3 ms·

> Ultimately, big problems with an open social web include: These two seem like the same problem: > moderation > spam You need some way of distinguishing hi

> Ultimately, big problems with an open social web include: These two seem like the same problem: > moderation > spam You need some way of distinguishing high quality from low quality posts. But we kind of already have that. Make likes public (what else are they even for?). Then show people posts from the people they follow or that the people they follow liked. Have a dislike button so that if you follow someone but always dislike the things they like, your client learns you don't want to see the things they like. Now you don't see trash unless you follow people who like trash, and then whose fault is that? > which now includes scrapers bringing your site to a crawl This is a completely independent problem from spam. It's also something decentralized networks are actually good at. If more devices are requesting some data then there are more sources of it. Let the bots get the data from each other. Track share ratios so high traffic nodes with bad ratios get banned for leeching and it's cheaper for them to get a cloud node somewhere with cheap bandwidth and actually upload than to buy residential proxies to fight bans. > good faith verification > posting transparency It's not clear what these are but they sound like kind of the same thing again and in particular they sound like elements in the authoritarian censorship toolbox which you don't actually need or want once you start showing people the posts they actually want to see instead of a bunch of spam from anons that nobody they follow likes.

johnnyanmac 1y ago
>You need some way of distinguishing high quality from low quality posts. Yes. But I see curation more as a 2nd order problems to solve once the bases are taken care of. Moderation focuses on addressing the low quality, while curation makes sure tye high quality posts receive focus. The tools needed for curation, stuff like filtering, finding similar posts/comments, popularity, following, are different from those needed to moderate, or self moderate (ignore, down voting, reporting). The latter poisons a site before it can really start to curate to its users. >This is a completely independent problem from spam. Yeah, thinking more about it, it probably is a distinct category. It simply has a similar result of making a site unable to function. >It's not clear what these are but they sound like kind of the same thing again I can clarify. In short, posting transparency focused more on the user and good faith verification focuses more on the content. (I'm also horrible with naming, so I welcome better terms to describe these) - Posting transparency at this point has one big goal: ensure you know when a human or a bot is posting. But it extends to ensuring there's no impersonation, that there's no abuse of alt accounts, and no voting manipulation. It can even extend in some domains to making sure e.g. That a person who says they worked at Google actually worked at Google. But this is definitely a step that can overstep privacies. - good faith verification refers more towards a duty to properly vet and fact check information that is posted. It may include addressing misinformation and hate, or removing non-transparent intimate advice like legal/medical claims without sources or proper licensing. It essentially boils down to making ensuring that "bad but popular" advice doesn't proliferate, as it it ought to do. >they sound like elements in the authoritarian censorship toolbox which you don't actually need or want once you start showing people the posts they actually want to see Yes, they are. I think we've seen enough examples of how dangerous "showing people what they actually want to see" can be if left unchecked. And the incentives to keep them up are equally dangerous in an ad-driven platform. Being able to address that naturally requires some more authorian approaches. That's why "good faith" is an important factor here. Any authoritarian act you introduce can only work on trust, and is easily broken by abuse. If we want incentives to change from "maximizing engagement" to "maximizing quality and community", we need to cull out malicious information. We already give some authoritarianism by having moderators we trust to remove spam and illegal content, so I don't see it as a giant overstep to make sure they can also do this.
- AnthonyMouse 1y ago
  > Moderation focuses on addressing the low quality, while curation makes sure tye high quality posts receive focus. This is effectively the same problem. The feed has a billion posts in it so if you're choosing from even the top half in terms of quality, the bottom decile is nowhere to be seen. > The latter poisons a site before it can really start to curate to its users. That's assuming you start off with a fire hose. Suppose you only see someone's posts in your feed if you a) visit their profile or b) someone you follow posted or liked it. > ensure you know when a human or a bot is posting. This is not possible and you should not attempt to do things that are known not to be possible. It doesn't matter what kind of verification you do. Humans can verify an account and then hand it to a bot to post things. Also, alts are good; people should be able to have an account for posting about computers and a different account for posting about cooking or travel or politics. What you're looking for is a way to rate limit account creation. But on day one you don't need that because your biggest problem is getting more users and by the time it's a problem you have a network effect and can just make them pay a pittance worth of cryptocurrency as a one-time fee if it's still a thing you want to do. > It can even extend in some domains to making sure e.g. That a person who says they worked at Google actually worked at Google. This is not a problem that social networks need to solve, but if it was you would just do it the way anybody else does it. If the user wants to know if someone really works for Google they contact the company and ask them, and if the company says no then you tell everybody that and anyone who doesn't believe you can contact the company themselves. > It may include addressing misinformation and hate, or removing non-transparent intimate advice like legal/medical claims without sources or proper licensing. If someone does something illegal then you have the government arrest them. If it isn't illegal then it isn't to be censored. There is nothing for a social media thing to be involved in here and the previous attempts to do it were in error. > It essentially boils down to making ensuring that "bad but popular" advice doesn't proliferate, as it it ought to do. To the extent that social media does such a thing, it does it exactly as above, i.e. as Reddit communities investigate things. If you want a professional organization dedicated to such things as an occupation, the thing you're trying to do is called investigative reporting, not social media. > I think we've seen enough examples of how dangerous "showing people what they actually want to see" can be if left unchecked. And the incentives to keep them up are equally dangerous in an ad-driven platform. No, they're much worse in an ad-driven platform, because then you're trying to maximize the amount of time people spend on the site and showing people rage bait and provocative trolling is an effective way to do that. What people want to see is like, a feed of fresh coupon codes that actually work, or good recipes for making your own food, or the result of the DIY project their buddy just finished. But showing you that doesn't make corporations the most money, so instead they show you somebody saying something political and provocative about vaccines because it gets people stirred up. Which is not actually what people want to see, which is why they're always complaining about it. > We already give some authoritarianism by having moderators we trust to remove spam and illegal content, so I don't see it as a giant overstep to make sure they can also do this. We should take away their ability to actually remove anything, censorship can GFTO, and instead give people a feed that they actually control and can therefore configure to not show that stuff because it is in reality not what they want to see.
tailrecursion 1y ago
> [...] show people posts from the people they follow or that the people they follow liked. Yes, this is a good system. It'll work particularly well at filtering spam because people largely agree what it is. One thing that will happen with your system is people will separate into cliques. But that's not the end of the world. Has anyone implemented Anthony's idea of using followees' likes to rank posts?
raincole 1y ago
> Then show people posts from the people they follow or that the people they follow liked Yeah, that's how you create echo-chambers where people truly believe vaccines cause autism, immigrants commit the most crimes, etc. When these people are inevitably showing at ballot, other people will call for more moderation. So no, this doesn't solve "moderation."
- dvdkon 1y ago
  But is being exposed to broader viewpoints or even being shown "correct" posts more often something that users of social media actually want? Social media is often seen as a public opinion manipulation machine, but I hope building another one of those is not the goal.

johnnyanmac 1y ago

>You need some way of distinguishing high quality from low quality posts. Yes. But I see curation more as a 2nd order problems to solve once the bases are taken care of. Moderation focuses on addressing the low quality, while curation makes sure tye high quality posts receive focus. The tools needed for curation, stuff like filtering, finding similar posts/comments, popularity, following, are different from those needed to moderate, or self moderate (ignore, down voting, reporting). The latter poisons a site before it can really start to curate to its users. >This is a completely independent problem from spam. Yeah, thinking more about it, it probably is a distinct category. It simply has a similar result of making a site unable to function. >It's not clear what these are but they sound like kind of the same thing again I can clarify. In short, posting transparency focused more on the user and good faith verification focuses more on the content. (I'm also horrible with naming, so I welcome better terms to describe these) - Posting transparency at this point has one big goal: ensure you know when a human or a bot is posting. But it extends to ensuring there's no impersonation, that there's no abuse of alt accounts, and no voting manipulation. It can even extend in some domains to making sure e.g. That a person who says they worked at Google actually worked at Google. But this is definitely a step that can overstep privacies. - good faith verification refers more towards a duty to properly vet and fact check information that is posted. It may include addressing misinformation and hate, or removing non-transparent intimate advice like legal/medical claims without sources or proper licensing. It essentially boils down to making ensuring that "bad but popular" advice doesn't proliferate, as it it ought to do. >they sound like elements in the authoritarian censorship toolbox which you don't actually need or want once you start showing people the posts they actually want to see Yes, they are. I think we've seen enough examples of how dangerous "showing people what they actually want to see" can be if left unchecked. And the incentives to keep them up are equally dangerous in an ad-driven platform. Being able to address that naturally requires some more authorian approaches. That's why "good faith" is an important factor here. Any authoritarian act you introduce can only work on trust, and is easily broken by abuse. If we want incentives to change from "maximizing engagement" to "maximizing quality and community", we need to cull out malicious information. We already give some authoritarianism by having moderators we trust to remove spam and illegal content, so I don't see it as a giant overstep to make sure they can also do this.

AnthonyMouse 1y ago
> Moderation focuses on addressing the low quality, while curation makes sure tye high quality posts receive focus. This is effectively the same problem. The feed has a billion posts in it so if you're choosing from even the top half in terms of quality, the bottom decile is nowhere to be seen. > The latter poisons a site before it can really start to curate to its users. That's assuming you start off with a fire hose. Suppose you only see someone's posts in your feed if you a) visit their profile or b) someone you follow posted or liked it. > ensure you know when a human or a bot is posting. This is not possible and you should not attempt to do things that are known not to be possible. It doesn't matter what kind of verification you do. Humans can verify an account and then hand it to a bot to post things. Also, alts are good; people should be able to have an account for posting about computers and a different account for posting about cooking or travel or politics. What you're looking for is a way to rate limit account creation. But on day one you don't need that because your biggest problem is getting more users and by the time it's a problem you have a network effect and can just make them pay a pittance worth of cryptocurrency as a one-time fee if it's still a thing you want to do. > It can even extend in some domains to making sure e.g. That a person who says they worked at Google actually worked at Google. This is not a problem that social networks need to solve, but if it was you would just do it the way anybody else does it. If the user wants to know if someone really works for Google they contact the company and ask them, and if the company says no then you tell everybody that and anyone who doesn't believe you can contact the company themselves. > It may include addressing misinformation and hate, or removing non-transparent intimate advice like legal/medical claims without sources or proper licensing. If someone does something illegal then you have the government arrest them. If it isn't illegal then it isn't to be censored. There is nothing for a social media thing to be involved in here and the previous attempts to do it were in error. > It essentially boils down to making ensuring that "bad but popular" advice doesn't proliferate, as it it ought to do. To the extent that social media does such a thing, it does it exactly as above, i.e. as Reddit communities investigate things. If you want a professional organization dedicated to such things as an occupation, the thing you're trying to do is called investigative reporting, not social media. > I think we've seen enough examples of how dangerous "showing people what they actually want to see" can be if left unchecked. And the incentives to keep them up are equally dangerous in an ad-driven platform. No, they're much worse in an ad-driven platform, because then you're trying to maximize the amount of time people spend on the site and showing people rage bait and provocative trolling is an effective way to do that. What people want to see is like, a feed of fresh coupon codes that actually work, or good recipes for making your own food, or the result of the DIY project their buddy just finished. But showing you that doesn't make corporations the most money, so instead they show you somebody saying something political and provocative about vaccines because it gets people stirred up. Which is not actually what people want to see, which is why they're always complaining about it. > We already give some authoritarianism by having moderators we trust to remove spam and illegal content, so I don't see it as a giant overstep to make sure they can also do this. We should take away their ability to actually remove anything, censorship can GFTO, and instead give people a feed that they actually control and can therefore configure to not show that stuff because it is in reality not what they want to see.

tailrecursion 1y ago

> [...] show people posts from the people they follow or that the people they follow liked. Yes, this is a good system. It'll work particularly well at filtering spam because people largely agree what it is. One thing that will happen with your system is people will separate into cliques. But that's not the end of the world. Has anyone implemented Anthony's idea of using followees' likes to rank posts?

raincole 1y ago

> Then show people posts from the people they follow or that the people they follow liked Yeah, that's how you create echo-chambers where people truly believe vaccines cause autism, immigrants commit the most crimes, etc. When these people are inevitably showing at ballot, other people will call for more moderation. So no, this doesn't solve "moderation."

dvdkon 1y ago
But is being exposed to broader viewpoints or even being shown "correct" posts more often something that users of social media actually want? Social media is often seen as a public opinion manipulation machine, but I hope building another one of those is not the goal.