4 ms·

The following link contains a racial slur in the text of a screenshot - you probably don't want to open it at work. https://www.reddit.com/r/ParlerWatch/commen

by throwaway19937 6y ago

The following link contains a racial slur in the text of a screenshot - you probably don't want to open it at work. https://www.reddit.com/r/ParlerWatch/comments/ktwmje/this_is_the_type_of_free_speech_parler_accepts/ https://www.reddit.com/r/ParlerWatch/comments/ktwmje/this_is... is a concrete example of speech that would be banned from Twitter and Facebook. I think it's telling that there are 25k up votes.

Moodles 6y ago
You have provided an exmaple (I assume, I didn't click the link because I have a productivity blocker which I obviously need to update for Hacker News, lol) of a bad thing on one platform, but not another. But ok, are there vice versa examples? What I'm getting at is, I understand Parler is generally a right-leaning platform, and therefore the types of "hate" will be right-leaning. Twitter is a generally left-leaning platform, so I would expect their type of hate to be generally left-leaning. So I think a tweet about e.g. storming the Capitol being banned on Twitter but not Parler isn't good evidence that Twitter is better. Because, for example, perhaps a violent Antifa tweet would be left alone on Twitter but censored on Parler. Perhaps some ML bot can quantity sentiment of tweets. That's the kind of evidence I would like to see.
cwkoss 6y ago
Twitter, Youtube, Facebook all get hundreds of comments of similar nature every hour. They tend to be pretty good about removing them promptly. If I start a new social media service, what are the objective measures I can use to ensure I'm moderating well and quickly enough? How should we allow this to interact with scale? If a new startup goes from 10 to 10,000 users overnight, and 0.5% of users post offensive comments that first day, how long does that startup have to remove them all before they are designated as "bad"? Also, how should we view overmoderation within this context? I think most people would find it injust and orwellian to ban a black sociology professor for posting an academic analysis of the use of the n-word through history, but I fear society is going to overcorrect.

Moodles 6y ago

You have provided an exmaple (I assume, I didn't click the link because I have a productivity blocker which I obviously need to update for Hacker News, lol) of a bad thing on one platform, but not another. But ok, are there vice versa examples? What I'm getting at is, I understand Parler is generally a right-leaning platform, and therefore the types of "hate" will be right-leaning. Twitter is a generally left-leaning platform, so I would expect their type of hate to be generally left-leaning. So I think a tweet about e.g. storming the Capitol being banned on Twitter but not Parler isn't good evidence that Twitter is better. Because, for example, perhaps a violent Antifa tweet would be left alone on Twitter but censored on Parler. Perhaps some ML bot can quantity sentiment of tweets. That's the kind of evidence I would like to see.

cwkoss 6y ago

Twitter, Youtube, Facebook all get hundreds of comments of similar nature every hour. They tend to be pretty good about removing them promptly. If I start a new social media service, what are the objective measures I can use to ensure I'm moderating well and quickly enough? How should we allow this to interact with scale? If a new startup goes from 10 to 10,000 users overnight, and 0.5% of users post offensive comments that first day, how long does that startup have to remove them all before they are designated as "bad"? Also, how should we view overmoderation within this context? I think most people would find it injust and orwellian to ban a black sociology professor for posting an academic analysis of the use of the n-word through history, but I fear society is going to overcorrect.