Me too. I am also sort of in all groups, especially in the beginning because I support this site and just thought contributing where it seem to need it as well as per my own interests made sense....
Me too.
I am also sort of in all groups, especially in the beginning because I support this site and just thought contributing where it seem to need it as well as per my own interests made sense.
When this site becomes more popular, I will like go back to lurking more, as I really am naturally an introvert even online.
I lurk because I simply don't know what to say. Plus anxiety causes me to go, "oh, I'd be hated for stating what I think, and people will probably think what I think too obvious".
I lurk because I simply don't know what to say. Plus anxiety causes me to go, "oh, I'd be hated for stating what I think, and people will probably think what I think too obvious".
I find that nine times out of ten someone has already said what I think, so there's no need for me to repeat the point and create more noise in a thread.
I find that nine times out of ten someone has already said what I think, so there's no need for me to repeat the point and create more noise in a thread.
I do definitely have those thought too, and I do find no matter the topic, there's an emotional cost to commenting. I was lucky enough to be on here near the beginning, when the community was...
I do definitely have those thought too, and I do find no matter the topic, there's an emotional cost to commenting. I was lucky enough to be on here near the beginning, when the community was generally more welcoming which has helped me.
Still generally Tildes has been really good, so I do find it easier to post here, and hopefully that's maintained.
I wish I could say that’s why I like using this site. I want to see it become the new Reddit, and I also want to be one of those users with a decade-old account that people look at and are like...
I wish I could say that’s why I like using this site. I want to see it become the new Reddit, and I also want to be one of those users with a decade-old account that people look at and are like “wow, he’s been here a while”
Ninja edit: The smaller community relative to Reddit also makes me like this one more because it feels like I can contribute more to conversations because of the lack of over-saturation of users.
In addition to this, I'd recommend you read this page from the docs on future mechanics, especially the parts involving the reputation system. https://docs.tildes.net/mechanics-future Quoting the...
In addition to this, I'd recommend you read this page from the docs on future mechanics, especially the parts involving the reputation system.
To be clear, I recognize that this is a dangerous type of system to implement, with the distinct risk of creating "power users" that have far too much influence. However, all systems have similar risks—even if all users are equal, people can form groups or abuse multiple accounts to increase their influence. These types of issues are social and can only be solved with oversight, accountability, and a willingness to punish people that abuse the system, not technology alone.
You can't exactly force people to post or comment, especially when they don't really have the time to do so. I don't post much recently because a lot of my time was taken up working, I still voted...
You can't exactly force people to post or comment, especially when they don't really have the time to do so. I don't post much recently because a lot of my time was taken up working, I still voted on submissions I found interesting though.
You can't force, but you might encourage. I'm thinking of online equivalents of a teacher calling on students in class, or soliciting manuscripts or submissions.
You can't force, but you might encourage.
I'm thinking of online equivalents of a teacher calling on students in class, or soliciting manuscripts or submissions.
The contexts in the examples you are thinking of are radically different, Tildes is (hopefully) not a full-time occupation for most people outside of it's administrators. Even if the will and the...
The contexts in the examples you are thinking of are radically different, Tildes is (hopefully) not a full-time occupation for most people outside of it's administrators. Even if the will and the incentives are there the hurdle you need to overcome isn't just that people are gun-shy with their opinions, but rather that they probably don't have the time to express them.
Given the 90-9-1 rule, a request system which explicitly solicited content from active lurkers (visiting site, interacting with content, but not posting or commenting) might be one approach....
Given the 90-9-1 rule, a request system which explicitly solicited content from active lurkers (visiting site, interacting with content, but not posting or commenting) might be one approach.
There's a lot of daylight between "hey, we'd really like to get some diversity of opinion and think you might have something valuable to say" every so often, and "yo, crank out 40 articles over the next five days".
Doing this successfully -- generating genuinely interesting and otherwise overlooked contributions, without crossing the lines into creepy or annoying -- is challenging. It may well require a human touch.
You can lower the amount of trust to very little by giving only very limited power at a given time. That way you can recruit a large part of the user base. There's different ways to do this, but...
You can lower the amount of trust to very little by giving only very limited power at a given time. That way you can recruit a large part of the user base.
There's different ways to do this, but for example for comments moderation, reported comments would get send to a jury of random users that would have the very limited power to judge whether that one particular comment is breaking the rule is was reported for or not. No need for a lot of trust for that.
I believe that both games have that function. CSGO you can have a 'super spectator' mode and be able to see the keypresses. LoL may have that function as well or both may not. I havent explored...
I believe that both games have that function. CSGO you can have a 'super spectator' mode and be able to see the keypresses. LoL may have that function as well or both may not. I havent explored them, but I do think that would be a good feature. Especially for games that want to hit the pro stage.
You could work from competence rather than raw activity, being on the right side of controversial decisions before they are settled should in my mind be worth a lot more than showing up every day...
You could work from competence rather than raw activity, being on the right side of controversial decisions before they are settled should in my mind be worth a lot more than showing up every day and upvoting the top items.
Upvoting something obviously amazing fast (because you're just sitting on the site all day) isn't really revealing much information to the system. It would've been upvoted anyway, you just happened to be there. In contrast upvoting something controversial which after exposure to lots of high-rep users is decided to be excellent demonstrates genuine competence.
Kat, sorry if you feel attacked, I did not mean to imply anything about your personal decision making (I haven't followed the older threads that closely to know either way!). Probably my "yous"...
Kat, sorry if you feel attacked, I did not mean to imply anything about your personal decision making (I haven't followed the older threads that closely to know either way!). Probably my "yous" should've been "ones" but then I'd sound even more of a pretentious ass.
That is incredibly subjective, and the exact opposite of something an algorithm should be trying to determine.
The thing which is subjective is the quality of the content being judged. But the users have to make a decision on this and the system has to decide which users have power to do this. Of course there can be big mod teams and many algorithmic mods needed to enact a ban, no doubt there will be direct admin interventions to adjust it, of course it can be rate limited and have all sorts of social aspects. But at some point for tildes to scale at all there must be a system which judges people based on past behavior and apportions power.
Ever done one of those colour vision eye tests? My argument is just that good judgment is measured best when judging difficult things.
There are already a lot of good thoughts in here, so I'm just going to quickly leave a few more, since I was hoping to both get some development done today as well as not... you know, work all...
There are already a lot of good thoughts in here, so I'm just going to quickly leave a few more, since I was hoping to both get some development done today as well as not... you know, work all day.
This is absolutely a problem and will happen, no matter what we do. Some people just get more invested than others and will find ways to have more influence, through either technical or social methods. This isn't even inherently a bad thing, but you have to be able to recognize and prevent it when some people start using it in ways that hurt the community instead of help it. This isn't something you can solve through technical measures alone.
A major factor is making sure that people with more "power" are accountable to others (though not necessarily to everyone), and also making sure that they can't act as the gatekeepers to prevent other people from gaining more influence as well. A lot of issues are caused by the responsibility being put on far too few people while they also have motivations to keep that higher-influence group small. Then you end up with those people burning out and feeling like they're in a constant battle with the community, while also pushing back against allowing more people to help.
It's mentioned in the Trust System page, but I think that making people's influence "decay" if they stop participating in a community regularly is really important. Communities evolve constantly, and just because someone understood the culture at one point doesn't mean that they still understand it if they stop paying attention for weeks or months.
I've been on Tildes from pretty early on, but just mostly as a lurker lately. I think I see what you are saying, but I don't necessarily see it as a problem. From what I've seen, the really active...
I've been on Tildes from pretty early on, but just mostly as a lurker lately. I think I see what you are saying, but I don't necessarily see it as a problem. From what I've seen, the really active users here have been a largely positive influence on the site. They are the ones who post topics, comment on threads thoughtfully, point new users in the right direction, and (usually) provide a model on what respectful behaviour looks like. Since the site is still pretty new, we need a group of people like this to keep it going. I don't really see it as them brute forcing their way into having influence over anything, just a group of people who are invested in the site and want to see it do well.
The main issue I see is how to get more people invested so that we can get a greater variety of perspectives and topics.
I'm always kind of happy when I recognize a username on here, makes it feel like a community. And well if in the new moderation system someone chooses to spend so much time on the site that they...
I'm always kind of happy when I recognize a username on here, makes it feel like a community.
And well if in the new moderation system someone chooses to spend so much time on the site that they have fingers in every pie - well they are kind of in the best position to make the disicions that accompany the power they have been given. They after all they already have a big stake in the success of the site.
You will see such effects in any informational or networked system.[1] Inevitably, a Zipf or power distribution will result.[2] Highly persistent patterns such as the 90-9-1 rule[3] mean that the...
You will see such effects in any informational or networked system.[1] Inevitably, a Zipf or power distribution will result.[2] Highly persistent patterns such as the 90-9-1 rule[3] mean that the pattern is exceedingly unlikely to change.
The question becomes far less "can we keep such a distribution from forming?"[4] but "what values will it reflect and what relationships will it emphasize or inhibit?". I'd like to suggest that the discussion might more fruitfully focus here.
An area I've been exploring over the past few years has been media and communications theory, a discipline covering a wide range of sins.[5] One element that's surfaced is that informational systems develop to favour rewarded behaviour. That is, if what's rewarded is popularity, you get very popular (and not very sophisticated) content. Money and advertising produce similar results. Power structures and institutions impose censorship to preserve their own interests.[6] Conspiracy theories (the crackpot type) seem to have their own dynamics.
If your interest is truth or accuracy, then that must be what you reward. In a corollary to Celine's 2nd Law (from Robert Anton Wilson's "Illuminatus" trilogy), not only is accurate information possible only in a non-punishing situation, but it's possible only in a situation which rewards accurate information and nothing else. Advertising, PR, propaganda, publish-or-perish, weaponised viral clickbait ... are all "something else". And they're ultimately hostile to accurate and useful information.[7]
Social systems develop the behaviours they explicitly or implicitly reward, and if undesireable patterns emerge, it is essential to examine and change those rewards.
And of course, truth and accuracy might not be your only aims: building and sustaining community, entertainment, amusement, phatic interactions, imperatives and calls to action, and, for a commercially-oriented venue, sustaining ongoing operations and development, are all related concerns. I don't see these as wholly incompatible.
There are other concerns, such as providing a voice to the voiceless, punching up (vs. down or horizontally), speaking truth to power, providing exposure to unknown or distant viewpoints or cultures (in space, time, or social dimensions -- your neighbour, parents, or children might be as removed as a distant tribal village or global megopolis, or some long-ago or far future civilisation). The history of notions of free speech, and particulaarrly the free-market fundamentalist underpinnings of "the marketplace of ideas" is also worth exploring. It differs today sharply from its putative originsfrom John Stuart Mill, particularly in "On Liberty".[8]
That is, they are values.
So maybe for now we might want to consider what Tilde's values are, how well it is serving them, what it rewards or discourages, and how it might best improve.[9]
Notes:
On Tildes, users (nodes / vertices) interact through posts and comments (links / edges).
Across any human interaction medium with inherently equal access, roughly 90% of particiipants lurk, contriuting essentially no content, 9% contribute about half the content, and 1% the other half. See Jacob Nielsen, "The 90-9-1 Rule for Participation Inequality in Social Media and Online Communities". My own 30+ years in online communities bears this out.
It may be possible to tamp down the high end by rate-limiting posts and comments, or otherwise imposing costs for transmitting into the channel. This can help against botspew and some human-initiated disruptive discussion, but does not induce more participation, directly, from others. In a live venue, e.g., a classroom, this might be accomplished by calling on individuals, perhaps randomly. This is more difficult to accomplish online. Though perhaps not impossible.
I've compiled a facetiously titled light reading list, which I'm still working through myself.
The Catholic Inquisition becomes all the more interesting in this light. There's also a whole cast of explorations of demagogic, populist, and fascist totalitarian dynamics as information-theoretic which caan be found in Walter Lippmann, H.L. Mencken, Antonio Gramsci, Norbert Wiener, Dwight MacDonald, the Frankfurt School, Hanah Arendt, Noam Chomsky, Umberto Eco, and others. See also: grasshoppers and locusts.
And this comment itself is an example and argument or what I'd like to see ... at least some of the time. Considered, substantive, supported explorations of a topic or question, preferably generating more of same, or betteer, actual work toward the goals discussed. Something the Internet was explicitly designed for (TBL, CERN), but today serves poorly, for many of the reasons expressed above. And, yes, I've had my doubts, dissapointments, and concerns with Tilde. Though not specifically on account of emergent prominent voices so much as what their emergence reveals of the site.
Your bolded thesis statement (might want to fix the typo) is spot on. I feel like I mentioned this in another thread, though did not go through the work of compiling resources to support my...
Social systems develop the behaviours they explicitly or implicitly rewaard, and if undesireable patterns emerge, it is essential to examine and change those rewards.
Your bolded thesis statement (might want to fix the typo) is spot on. I feel like I mentioned this in another thread, though did not go through the work of compiling resources to support my assertion.
The trouble that Tildes is going to face, along with most other systems, is ensuring the rewards are encouraging the correct behaviors. I think it's fairly easy to establish goals and come up with some ideas that one would think would work. It'll be important to set up a system that measures the affects of the rewards that are established.
Very much. The principle ef emergent properties, with new and distinct behaviours appearing at various size thresholds, also means that what works initially will not work at scale. The converse is...
Very much. The principle ef emergent properties, with new and distinct behaviours appearing at various size thresholds, also means that what works initially will not work at scale. The converse is that if you begin by focussing on problems-at-scale, you'll likely bollux up the small stuff. I suspect that this was a large part of Imzy's failure: they were so focussed on Big Reddit problems that they failed to consider Small Reddit problems. Though there were other bad bits as well.
I don't know what the threshold points are, though a general empiricsl rule of physics is that scaling by an order of magnitude puts you in a new domain. In behavioural networks, that seems sometimes too big, sometimes too small.
Networks seem to exhibit both topology and scale effects. Topologies include null, unary, pair, chain, ring, star, web, tree, mesh, and compound.
Scale begins similarly: null, unary, pair. At 3 you've a choice of topologies (chain/mesh), at an n=5 mesh, the link count exceeds nodes. Most core government executive councils seem to settle on 5 ≤ n < 10 (US and UK key cabinet seats, Politburo Standing Committees, interesting history). A uni seminar is best below about 15, classsroom size at 20-25. As interactions grow above this stage, central inteeractivity falls -- the centre is in broadcast mode, and subgroupsbstart forming.
All of this is tiny by social network staandards, but it is where actual interactions occur. The system should foster this scale.
Group dynamics seem to shift fairly markedly at, roughly, 50, 150, 300, 1,000, and larger scales. Above 1,000, the order-of-magnitude rule-of-thumb seems to hold well. The lifespan of the group (ad hoc, days, weeks, months, years, ...) also matters.
Between 10k and 100k, I've seen palpable shifts in most online forums. Usenet as of 1990 was roughly 200,000 mostly University students and affilliates (John Quarterman, The Matrix). I'd watched Slashdot, Kuro5hin, Digg, Reddit, and G+ through similar type growth phases.
(Oh yeah: decline-phase modes are a whole 'nother story.)
Founding cohort matters tremendously, as does institutional affilliation. Usenet, and early Facebook, were both based on uni populations, with some level of accountability to either their institution's administrators or peers. Suasion worked. In general populations, you lose that lever, and expansion of both networks beyond this base rapidly evolved behaviour, including abuse. A founding culture can still leave lasting marks. (This was IMO another of Imzy's failings, see also: Voat, Gab.)
Early-phase growth hacking by sniping high-value contributors is a good way to precipitate both growth of new and decline of old networks.
At national/international scales, bad actors, or simply asocial/oblivious individuals or groups can enter. There is also the problem of determining common standards for fundamental concepts. The question of labels to apply to places, groups, dates, cultures, etc., is a frequently recurring example (there are ... numerous others, though this is one I've been raising on Mastodon lately.). Complexity complexifies.
Thanks for the typo catch. I'm on a tablet w/ soft keyboard. It's a damned pain.
A lot of what I think has already been covered in this thread. Just my two cents, I can see what you're getting at, but don't really agree. Power users are naturally going to come up in any...
A lot of what I think has already been covered in this thread.
Just my two cents, I can see what you're getting at, but don't really agree. Power users are naturally going to come up in any system, and as long as they're also held accountable the same as any user that's fine.
And just what I've noticed so far, when a power user makes a post or comment, I actually feel they are held to a slightly higher standard. I feel they get called faster on low-effort or mean comments. Not to say that will always happen.
We've talked a lot about moderation, editing, curation, posting mechanics, all sorts of topics. The one topic we haven't really talked about is the community leader roles. Who speaks for the...
We've talked a lot about moderation, editing, curation, posting mechanics, all sorts of topics. The one topic we haven't really talked about is the community leader roles. Who speaks for the groups? Why do they speak for the group? Do we want all of our trusted users with powers invisible behind an alias so they can't be witch-hunted? Or out in the open so they have more accountability and a visible leadership role in a community? What tools do we need to invent so that leaders can do a better job?
This isn't overly important in small groups, but on a website with thousands of groups all interacting with each other it becomes a critical issue. These leaders will become like representatives for their groups. That's a governance model we'll need to discuss and build to have any hope of operating well at a large scale.
And will those community leaders have a limitation on length of service? People burn out and cling on to something that they perceive as giving them status and authority, to the detriment of both...
And will those community leaders have a limitation on length of service? People burn out and cling on to something that they perceive as giving them status and authority, to the detriment of both themselves and a community.
I like the concept of continual turnover of mods/leaders, drawn from active members. It needs to be a slow considered process of renewal though. I'm kind of against, for want of a better description, a "tenured system".
People find it hard to adapt to changes after a while which hurts them and their communities.
I don't like referencing reddit too much, but the "legacy mods" and many many inactive mods in huge subreddits are incredibly detrimental to the site as a whole, and a big reason why I think a slow and steady, but constant induction of new and enthusiastic people is pretty important.
That's access to moderation, not leadership. They aren't the same thing. Being a mod is just which controls you get to click in the interface. The reply box is something much more powerful, and...
That's access to moderation, not leadership. They aren't the same thing.
Being a mod is just which controls you get to click in the interface. The reply box is something much more powerful, and everyone gets it on day one. Leaders make replies, and everyone's opinion changes to fall in line with them because they are persuasive. They do this often enough, and they have a little cult following, another group who hates their guts, and everyone else's attention.
Leaders/power users are going to appear because leadership has nothing to do with what bits/dials you have on the website. It's about behavior and how well you can get your points across. Do that on behalf of a community and they will get behind you. If no one is doing it, someone will get nominated to do it by the other users/mods who don't want to do it themselves.
We've talked about solving the moderator problem with the trust system. We haven't talked about a framework for these leader-types. People who become mods don't somehow automatically make for good leadership, and leadership skills doesn't mean you make a good moderator. I think if we equate leaders to moderators (like reddit does) we're going to end up in trouble down the road.
Let me also just throw my lot in the pile that thinks it's a non-issue. Spending a lot of time on the site or having broad interests is non a punishable offense, a prolific user gaining mod...
Let me also just throw my lot in the pile that thinks it's a non-issue. Spending a lot of time on the site or having broad interests is non a punishable offense, a prolific user gaining mod privileges through activity did not game the system and is not less deserving of those privileges, anything like a catch up mechanic is silly and only useful in games like Mario cart, and if you are intimidated or dissuaded from posting because a popular user has posted, then that's not a fault of the system it the user, it's on you.
If somebody was prolific because of bitting or account sharing that's different, but I can't even fathom how a normal, rule abiding user that spends a lot of time on the site could ever be a bad thing.
But if they abuse their abilities then those abilities can be taken away. What makes somebody who got there in 3 months because they were prolific vs someone who got there in a year because they...
But if they abuse their abilities then those abilities can be taken away. What makes somebody who got there in 3 months because they were prolific vs someone who got there in a year because they only browse 15 minutes a day?
But they also can't get powers without messaging deimos either, so it's moot. By the time they can get power through automated systems we'll also have the systems to remove them from power
But they also can't get powers without messaging deimos either, so it's moot. By the time they can get power through automated systems we'll also have the systems to remove them from power
Without means for mitigating crapfloods, crapfloods will happen (as will other forms of abuse). Undermoderation is also a problem. And a sufficiently attractive platform will become aa magnet for...
Without means for mitigating crapfloods, crapfloods will happen (as will other forms of abuse). Undermoderation is also a problem. And a sufficiently attractive platform will become aa magnet for such behaviour. That's effectively Facebook's problem, wwith several compounding factors.
“The major problem—one of the major problems, for there are several—one of the many major problems with governing people is that of whom you get to do it; or rather of who manages to get people to...
“The major problem—one of the major problems, for there are several—one of the many major problems with governing people is that of whom you get to do it; or rather of who manages to get people to let them do it to them.
To summarize: it is a well-known fact that those people who must want to rule people are, ipso facto, those least suited to do it.
To summarize the summary: anyone who is capable of getting themselves made President Moderator should on no account be allowed to do the job.”
Me too.
I am also sort of in all groups, especially in the beginning because I support this site and just thought contributing where it seem to need it as well as per my own interests made sense.
When this site becomes more popular, I will like go back to lurking more, as I really am naturally an introvert even online.
I lurk because I simply don't know what to say. Plus anxiety causes me to go, "oh, I'd be hated for stating what I think, and people will probably think what I think too obvious".
I find that nine times out of ten someone has already said what I think, so there's no need for me to repeat the point and create more noise in a thread.
I do definitely have those thought too, and I do find no matter the topic, there's an emotional cost to commenting. I was lucky enough to be on here near the beginning, when the community was generally more welcoming which has helped me.
Still generally Tildes has been really good, so I do find it easier to post here, and hopefully that's maintained.
I wish I could say that’s why I like using this site. I want to see it become the new Reddit, and I also want to be one of those users with a decade-old account that people look at and are like “wow, he’s been here a while”
Ninja edit: The smaller community relative to Reddit also makes me like this one more because it feels like I can contribute more to conversations because of the lack of over-saturation of users.
In addition to this, I'd recommend you read this page from the docs on future mechanics, especially the parts involving the reputation system.
https://docs.tildes.net/mechanics-future
Quoting the final passage from Deimos,
You can't exactly force people to post or comment, especially when they don't really have the time to do so. I don't post much recently because a lot of my time was taken up working, I still voted on submissions I found interesting though.
You can't force, but you might encourage.
I'm thinking of online equivalents of a teacher calling on students in class, or soliciting manuscripts or submissions.
The contexts in the examples you are thinking of are radically different, Tildes is (hopefully) not a full-time occupation for most people outside of it's administrators. Even if the will and the incentives are there the hurdle you need to overcome isn't just that people are gun-shy with their opinions, but rather that they probably don't have the time to express them.
Given the 90-9-1 rule, a request system which explicitly solicited content from active lurkers (visiting site, interacting with content, but not posting or commenting) might be one approach.
There's a lot of daylight between "hey, we'd really like to get some diversity of opinion and think you might have something valuable to say" every so often, and "yo, crank out 40 articles over the next five days".
Doing this successfully -- generating genuinely interesting and otherwise overlooked contributions, without crossing the lines into creepy or annoying -- is challenging. It may well require a human touch.
You can lower the amount of trust to very little by giving only very limited power at a given time. That way you can recruit a large part of the user base.
There's different ways to do this, but for example for comments moderation, reported comments would get send to a jury of random users that would have the very limited power to judge whether that one particular comment is breaking the rule is was reported for or not. No need for a lot of trust for that.
I believe that both games have that function. CSGO you can have a 'super spectator' mode and be able to see the keypresses. LoL may have that function as well or both may not. I havent explored them, but I do think that would be a good feature. Especially for games that want to hit the pro stage.
You could work from competence rather than raw activity, being on the right side of controversial decisions before they are settled should in my mind be worth a lot more than showing up every day and upvoting the top items.
Upvoting something obviously amazing fast (because you're just sitting on the site all day) isn't really revealing much information to the system. It would've been upvoted anyway, you just happened to be there. In contrast upvoting something controversial which after exposure to lots of high-rep users is decided to be excellent demonstrates genuine competence.
Kat, sorry if you feel attacked, I did not mean to imply anything about your personal decision making (I haven't followed the older threads that closely to know either way!). Probably my "yous" should've been "ones" but then I'd sound even more of a pretentious ass.
The thing which is subjective is the quality of the content being judged. But the users have to make a decision on this and the system has to decide which users have power to do this. Of course there can be big mod teams and many algorithmic mods needed to enact a ban, no doubt there will be direct admin interventions to adjust it, of course it can be rate limited and have all sorts of social aspects. But at some point for tildes to scale at all there must be a system which judges people based on past behavior and apportions power.
Ever done one of those colour vision eye tests? My argument is just that good judgment is measured best when judging difficult things.
There are already a lot of good thoughts in here, so I'm just going to quickly leave a few more, since I was hoping to both get some development done today as well as not... you know, work all day.
This is absolutely a problem and will happen, no matter what we do. Some people just get more invested than others and will find ways to have more influence, through either technical or social methods. This isn't even inherently a bad thing, but you have to be able to recognize and prevent it when some people start using it in ways that hurt the community instead of help it. This isn't something you can solve through technical measures alone.
A major factor is making sure that people with more "power" are accountable to others (though not necessarily to everyone), and also making sure that they can't act as the gatekeepers to prevent other people from gaining more influence as well. A lot of issues are caused by the responsibility being put on far too few people while they also have motivations to keep that higher-influence group small. Then you end up with those people burning out and feeling like they're in a constant battle with the community, while also pushing back against allowing more people to help.
It's mentioned in the Trust System page, but I think that making people's influence "decay" if they stop participating in a community regularly is really important. Communities evolve constantly, and just because someone understood the culture at one point doesn't mean that they still understand it if they stop paying attention for weeks or months.
I've been on Tildes from pretty early on, but just mostly as a lurker lately. I think I see what you are saying, but I don't necessarily see it as a problem. From what I've seen, the really active users here have been a largely positive influence on the site. They are the ones who post topics, comment on threads thoughtfully, point new users in the right direction, and (usually) provide a model on what respectful behaviour looks like. Since the site is still pretty new, we need a group of people like this to keep it going. I don't really see it as them brute forcing their way into having influence over anything, just a group of people who are invested in the site and want to see it do well.
The main issue I see is how to get more people invested so that we can get a greater variety of perspectives and topics.
I'm always kind of happy when I recognize a username on here, makes it feel like a community.
And well if in the new moderation system someone chooses to spend so much time on the site that they have fingers in every pie - well they are kind of in the best position to make the disicions that accompany the power they have been given. They after all they already have a big stake in the success of the site.
You will see such effects in any informational or networked system.[1] Inevitably, a Zipf or power distribution will result.[2] Highly persistent patterns such as the 90-9-1 rule[3] mean that the pattern is exceedingly unlikely to change.
The question becomes far less "can we keep such a distribution from forming?"[4] but "what values will it reflect and what relationships will it emphasize or inhibit?". I'd like to suggest that the discussion might more fruitfully focus here.
An area I've been exploring over the past few years has been media and communications theory, a discipline covering a wide range of sins.[5] One element that's surfaced is that informational systems develop to favour rewarded behaviour. That is, if what's rewarded is popularity, you get very popular (and not very sophisticated) content. Money and advertising produce similar results. Power structures and institutions impose censorship to preserve their own interests.[6] Conspiracy theories (the crackpot type) seem to have their own dynamics.
If your interest is truth or accuracy, then that must be what you reward. In a corollary to Celine's 2nd Law (from Robert Anton Wilson's "Illuminatus" trilogy), not only is accurate information possible only in a non-punishing situation, but it's possible only in a situation which rewards accurate information and nothing else. Advertising, PR, propaganda, publish-or-perish, weaponised viral clickbait ... are all "something else". And they're ultimately hostile to accurate and useful information.[7]
Social systems develop the behaviours they explicitly or implicitly reward, and if undesireable patterns emerge, it is essential to examine and change those rewards.
And of course, truth and accuracy might not be your only aims: building and sustaining community, entertainment, amusement, phatic interactions, imperatives and calls to action, and, for a commercially-oriented venue, sustaining ongoing operations and development, are all related concerns. I don't see these as wholly incompatible.
There are other concerns, such as providing a voice to the voiceless, punching up (vs. down or horizontally), speaking truth to power, providing exposure to unknown or distant viewpoints or cultures (in space, time, or social dimensions -- your neighbour, parents, or children might be as removed as a distant tribal village or global megopolis, or some long-ago or far future civilisation). The history of notions of free speech, and particulaarrly the free-market fundamentalist underpinnings of "the marketplace of ideas" is also worth exploring. It differs today sharply from its putative originsfrom John Stuart Mill, particularly in "On Liberty".[8]
That is, they are values.
So maybe for now we might want to consider what Tilde's values are, how well it is serving them, what it rewards or discourages, and how it might best improve.[9]
Notes:
On Tildes, users (nodes / vertices) interact through posts and comments (links / edges).
VSauce has a surprisingly good explainer, The Zipf Mystery.
Across any human interaction medium with inherently equal access, roughly 90% of particiipants lurk, contriuting essentially no content, 9% contribute about half the content, and 1% the other half. See Jacob Nielsen, "The 90-9-1 Rule for Participation Inequality in Social Media and Online Communities". My own 30+ years in online communities bears this out.
It may be possible to tamp down the high end by rate-limiting posts and comments, or otherwise imposing costs for transmitting into the channel. This can help against botspew and some human-initiated disruptive discussion, but does not induce more participation, directly, from others. In a live venue, e.g., a classroom, this might be accomplished by calling on individuals, perhaps randomly. This is more difficult to accomplish online. Though perhaps not impossible.
I've compiled a facetiously titled light reading list, which I'm still working through myself.
The Catholic Inquisition becomes all the more interesting in this light. There's also a whole cast of explorations of demagogic, populist, and fascist totalitarian dynamics as information-theoretic which caan be found in Walter Lippmann, H.L. Mencken, Antonio Gramsci, Norbert Wiener, Dwight MacDonald, the Frankfurt School, Hanah Arendt, Noam Chomsky, Umberto Eco, and others. See also: grasshoppers and locusts.
Some further development, in the context of micropayments: "Repudiation as the micropayments killer feature (Not)".
See "Manufacturing consent" is the rebuttal to "The marketplace of ideas", featuring Jill Gordon's "John Stuart Mill and 'The Marketplace of Ideas'" (available via Sci-Hub). Hans Jensens's treatment of Mill is also of interest.
And this comment itself is an example and argument or what I'd like to see ... at least some of the time. Considered, substantive, supported explorations of a topic or question, preferably generating more of same, or betteer, actual work toward the goals discussed. Something the Internet was explicitly designed for (TBL, CERN), but today serves poorly, for many of the reasons expressed above. And, yes, I've had my doubts, dissapointments, and concerns with Tilde. Though not specifically on account of emergent prominent voices so much as what their emergence reveals of the site.
Your bolded thesis statement (might want to fix the typo) is spot on. I feel like I mentioned this in another thread, though did not go through the work of compiling resources to support my assertion.
The trouble that Tildes is going to face, along with most other systems, is ensuring the rewards are encouraging the correct behaviors. I think it's fairly easy to establish goals and come up with some ideas that one would think would work. It'll be important to set up a system that measures the affects of the rewards that are established.
Very much. The principle ef emergent properties, with new and distinct behaviours appearing at various size thresholds, also means that what works initially will not work at scale. The converse is that if you begin by focussing on problems-at-scale, you'll likely bollux up the small stuff. I suspect that this was a large part of Imzy's failure: they were so focussed on Big Reddit problems that they failed to consider Small Reddit problems. Though there were other bad bits as well.
I don't know what the threshold points are, though a general empiricsl rule of physics is that scaling by an order of magnitude puts you in a new domain. In behavioural networks, that seems sometimes too big, sometimes too small.
Networks seem to exhibit both topology and scale effects. Topologies include null, unary, pair, chain, ring, star, web, tree, mesh, and compound.
Scale begins similarly: null, unary, pair. At 3 you've a choice of topologies (chain/mesh), at an n=5 mesh, the link count exceeds nodes. Most core government executive councils seem to settle on 5 ≤ n < 10 (US and UK key cabinet seats, Politburo Standing Committees, interesting history). A uni seminar is best below about 15, classsroom size at 20-25. As interactions grow above this stage, central inteeractivity falls -- the centre is in broadcast mode, and subgroupsbstart forming.
All of this is tiny by social network staandards, but it is where actual interactions occur. The system should foster this scale.
Group dynamics seem to shift fairly markedly at, roughly, 50, 150, 300, 1,000, and larger scales. Above 1,000, the order-of-magnitude rule-of-thumb seems to hold well. The lifespan of the group (ad hoc, days, weeks, months, years, ...) also matters.
Between 10k and 100k, I've seen palpable shifts in most online forums. Usenet as of 1990 was roughly 200,000 mostly University students and affilliates (John Quarterman, The Matrix). I'd watched Slashdot, Kuro5hin, Digg, Reddit, and G+ through similar type growth phases.
(Oh yeah: decline-phase modes are a whole 'nother story.)
Founding cohort matters tremendously, as does institutional affilliation. Usenet, and early Facebook, were both based on uni populations, with some level of accountability to either their institution's administrators or peers. Suasion worked. In general populations, you lose that lever, and expansion of both networks beyond this base rapidly evolved behaviour, including abuse. A founding culture can still leave lasting marks. (This was IMO another of Imzy's failings, see also: Voat, Gab.)
Early-phase growth hacking by sniping high-value contributors is a good way to precipitate both growth of new and decline of old networks.
At national/international scales, bad actors, or simply asocial/oblivious individuals or groups can enter. There is also the problem of determining common standards for fundamental concepts. The question of labels to apply to places, groups, dates, cultures, etc., is a frequently recurring example (there are ... numerous others, though this is one I've been raising on Mastodon lately.). Complexity complexifies.
Thanks for the typo catch. I'm on a tablet w/ soft keyboard. It's a damned pain.
A lot of what I think has already been covered in this thread.
Just my two cents, I can see what you're getting at, but don't really agree. Power users are naturally going to come up in any system, and as long as they're also held accountable the same as any user that's fine.
And just what I've noticed so far, when a power user makes a post or comment, I actually feel they are held to a slightly higher standard. I feel they get called faster on low-effort or mean comments. Not to say that will always happen.
We've talked a lot about moderation, editing, curation, posting mechanics, all sorts of topics. The one topic we haven't really talked about is the community leader roles. Who speaks for the groups? Why do they speak for the group? Do we want all of our trusted users with powers invisible behind an alias so they can't be witch-hunted? Or out in the open so they have more accountability and a visible leadership role in a community? What tools do we need to invent so that leaders can do a better job?
This isn't overly important in small groups, but on a website with thousands of groups all interacting with each other it becomes a critical issue. These leaders will become like representatives for their groups. That's a governance model we'll need to discuss and build to have any hope of operating well at a large scale.
And will those community leaders have a limitation on length of service? People burn out and cling on to something that they perceive as giving them status and authority, to the detriment of both themselves and a community.
I like the concept of continual turnover of mods/leaders, drawn from active members. It needs to be a slow considered process of renewal though. I'm kind of against, for want of a better description, a "tenured system".
People find it hard to adapt to changes after a while which hurts them and their communities.
I don't like referencing reddit too much, but the "legacy mods" and many many inactive mods in huge subreddits are incredibly detrimental to the site as a whole, and a big reason why I think a slow and steady, but constant induction of new and enthusiastic people is pretty important.
That's access to moderation, not leadership. They aren't the same thing.
Being a mod is just which controls you get to click in the interface. The reply box is something much more powerful, and everyone gets it on day one. Leaders make replies, and everyone's opinion changes to fall in line with them because they are persuasive. They do this often enough, and they have a little cult following, another group who hates their guts, and everyone else's attention.
Leaders/power users are going to appear because leadership has nothing to do with what bits/dials you have on the website. It's about behavior and how well you can get your points across. Do that on behalf of a community and they will get behind you. If no one is doing it, someone will get nominated to do it by the other users/mods who don't want to do it themselves.
We've talked about solving the moderator problem with the trust system. We haven't talked about a framework for these leader-types. People who become mods don't somehow automatically make for good leadership, and leadership skills doesn't mean you make a good moderator. I think if we equate leaders to moderators (like reddit does) we're going to end up in trouble down the road.
Yeah, I was definitely conflating the two things.
Let me also just throw my lot in the pile that thinks it's a non-issue. Spending a lot of time on the site or having broad interests is non a punishable offense, a prolific user gaining mod privileges through activity did not game the system and is not less deserving of those privileges, anything like a catch up mechanic is silly and only useful in games like Mario cart, and if you are intimidated or dissuaded from posting because a popular user has posted, then that's not a fault of the system it the user, it's on you.
If somebody was prolific because of bitting or account sharing that's different, but I can't even fathom how a normal, rule abiding user that spends a lot of time on the site could ever be a bad thing.
But if they abuse their abilities then those abilities can be taken away. What makes somebody who got there in 3 months because they were prolific vs someone who got there in a year because they only browse 15 minutes a day?
You need a mechanism for doing this which scales. Presently that's not clearly provided, other than messaging or emailing @deimos.
But they also can't get powers without messaging deimos either, so it's moot. By the time they can get power through automated systems we'll also have the systems to remove them from power
Without means for mitigating crapfloods, crapfloods will happen (as will other forms of abuse). Undermoderation is also a problem. And a sufficiently attractive platform will become aa magnet for such behaviour. That's effectively Facebook's problem, wwith several compounding factors.
(Also Twitter and Reddit.)
“The major problem—one of the major problems, for there are several—one of the many major problems with governing people is that of whom you get to do it; or rather of who manages to get people to let them do it to them.
To summarize: it is a well-known fact that those people who must want to rule people are, ipso facto, those least suited to do it.
To summarize the summary: anyone who is capable of getting themselves made
PresidentModerator should on no account be allowed to do the job.”-Douglas Adams (modified).