Topic: Someone please make a steem clone - page 9. (Read 14356 times)

You call it parasitic, I call it rent. You pay orders of magnitude more rent for being on the busiest street downtown as compared to out on a country road in the middle of nowhere. But in most cases it is worth it because that's where the customers are. I don't think the value of being on a widely popular platform can be denied, but ultimately it comes down to price. Probably some specialists will not like the platform and its rent and will go elsewhere or self-host, just as happens with bloggers now. The numbers on how this will work out are entirely unclear.

As I said on a Steem comment reply, I don't think n^2 is written in stone either. It could be changed if the price is unworkable. It won't be changed to linear because that would indeed break the system, but there are a lot of other options.

hone away...

Afaics, the diversity of the content is orthogonal to the cognition that to earn significantly, you must cater to the demographics that are already on Steem. Meaning it is possible to find diverse topics that are judged to have a high value by Steem's current demographics, but this is normally because it is judged that the content is new and thus widens the appeal to potential readership. But a serious blogger isn't interested in always being new, because they can't be. They are interesting in being rewarded for being expert in a coterie. But coteries are a negative economically on Steem because of quadratic weighting (except for now the cyptonerd and future of steem intersection coterie which is why two of my blogs paid well, and of course the boob coterie of young enough cryptonerd males to still get an erection). You actively want to avoid forming a coterie.

Realize that the reward pool is partitioned by quadratic weighing, so the larger your vote count, then quadratically more of the pie you get.

Serious bloggers are not stupid. They will eventually figure out the economics of Steem is parasitic on coteries. Articles will be written about this. It can't be hidden.

Expect me to home in on the most salient issues. That is one of my vocations.

My guess is the chants of "too many boobs," is probably wrongheaded in this regard--it might rather be, "too many BTC and ETH posts..." Though the trending page is looking cosmopolis--at least more so than a few weeks ago.

Did you not read this?

Esoteric might help for those are very astute at who they follow, but I thought we were talking about crossing the chasm to mass adoption. Even I have followed people at Steem because I like a few of their posts, but that doesn't mean I will be interested in everything they might share.

But that isn't even the main dilemma (all social networks have this problem and any innovation will be a first mover in that way), which is really the economic problem of who is incentivized to invest their time and effort in Steem.

A share button (and following people with esoteric tastes)--fixes this. My guess is an algorithm can't fix closed mindedness--if anything, it will likely exacerbate it.

I posit the above violates the Pareto principle (e.g. 80/20 rule) of efficiency. Focus beats a firehouse. Serious bloggers will focus on how they can earn the most.

Also I have a specific economic incentive innovation over Steem (that afaics Steem can't copy) which will make it very clear to the serious bloggers that they should choose my competitor to Steem instead of Steem. I believe they will clearly see that the quadratic weight of rewards pulls their investment away from their following towards the groupthink. In other words, I think serious bloggers will see Steem as parasitic on them and their following.

Orders-of-magnitude and squaring are very powerful concepts. Principle of least power may apply in spades. One can do more harm than they foresee by applying too much algorithmic power. The elephant is unaware of all the things he crushes under his feet.

I could be wrong of course.

Yes I agree for the most part. Enhancement is needed there. A market solution may be people who offer feeds that feature new and undiscovered content. There is already one blogger doing this that I know about. I bet there are more and I can't find them!


If it gets stuck in a crypto demographic then I agree it will fail.

However, I'm not sure that the push toward what you call groupthink (and I call raw popularity) will reinforce this or overcome it. If people join who are not cryptonerds, then the content that appeals both to those inside and outside the crypto ghetto will do the best. Content that appeals to smaller groups will in some sense subsidize the content that appeals to the broadest audience. Perhaps it is a reasonable cost to pay such rent to gain the network benefits of sharing a widely-popular platform as a small niche that most people don't care about. Or perhaps you are right and niche interest groups won't stay at all.

I didn't qualify with "individual" and that was intentional. What I mean (which you understood) is that the ranking by value judgement is a also a relevance offered in the "trending" and "hot" relevance sortings (and those sortings within tags).

Since the other relevance features (active, new, tags but noting tags are sorted by new, active, hot, or trending) are not very personally relevant either, then I claim the only relevance currently offered is really mostly the value rankings.

Follows won't improve that much for discovering new content outside of the people you already follow.


Even if so, I posited that the quadratic weighting will still influence blog posts towards a groupthink. I posit the quadratic weight will fight against any attempt to develop relevance coteries, because it gives the economic benefits to those who can cater to the groupthink. This I believe is a catastrophic design decision and the Achilles Heel of Steem.

One can argue that if the demographics are diverse, then the quadratics will be split across many diverse groups, but not only is it a chicken-and-egg dilemma, but also power-law distributions don't work that way. They are always trending to more centralization and the free market creates mechanisms to route around the failure of overconcentration (and heck Steem started with a sneaky mine at 90% concentrated into 0.1% which is orders-of-magnitude worse than a stable power distribution).

OTOH, Steem has the advantage of being an open database that any one can leverage. Open systems are more resilient. But Steem is not entirely open source (can't be forked and there is less economic incentive thus far for creating a Steemit clone)


I thought of that too (and even mentioned the initial raw popularity lead is conceded), but remember the entire point was to move crypto out of our tiny nerd demographic and if the groupthink is economically enforcing the walled garden, then I think it is more negative than positive.

Edit: it is true that Facebook started with a smaller demographic to build a base and leveraged that to cross the chasm. But the problem I see is the quadratic weighting may be a wall that prevents that.

It doesn't actually do anything for (individual) relevance since the recommended feature was removed so I think that is incorrect.

Some alternate interface and new features may make sense to try to address that need. I think we will learn more as already-planned and developed features are rolled out.


To some extent a groupthink notion of value does add a lot of value (possibly the most value) to a social media platforms, as personally unappealing as that may be to you (or even me). There is reason you and I probably don't spend a lot of time on social media platforms. The nature of the beast is to try to bring people together on common ground so they can be starstruck at who has the most followers, "Like" each other's updates all day long, and occasionally engage in drama (which itself requires commonality otherwise you don't care). Teenage girls love it. Teenage girls are not typically iconoclasts.

There may be room for multiple dimensions of value, but raw popularity can't be dismissed as a huge value driver.

Agreed, see also I was thinking similarly per my edit of my prior post while you were composing your reply.

Steem's current relevance filters are:

* tags
* active, trending, hot, new
* and soon follows


The value question is very important to study.

Does your (or a groupthink's) judgement of value add the most value to Steem? I argue no. Because I posit a groupthink can perpetuate the singular demographics that can prevent Steem from crossing the chasm to mainstream diversity of demographics.

Rather I think the greatest value comes from the most users engaging in investing their time and effort in the site. For this you may need a proliferation of smaller coteries which are implausibly numerous for the groupthink to recognize. It is surely beyond Dunbar number upper bound of the cognitive load a human can handle in social networks.

So that is why it may not be wise to consider value and relevance orthogonally.


It depends on the what an upvote is supposed to mean and is actually doing. As it stands now, it is both a relevance and value judgement feature. But it has problems.

One example could argue that people want to view the upvote as an expression of something, not as an insignificant (insulting level) of monetary tip. The quadratic weighting can make each vote more monetarily significant when combined into a groupthink. But the groupthink has negative downsides/tradeoffs.

Voting in Steem does not express personal relevance. Prior to the recommended feature being removed, it could be interpreted as having in part that effect, but it no longer does, and that is certainly not its essential function.

Voting in Steem is more like tipping or crowdfunding, with a social component that increases rewards if more people approve.

For example, I'm not a fan of the Dollar Vigilante and I don't really care whether I see more of his stuff or not, I don't even know whether I agree with him on most things or not, but I voted for his posts because I think funding him adds value.

Maybe the upvote button should be replaced with a dollar sign/coin icon.

The question of personal relevance and own eyes is interesting, because I might very well want to see content that I think is crap so I can downvote it. In fact I'm more interested in seeing that than content I'm neutral about.

I wasn't proposing a poll for its feature value (although that is another perhaps useful feature), but rather as a release valve for those who will otherwise "abuse" the upvoting to reflect their agreement or disagreement, rather than using upvoting exclusively for expressing relevance of what they want to be see highly ranked for their own eyes ongoing. Disagreement isn't always "I don't want to see this genre of content in the future". See the problem is the two meanings of voting are currently conflated. I am still contemplating how to best handle relevance and voting. Not confident that a poll as a release valve is the optimum solution.

Edit: I thought about using time spent on a page as a metric of relevance and leave the upvoting for agreement+tipping (eliminating the downvoting), but time spent on a page is variable due to varying length and complexity of the content.

I don't think every post needs that, but a poll feature could be useful in making interesting posts.

I was thinking a volume button where you could adjust your tip (because what guy hasn't ever felt the urge to make it rain at the "scrip club"?). Though I think this goes hand in hand with a funded wallet.

Pacman Jones + scrip club = a 2-3 year bit on Jim Rome--can't imagine the curation miles that guy would rack up.

Reading this I thought there may be a word that could represent this functionality better than "upvote"

What do you think about using the word "magnetize"?

Pollsters would agree and that's why we get groupthink politics--not that there's anything wrong with it--just that enough metrics gets you one size fits all products.

Newbium is really great!

People want to express their reaction to content. If we conflate their opinion of agreement with the relevance of the content for their attention, then we don't have an attention relevance metric rather some Frankenstein. I may disagree with a blog post and be compelled to express it with a quick click, but it may not mean I want to hide that content from my future attention. Without an accurate relevance algorithm, we will encourage groupthink, because attention (i.e. ranking) drives votes and rewards.

Wait, so you're trying to discourage groupthink with a level of agreement/disagreement button--some things we're better off not knowing why--the imagination sometimes gets jolted by its unknowingness.