4 ms·

Moderation decisions should be transparent, rules-based, and follow basic due process principles. That is not so far from what we have already and it is a spe

Moderation decisions should be transparent, rules-based, and follow basic due process principles. That is not so far from what we have already and it is a spectacular failure! No, moderation decisions should be transparently arbitrary (as opposed to rules-based), and offer no 'due process' whatsoever. I would prefer a social media ecosystem with dozens of players and a philosophy more like this: Remember you are a guest on our server; we can kick you off if we think you're a liability to our reputation; and if we do kick you off, you won't be getting any months-long court of appeals because we owe you nothing but an archive of your data. Don't be an idiot.

naasking 5y ago
> That is not so far from what we have already and it is a spectacular failure! We have nothing like a transparent, rules-based system based on basic due process. The rules are vague and broad, the transparency is basically nil, and due process does not exist at all. People are banned and then rules are invented post-hoc to justify them, there is no discovery, and corporations have total authority to do as they please without restriction. In what non-authoritarian country does this resemble anything like a legal system?
- uniqueid 5y ago
  Twitter, FB, etc still try to do rule-based moderation. Sometimes they later discover there's problem content that the rules don't cover. They still think they need to do (corporate) soul-searching and provide justifications to the public when they moderate. Since the way they moderate is awful, they often lie or reverse course when doing so. They still have these complicated appeals boards and processes. But since it's absurd to have some committee deliberating about most 'tweets' these processes are slow or seldom-used. So I consider what they do 'not so far' from what the essay recommends because it's the same paradigm, just terribly implemented. My opinion is that it's not desirable, and maybe impossible, to implement moderation well using this paradigm. If a guy walks into a hotel and takes a crap in the lobby, the hotel doesn't check a flowchart before kicking him out; the hotel doesn't need any justification aside from 'he defecated in the lobby', and if the guy petitions the hotel to crap - on a later occasion - in the lobby again, the hotel won't assemble a committee to tell him 'no'
  - vorpalhex 5y ago
    The hotel can't keep the guys belongings in that case. When your business transaction is fleeting, you can handle exceptions easily. You can just stop selling coffee to the guy who refuses to pay or makes rude comments. On the other hand, you can't kick out a tenant because they are rude or even if they shit on the floor. You have to go through a process that has external, third party controls. The tenant is allowed to retrieve their property. You can't prevent the tenant from being rented to again. Google, Facebook and co are basically digital landlords. The tenants get rights too.
    - uniqueid 5y ago
      I do agree these companies have an obligation to provide a user with the data he or she creates. We might disagree over how far that goes. Personally I think they just owe the user a zip archive. My hotel analogy probably complicated things. The main problem I have with moderation on social media is that I think moderation is necessary (to keep the world relatively sane) but also impossible to implement objectively. So it's better to embrace the subjectivity than to obfuscate it with flowcharts and procedures.
      - vorpalhex 5y ago
        Yeah I agree that I don't think any platform is going to be objective, and I think a lot of users also don't want strictly objective moderators. One of the reasons I want the fediverse to do well is because it can solve this case. Your local instance can be moderated as you desire - heavy handedly or barely at all - and you can still communicate with other folks. However, we do need some kind of user rights to enable platform migrating. Something like mobile phone carriers where they have to unlock your phone and let you port your number. Social media companies need to let your identity and data be portable.
  - naasking 5y ago
    > Twitter, FB, etc still try to do rule-based moderation. Sometimes they later discover there's problem content that the rules don't cover. Ex post facto laws are prohibited, ie. retroactively make something a crime and then charge people for breaking that law. This is something Twitter et all do all the time. The law is rules-based, and something like that is what this article is recommending. Whatever Twitter et al are doing is not that. We can take a cynical guess at what they're doing though: negative PR minimization. Negative PR is proportional to the size of the mob that's upset about it, so this ultimately reduces to mob rule. The law is intended to eliminate mob rule. > They still have these complicated appeals boards and processes. Their appeals process is not complicated. It's "my word goes, so you'll take my judgment and like it; you have no representation or say in these decisions, and I'm not going to tell you what rule you broke or specifically how you broke that rule; I can in fact change any rule arbitrarily so that you are ex post facto guilty, and there's nothing you can do about it, but we're going to be performative about abiding to 'rules' to project the illusion of impartiality and fairness, unless of course the rules inconvenience us". There are countless examples of this. Governments also try these tactics from time to time, but with the separation of powers they are often smacked back in line by the courts. Do you see a comparable separation of powers in these corporations? > So I consider what they do 'not so far' from what the essay recommends because it's the same paradigm, just terribly implemented. I disagree. It's not the same paradigm because the principles and processes at play are not remotely the same. These companies are acting on the principle maximizing their profitability, which I think leads to the negative PR minimization principle, which ultimately reduces to a form of mob rule. This is often the opposite of fair and just, and this article is about requiring these corporations to legally uphold other principles to re-align their incentives for fairer outcomes.

naasking 5y ago

> That is not so far from what we have already and it is a spectacular failure! We have nothing like a transparent, rules-based system based on basic due process. The rules are vague and broad, the transparency is basically nil, and due process does not exist at all. People are banned and then rules are invented post-hoc to justify them, there is no discovery, and corporations have total authority to do as they please without restriction. In what non-authoritarian country does this resemble anything like a legal system?

uniqueid 5y ago
Twitter, FB, etc still try to do rule-based moderation. Sometimes they later discover there's problem content that the rules don't cover. They still think they need to do (corporate) soul-searching and provide justifications to the public when they moderate. Since the way they moderate is awful, they often lie or reverse course when doing so. They still have these complicated appeals boards and processes. But since it's absurd to have some committee deliberating about most 'tweets' these processes are slow or seldom-used. So I consider what they do 'not so far' from what the essay recommends because it's the same paradigm, just terribly implemented. My opinion is that it's not desirable, and maybe impossible, to implement moderation well using this paradigm. If a guy walks into a hotel and takes a crap in the lobby, the hotel doesn't check a flowchart before kicking him out; the hotel doesn't need any justification aside from 'he defecated in the lobby', and if the guy petitions the hotel to crap - on a later occasion - in the lobby again, the hotel won't assemble a committee to tell him 'no'
- vorpalhex 5y ago
  The hotel can't keep the guys belongings in that case. When your business transaction is fleeting, you can handle exceptions easily. You can just stop selling coffee to the guy who refuses to pay or makes rude comments. On the other hand, you can't kick out a tenant because they are rude or even if they shit on the floor. You have to go through a process that has external, third party controls. The tenant is allowed to retrieve their property. You can't prevent the tenant from being rented to again. Google, Facebook and co are basically digital landlords. The tenants get rights too.
  - uniqueid 5y ago
    I do agree these companies have an obligation to provide a user with the data he or she creates. We might disagree over how far that goes. Personally I think they just owe the user a zip archive. My hotel analogy probably complicated things. The main problem I have with moderation on social media is that I think moderation is necessary (to keep the world relatively sane) but also impossible to implement objectively. So it's better to embrace the subjectivity than to obfuscate it with flowcharts and procedures.
    - vorpalhex 5y ago
      Yeah I agree that I don't think any platform is going to be objective, and I think a lot of users also don't want strictly objective moderators. One of the reasons I want the fediverse to do well is because it can solve this case. Your local instance can be moderated as you desire - heavy handedly or barely at all - and you can still communicate with other folks. However, we do need some kind of user rights to enable platform migrating. Something like mobile phone carriers where they have to unlock your phone and let you port your number. Social media companies need to let your identity and data be portable.
- naasking 5y ago
  > Twitter, FB, etc still try to do rule-based moderation. Sometimes they later discover there's problem content that the rules don't cover. Ex post facto laws are prohibited, ie. retroactively make something a crime and then charge people for breaking that law. This is something Twitter et all do all the time. The law is rules-based, and something like that is what this article is recommending. Whatever Twitter et al are doing is not that. We can take a cynical guess at what they're doing though: negative PR minimization. Negative PR is proportional to the size of the mob that's upset about it, so this ultimately reduces to mob rule. The law is intended to eliminate mob rule. > They still have these complicated appeals boards and processes. Their appeals process is not complicated. It's "my word goes, so you'll take my judgment and like it; you have no representation or say in these decisions, and I'm not going to tell you what rule you broke or specifically how you broke that rule; I can in fact change any rule arbitrarily so that you are ex post facto guilty, and there's nothing you can do about it, but we're going to be performative about abiding to 'rules' to project the illusion of impartiality and fairness, unless of course the rules inconvenience us". There are countless examples of this. Governments also try these tactics from time to time, but with the separation of powers they are often smacked back in line by the courts. Do you see a comparable separation of powers in these corporations? > So I consider what they do 'not so far' from what the essay recommends because it's the same paradigm, just terribly implemented. I disagree. It's not the same paradigm because the principles and processes at play are not remotely the same. These companies are acting on the principle maximizing their profitability, which I think leads to the negative PR minimization principle, which ultimately reduces to a form of mob rule. This is often the opposite of fair and just, and this article is about requiring these corporations to legally uphold other principles to re-align their incentives for fairer outcomes.