for disengagement and unaccountability. It’s completely contradictory to the concept of DPoS and its theoretical ability to mitigate abuse.
I know that this proposition will be unpopular. Lots of people are making lots of effortless money from it and some people are actually putting it to good use. But blockchain protocols probably shouldn’t be devised based on the idea that people will just need to act in good faith and police each other in order to prevent exploitation. It’s not a rational or scalable solution.
If there’s a way to skirt responsibility and accountability in favor of financial gain, then that’s what will ultimately happen. The idea is to not provide that mechanism. The protocols should be written in a way that makes this as hard as possible. Absolute prevention is not reasonable, but we shouldn’t create the vectors for abuse or exploitation. We shouldn’t encourage the degradation of integrity by making it easier to achieve and making it more lucrative.
We have repeatedly witnessed delegatees acting completely irresponsible with their voting power, whether it has been used to simply upvote themselves or their friends, to turn around and sell votes with that delegation to spammers, plagiarists, and what is considered to be “shitposting,” or to blatantly scam other users by promising a variety of undelivered “services.” And no matter how many times it’s discovered and publicly identified, it still continues – even from the same delegators and the same delegatees.
This will only get worse as the platform scales, if it’s not adequately addressed. The best way to address this type of issue is to make the appropriate changes to the blockchain protocols, not by asking nicely for the exploiters to please stop exploiting or by asking spammers and "shitposters" to please stop renting delegation to reward themselves from the shared pool for unseen/unpopular content.
And let’s not forget the elephant in the room, which is the sign-up process from STINC.
Due to the delegation hard fork, STINC has been able to switch from actually paying for account creation from their pool of ninja-mined stake to simply delegating and undelegating stake as needed in order to create accounts indefinitely.
The results have been abysmal. Sign-ups still take weeks and some potential users are never approved (for a variety of reasons). Massive bot-nets (including one that was over 20,000 accounts) have been created and approved by STINC’s so-called “manual approval” process. Accounts that exploit/spam the network are able to do so for prolonged periods of time before delegation is finally removed...if ever.
Accountability is often non-existent.
The question we ought to ask here is, if that ninja-mined stake is not to be used/redistributed via responsible account creation, then why did STINC ninja-mine it and why do they continue to hold onto it? The original distribution of tokens has been a huge vector for attacking and ridiculing the blockchain and its so-called leadership. The excuse for this ninja-mine has always been that the stake was needed for account creation and/or other development and marketing of the blockchain. If account creation no longer represents a considerable cost (and may become even less so as a result of proposed hard forks coming soon), then that stake is no longer necessary for it.
The responsible thing to do in this case, in terms of the overall network, would be to burn it. A considerable amount of tokens that are no longer necessary for normal operations would be eliminated and one of the largest concerns of investors, users, and critics could be alleviated. It’s a win for everyone. But as I stated already – I know this won’t happen. It should and it would if we had better leadership, but we are not fortunate enough to have that here.
There should be no more added benefit for an entity such as STINC to change protocols that improve their potential profits at the expense of the integrity of the network. Delegation has done exactly that. For that reason and the other reasons stated above, it ought to be rolled back.
Change the posting rewards split to at least 50/50 for authors and curators.
The investment token on the Steem blockchain is STEEM. If we want to encourage investment (which increases the value for all of us), we need to encourage the buying and holding of STEEM, which means that we need better incentives for Steem Power and stakeholders. Without the speculative investments, STEEM loses value as a traded token and that consequently reduces the rewarded value of content on the Steem blockchain.
Why do we see so many people leasing out their SP for quick liquid payouts or not powering up their STEEM? Because there’s not enough incentive to care about voting on content for curation rewards. When the maximum amount of rewards you can earn is 25% of an entire post payout (it’s actually much less – closer to half of that, around 12%), divided amongst all curators, it’s no surprise that people move to automation and generally don’t care about engaging.
If we want more and better curation, we need to incentivize it. If we want more investment, we need to incentivize it. And if we truly understand that social media is not comprised of 100% bloggers and commenters – that the simple act of voting on content is actually the largest chunk of interaction – then we need to make it more lucrative for content consumers to easily invest and earn for the quality work that they contribute.
Curation rewards are that mechanism for bringing in more investors, for likely keeping them more engaged, and for bringing in a wider base of users that are not solely here to create low-quality content and not promote it, then hope for big rewards...and then ultimately become disappointed for not earning from their lack of quality, effort, and experience with social media and networking.
Change the daily voting target to at least 20 or more daily votes.
Prior to hard fork 19, the daily voting target was 40 – which meant that a user could cast a 100% vote on 40 posts per day and still completely recharge their voting power after 24 hours of not voting.
With the acceptance of hard fork 19, the voting target was changed to 10 votes per day, giving users the ability to cast a vote with four times as much power. If your 100% vote was worth $1.00 prior to hard fork 19, it would have become $4.00 after acceptance, but you could only cast 10 votes per day, ultimately representing no change in reward allocation if your voting habits had not changed (and assuming that the rewards algorithm wasn’t changed, which it was).
In theory, this would be a boon to those who actually engaged less on the platform. It was advertised as a way for these less/minimally active users to “compete with bots.”
In reality, it simply further reduces the incentive to be more active on the platform and, in turn, reduces the incentive to find more/better content to upvote while also increasing the incentive to self-vote. The effect on bot influence seems to be quite minimal at best (no thanks to other protocol changes already discussed). So, it did in fact represent a windfall to less active users while having essentially no impact on bot influence. One could argue that bots/automation are even worse today.
A lower vote target also increases cognitive load by requiring increased fractionalizing of voting power for users who wish to be more engaged, regardless of account size or experience with Steem and its interfaces.
Despite many well-reasoned objections to this protocol change last year, the rationale was apparently ultimately accepted and the protocol was rubberstamped for approval. It was never explained why less active users ought to have more influence than the most active and engaged users on the platform. The argument that self-voting (made more lucrative by this and other protocol changes) would become more lucrative and prevalent was ignored.
Just as the voting algorithm’s superlinear curve was too much and the completely linear “curve” is too low, the daily voting target of 40 may have been too high (I don’t believe it was) but a target of 10 votes per day is certainly too low. In my opinion, the minimum target shouldn’t be lower than 20 and I think that number should at least be tested.
These proposed changes aren’t THE answers.
There is no simple fix to the current mess. There never is. But the ideas proposed here have a better chance of aligning incentives with users and investors than the protocols currently in place. I also want to stress one thing that I’ve stated many times over the past year or more, and that has to do with content creation vs. content consumption.
The vast majority of social media users are not content creators. However, here on Steem/Steemit, there’s an emphasis on creators – how much they can earn, how their content is “permanently” stored, and/or how their voice matters. Most of the arguments about hard fork changes have been focused on making money for content creators and how creators will benefit from this or that protocol tweak or overhaul.
This is completely backwards.
Active users on Steem have been recently calculated at around 60,000. This includes every account making a blockchain transaction, which – most likely – is comprised of mostly bots, trail voting, and a variety of spam. Daily unique visitors to Steemit.com alone have reportedly been around 250,000 – which represents only a 55% market share of the Steem blockchain.
Even if we were to assume that all users and visitors were unique humans and all users were in fact content creators on Steemit.com, we can see that content creators would still be less than 25% of the total. And this is the best-case scenario for them.
The larger market of users who are incentivized to invest in and engage on the platform is clearly the content consumer. But there’s an even bigger point to make here that we simply can’t dismiss:
Those content consumers represent a potentially and drastically larger amount of investment value than content creators.
If Steem/Steemit was marketed as a place for content consumers to make money for content discovery and that they could get started with their crypto earning from a minimal allocation/investment of Steem Power, would we get more or less consumption/curation and investment? If that curation/investment was better rewarded via a posting/curation rewards split closer to 50/50 (or even more in favor of curation rewards), would we get more or less users willing to sign up, invest, curate content, and actually hold SP long-term? If we had more people purchasing STEEM, powering it up, and curating content with better curation incentives, would we get a better or worse pool of content? Would we get higher or lower payouts for that quality content?
In the effort to cater to content creators the opportunity to reach a vastly larger number of social media users is continually missed. In the effort to cater to those who use the Steem currencies to simply cash out and “earn a living as a blogger,” we add to the downward price pressures of the very currency that they need to earn that living. If we simply offered more incentive to the other 75%+ of potential users that could earn their STEEM by being invested and curating good content, rather than being a creator and relying on post rewards, there’s a good chance that we could see more investment, higher prices, better rewards all around, and – most importantly – a better pool of content creators.
The current marketing and incentives cater to people who mostly have no idea how to create good content, don’t know how to network and market themselves, show up expecting to earn a lot of money, don’t invest anything to help with the money-making aspect of the system, then become disillusioned with the entire platform.
This should probably change.
Please consider this when thinking about some of my proposals above. There is a lot of potential value in bringing over successful social media personalities – primarily for their followings. Incentivize those millions and millions of followers to show up and vote/comment – not to create more bad content – and that could be the key to greater growth, investment, and innovation.
Figure out a way to monetize those eyeballs and the additional traffic from better content creators, then distribute some of those profits to users, and there’s your ticket to a true crypto and social media revolution.