Issues With Using Stake To Filter Content
I have been critical in the past about the lack of innovative interfaces on the Steem blockchain being used to present information in unique and interesting ways. Unfortunately, rather than tackle this issue, most applications simply adopt the page filters that are already found on the blockchain. Those being "trending", "hot", "active", and "new". I'm not really sure what active does, but "new" is a stream of spam and "trending" and "hot" show posts that make money fast and posts that make a lot of money. Unfortunately, those properties do not make content good.
Steemit's Fire Hose Approach To Content Presentation
In terms of information presentation, Steemit is dominated by sites like Reddit which has elegant and simple solutions to content presentation. One Account, one vote and sub-communities that develop around individual subreddits. Pretty basic stuff and vastly superior to Steemit, in my humble opinion, in terms of ease of finding interesting, quality content.
But my goal is not to advertise Reddit's superior content, but explore why the current applications are so broken in the ways they present information to the audience. Certainly, we have to admit that they are doing some things right in order to gain the massive following and popularity among large numbers of people.
With traditional media platforms, each individual audience member is treated equally. Each person has a voice and their vote can impact how much attention a video or article receives. With Steem, this is not true. The audience is now filled with individuals with varying influence depending on their stake in the ecosystem. Those with more money have more influence.
Now, there is nothing inherently wrong about this statement. In fact, many will argue that this attracts more investment into the ecosystem. The issue comes when that influence is used to determine the validity or value of a certain piece of information. You then have a huge bias to the interest of large stakeholders and you can largely disregard the opinions of smaller stakeholders to some extent.
Thus, the front page "trending" and "hot" pages represent the content that large stakeholders find interesting, not the content that the community or sub-groups within the community find interesting. Essentially, the opinion of the rich is the only opinion that really matters since articles are organized simply by the amount of stakeholder support that backs them.
So, that's why a lot of these pages are littered with uninteresting content. Because the content is only interesting to the whale who self-upvoted it or the user that spend a ton of money buying bot votes. But there is no decentralized opinion or consensus on the content itself. These pages are simply additional "promotion" pages.
And all of these additional applications have adopted this poor system of information presentation to the users. The end result are cheap looking websites with inferior content, not necessarily because the content available is mediocre, but the good stuff is hidden behind the garbage that whales push up to the top.
Now I'm not saying the whales are really at fault here, but they aren't professional curators of art either. And not only that. If someone wants to organize the information better they either have to develop their own application or spend a ton of money on Steem. And that's frankly not appealing.
Users should not be expected to buy more Steem in order to be able to access more of the content they like. They should only buy more in order to facilitating earning more.
I'm not going to lie that I spend a good amount of time on other sites that have more interesting information. They've had years to figure this out. Unfortunately, applications haven't figured out that interesting content, not money, is what gave these other platforms their meteoric rises and mass adoption.
Sure, content creators and advertisers make large sums of money off of YouTube videos, but the site garnered popularity when no money and expectation for money were on anyone's mind. The ability to access interesting information is addictive and keeps people coming back for more and inviting their friends for the ride.
I'm not inviting anyone to the cesspool that is Steemit. That's not to say there are not some nuggets of gold out there, but they are hidden beneath an ocean of mediocrity that is simply there by a poorly designed interfaces. There is complete disregard for the majority of the audience in order for the few to flaunt their garbage on "trending".
But, I enjoy posting here. I enjoy only reaching a few eyeballs than many. It makes things more manageable and I can spend the time talking to interesting people. Most people are not content with such a grind and would rather use other platforms in spite of the rewards. Just think about that. The interface is so bad that some users quit and convert to other platforms. User retention is more important than total accounts created.
But, complaining is a waste of time unless you are willing to propose solutions and I have proposed lots of ideas over the months on different fixes and adjustments I would like to see. I have also developed alternative experimental ways of showing information like Q-Filter which attempts to solve this problem. I have also given away lots of SBD in the previous posts in order to foster better discussion and show that it is possible to not have one's soul completely corrupted for money's sake.
But that's enough about me. Let's hear what you have to say. What do you think Steemit's biggest issue is? How would you prefer to see information portrayed? This is the last post in which I will be giving away 100% of the SBD of the post and I have to thank the now over 500 (probably 20) followers that occasionally take the time to read what I have to say. Thank you.
And that's how it will remain, as it has been since the beginning, unless people want and agree to change how they vote, or how the voting system works. Quality isn't of much concern for many, I've been mocked for talking about the importance of quality, and those posts that mocked me got big payouts because many power players don't care and don't want quality information. The consciousness of people needs to change for the behavior to change, and that's not likely going to happen. Those with power want to use their power the way they want, and don't want to lose how they want to personally influence the site, regardless of how different behavior can help to improve and make Steemit more attractive.
That's why we need competition. Someone really should fork the Steem network, make some tweaks, make a very public announcement announcing the start of the new chain, and compete. Or someone can build a new blockchain from the ground up if that is better. I really could care less which method is taken, but I want something to scare Steem into action.
This problem is one of greed. Rather than try to make the site better many whales would rather just cash out for short term gains. Sometimes it blows my mind how shortsighted people can be.
Honestly, I'm not too worried, only impatient. It's only a matter of time before Steemit changes or another superior platform comes along.
The nice thing is that the bid bots can't keep up with subsidizing crap content with a ten steem dollar attached to it. The losses on there are quite staggering, the value in them are getting absolutely sucked up much faster than I had ever seen in my short three months here. Bots are showing no value in placing a bid over 40 minutes out from when the timer expires, relatively close when the vote is cast. The crap posts have flooded the market which is both good and bad from the perspective of generating more of an elite level of posting, lol, which I am not saying I am in by any means. Low effort may be a better way to describe a crap post I guess.
Yesterday, I thought about ways of how to improve the feed. I usually ignore the trending and hot categories because as you write, it's about what whales like and about who's paying the most to the bots. So, not really interesting.
The only good category is the new one, because that's where you don't have the money bias in. Instead there is about 80% BS that you have to filter out first, but then you can find interesting content. (Strangely, I still have about the same share of stuff I don't need in my feed. None of my unfollow purges helped so far...)
A system level approach to fix the content filter problem is out of the question. The people in charge aren't interested in an improvement. That's why I thought into the direction of topic bots that only resteem articles that have for example as their first tag "politics". In this category the bot would resteem everything, except maybe one of the other tags is one of the notorious spam signals like "dmania" or "colorchallenge". Also important is the language of the articles that get resteemed. You could name that bot @resteem-en-politics, which resteems politics articles in English. Another one could be named @resteem-en-sports which is doing the same with sports. And one could be named @resteem-de-deutsch, @resteem-en-philosophy, @resteem-es-php etc.
All topics could get their own bot and these bots do nothing else but resteeming a single topic in a single language. What you as a user have to do is to follow that bot and you exclusively get content of your liking into your feed.
I think that would be the most simple and efficient solution. Maybe there could be combined bots where two specific tags are necessary to thin out the amount, if that is necessary (e.g. @resteem-en-sports+racing).(I don't know how many articles are produced daily per category, but that should be added as information so you know how many posts to expect.)
Plus of course a blacklist for spammers and scammers.
The programming for that I believe is very simple:
If you do it right, I guess per language the average user would follow about 20-30 of these bots, of which everyone resteems perhalps about 5 posts per day. This wouldn't be more than the daily output of a big online news outlet.
The biggest problem is how to acquire all these accounts, because you need different phone numbers and IP addresses since the developers try to limit the number of accounts to one per person. So effectively, you'd have to talk to the developers and get their approval - something which could prove to be impossible.
Other biggest problems of Steemit are as I think the reward pool rape by the big bots as I described it in a recent post and the gap between witnesses&developers to the real world and their sheer ignorance, greed and incompetence. This must change and I intend to create an account for which I will collect as many votes for witnesses as possible. With that I will try to pressure them into the direction of more quality, more professionalism, more transparency and more work. Too many witnesses are lazy and just ripping off the system. Either they improve or they won't get a vote. At least theoretically, that should work.^^
You are talking proxy voting? Hmmm, interesting ...
You can create as many accounts as you want if you fund them with SP. The phone number/IP thing is just for free accounts. You can also create accounts through an API or SteemConnect (tutorial for SteemConnect)
I currently own this account plus two others (@steemitbooks and @shadeslayer) and I've also created one for a friend.
thx for the info! The creation of theme-based resteem bots would require to creat dozens of them, which makes it very expensive. The way via different IPs/Tels is the affordable way to do so.
I didn't know about the Steemconnect way of creating an account, is it for free and can you create multiple accounts with it?
The Steemconnect way is just like the way normal accounts are created but more accessible since you don't need to run any code. The only cost is the Steem or SP delegation that the new account needs. You don't ever lose any Steem, it just gets transferred to your new account. You can create unlimited accounts this way (all of the accounts I mentioned were created that way)
Steemit requires phone verification because they have to create accounts the same way (they are "paying" the registration fee with their delegation) and they don't want to pay for your secondary accounts.
So.. I only have to provide enough Steem Power as delegation, but I don't have to pay money? If I understand the tutorial correctly, it costs quite a bit:
I don't find this account_creation_fee... can you direct me there, please?
They removed that stat from steemd 😛 Last I knew though, the fee was 0.1 Steem. So if you want to transfer Steem/Power Up your account multiply that by 30 (3 Steem) and put that amount (or more) in the Steem tab. There isn't really a fee, it's just an amount you have to transfer to that account.
If you want to delegate instead of Powering Up the account, use this website to calculate the number of vests you want to delegate. Then put that amount in the Vests tab. I'm not totally sure, but I think you also need to put the base fee (0.1) in the Steem tab when you delegate.
I'd suggest you spend at least 6 Steem to set the account up.
Make sure to copy the password before you click any buttons 😉
Thanks a lot for riding me through this. This platform is so ridiculously complex.. just as if that all was necessary^^
Ok, I'm now standing in front of the next hurdle. I managed to get to the "create account". Since all my Steem is powered up, I intend to delegate Vests. In total 200,000 which is equal to ~100 SP (should be enough). So...
And I guess after that should be coming something like "your account has been created". But I get this error message:
Question is, what did I do wrong? Do the Vests that I delegate have to be powered down? I guess not. Or does it require more than zero Steem? You probably don't know that, but I ask anyway, you never know^^
Oh, and I tried to login to Steemconnect to see what my "console" there has to offer me, but somehow they've removed the login. lol.. Steemit is worse than Linux before Ubuntu.
Haven't delegated to create an account before but I believe you do have to put the registration fee (0.1) in the SP field. If that doesn't work you could try upping it to 0.5 since that is what Steemit gives to accounts.
Resteem bots are an interesting idea, but the issue with those is that they can overrun your feed. But they can clean out some of the junk and pollution through cross-tagging. But you could incorporate those with additional filters to create some interesting feeds. Or you could find a way to extract the posts from those bots and build an application on top of those bots.
I foresee that some curation organizations will develop in the future such that they simply resteem content, but spend copious amounts of time looking for specific posts, hire qualified writers, etc.
Yes, the problem of getting overran is always there. I would try to get that under control by looking at the average number of posts and if it goes up it implies a manual rewiew. Another rule could be to only allow users with a certain reputation and only articles with a minimum amount of words. Or as you say it: Add other filters.
If such a resteem bot grows popular, it could even charge money if someone wants his content to be added, but without guarantee of a resteem.
edit
Something unrelated: Can you give me a subdomain of your herokuapp.com URL (plus some space and a PHP access if possible)? I would give you 5 SBD for it and I want to start building a site for my little witness quality control project.
edit2
I just thought that thanks to the public blockchain it would also be possible to automatically resteem content that has been upvoted or resteemed by certain talented users who are good curators. You basically take a free-ride on their activity.
So Heroku manages and owns subdomains of herokuapp.com. They allow you to deploy small applications on their servers for free and then charge you if you want to scale up. Unfortunately, I don't own the domain or server space.
Instead of resteeming posts, the bots would do better to compile a post with links to the articles and the first paragraph. This would help with not generating tons of posts in the followers' feeds.
Really though, it is better for separate curators to handpick articles then to leave the work up to bots. I'm currently working on getting a Discord server set up for my @steemitbooks account where people can drop links to book related posts and I (or other people I assign as curators) can look through and select the best ones to feature in a curation post. While it requires more work it will lead to better quality feed then if you let a bot do it.
Yes, currently, the manual selection is probably the better way of spreading content.
I definitely agree that Trending and Hot are useless. About the only category they are useful for browsing is #contest and their usefulness is limited even there. New is slightly more useful but, like you said, full of junk.
Something I've though about working on, after I complete my current downvote bot project, is to create an mute list with SteemConnect to let you mute all the spammers on a blacklist with minimal effort. Don't know how many people would use it, but it might be useful.
Right now though, I mainly upvote content in my feed since I try to keep it to stuff I'm interested in or original authors that need support. Most users in it come from my infrequent browsing of the feed or Discord users I meet.
I don't think the current site is where Steemit needs to be heading. I'd like to see more projects built on top of it like Utopian, dTube, and dLive. I think that is where Steem's potential lies. I'd personally love a chat room where you could upvote messages
That said, Steemit is around to stay and I think that as time goes on and more quality users join the platform, we will be able to reduce spam. Most spam accounts can be flagged into oblivion with concentrated effort. The whales that abuse the system are much harder to stop. Our only hope there is to get some good whales that contribute to the platform instead of exploiting it.
Just my rambling thoughts 😉
I keep coming back to your wall for I know I can expect some valuable log and not a hoax. Having said that I do feel Whales have a responsibility to filter what's promoted in order to maintain quality here also trending and new doesn't really justify the attention.
I personally look for my favorite tag (as per mood) and try replying to new accounts with good content. That's my way of paying back to steemit.
I have talked a lot about the problems of Steemit and how to fix them. Everything about this platform revolves around investors and it is so absolutely unnecessary and unproductive at this point.
My post about improved filters directly addresses this issue. Tracking average payouts could help us.
If a whale makes an average of $500 per post and they have a post that's paying out $250, is this 'hot' or 'trending'? No, it's probably garbage, but that didn't stop it from getting massively more exposure than it should have.
Posts in 'hot' and 'trending' should get there from a combination of getting paid out more than average, making sure the author isn't spamming posts, upvotes by confirmed unique accounts that the community trusts, and straight up luck (lottery).
Unfortunately this site's only metric for trust are money and a laughable flagging system where people like @berniesanders have -17 reputation. These measures of trust are hilariously embarrassing and we the community should be doing everything in our power to implement more intelligent dynamic systems for our online mini-government.
Unfortunately our online mini-government works the same as normal government in that there are insider groups with special interests and nothing gets done.
Still, we are so close. The second a group of developers takes all this amazing open source technology and makes something truly amazing the entire world is going to benefit.
Did you ever try a version of your filter that actually makes all votes worth the same? I always wondered what that would like like. Of course if that was the criteria the nature of the bots might change. Might be even better though. Or even things like comments (spam filtered).
Are the hot and trending sections in the core blockchain code? Hum. I never bother to check that. Curious.... I'll see if I can verify that right now or find the right reference to it...
The current version of Q-Filter ranks users votes by how much they use Q-Filter and doesn't take SP or Rep into account. So it's basically making all votes the same.
The votes are different in that there are different weights, but everyone's starting point is the same and since work is required to evaluate posts to discover the score that essentially prevents whales from ever existing. So more egalitarian, but not perfectly equal.
They are in the API and are used by developers. I'm not sure whether or not they are in the core blockchain code, but even if they are not, they are still used by all of the applications that utilize the blockchain.
As for the filters I have been playing with, having each vote have a static voting weight never really occurred to me. There are so many ways to attack such a system that I never considered it. But I would imagine that in the type of filter I am using it would simply sort posts pseudo-randomly since most people vote at 100% with a few lesser percent whale votes in between.
Yeah I agree it's easy to game, however it isn't much different than Reddit's system, right? Maybe something else with comments (spam filtered) as a signal on top. I assume that it would be a much better indicator than what we have now and we might have something interesting to see from it.
It would be the same as Reddit's system, with one major exception. The money incentive. That changes people's behavior and gives us different results. Rather than vote for stuff people are in favor in, they vote for what is profitable. Sometimes they overlap, sometimes they don't.
Secondly, I originally designed the algorithms I built to build better Reddit feeds since Reddit feeds are often polluted with ideological shills. Though since Reddit is better at organizing stuff than Steemit, bringing Reddit-like filters to Steemit wouldn't be such a bad idea.
Ah yes, that is a good point about improving reddit's filters as well :).
That being said, I was simply wondering what the trending/hot would look like today with these filters (in the state where everyone is gaming the vote-weight version of the filters).
If it ever gets adjusted the other way, I'm sure you'll see a shift in the feed to game the new system.
Steemit’s biggest issue is that Steem Inc. isn’t interested in developing Steemit as a high quality platform. It’s both by design and by lack of interest (read: they are too vested to be the dominant long term player/interface of Steem).
Steemit is the initial centralized roll-out platform to bring eyeballs to the blockchain. Nothing more, nothing less.
Beyond that, one could even wonder whether Steem Inc. have what it takes to take the Steem ecosystem, with all its issues, to the next level.
So far we know that they don’t have what it takes to launch the long promised, and emphasized at SteemFest to be ready last year, communities. Same applies to the iOS app.
Somewhere, the issue is that there is a conflict between conflict of interest and desire to go full-out on Steemit. Given how vested Steem Inc. and top level employees are, that’s an entirely understandable ambiguity to struggle with.
I would be fine with Steemit continuing to suck, but all interfaces suck and all interfaces use the same trending, hot feeds that the blockchain provides via their API. You would imagine that one of these applications could fix the issue, but why bother when you have huge delegations to abuse at your disposal. The applications are cool, but the way they present information still is beyond primitive. Honestly, D-Tube is the only application I actually see accomplishing real development and even remotely addressing this issue.
The question is whether the third party developers with pick up the slack and whether they will be rewarded with cold hard Steem rather than a delegation. It's ridiculous that Steemit has millions of dollars of stake at their disposal and can not do anything but HODL it and hope the community saves them from their apathetic incompetence.
Agreed,
There is a steem blockchain with developers and there is a steemit the website with no developers.
How far is this from the truth? It would explain a lot of questions...
Great info once again from the visionary @greer184! Thanks for the effort and excellence on a whole other level. I am putting q-filter hashtag on my posts until I am asked to discontinue doing so. I love the discord and commentary.