3 ms·

Seems like this is an easy problem to solve. Allow users to control their own content filters instead of removing the offending content from the platform altog

Seems like this is an easy problem to solve. Allow users to control their own content filters instead of removing the offending content from the platform altogether. Unless this isn't actually about that, and is really just people trying to shut down opinions they don't agree with...

programmarchy 6y ago
I've wondered this myself. Platforms could censor illegal material as required by law, then leave the gray areas to users. There's a pretty big technical challenge, though, because instead of training a single one-size-fits-all algorithm, you suddenly have to train N algorithms, where N is the number of users on your platform. In order to scale, this training may have to be done on the client. But in addition to the technical challenge, I doubt these platforms want to cede the enormous power they have over information distribution. Right now Twitter can flex on the president of a global hegemon, which is crazy power. Why would they give that up?
- travisoneill1 6y ago
  N is the number of content types (still a lot), so FB only needs one misinformation filter, and you the user could choose whether or not you want that one.
javagram 6y ago
This is not about user filters. I haven’t seen trump’s post on my Facebook because I don’t follow him. This is about employees who feel that Trump is so dangerous that his posts should be deleted or hidden to prevent his willing followers from seeing them. Needless to say, any open protocol Facebook alternative would still allow Trump to broadcast his evil messages. Just like he is able to send emails and texts with the messages to his supporters today. Is the next step AT&T and Verizon and TMobile employees demanding trump campaign SMSes be censored ? What is the meaningful difference between that and a tweet/post?

3 ms·

Seems like this is an easy problem to solve. Allow users to control their own content filters instead of removing the offending content from the platform altog

by travisoneill1 6y ago

programmarchy 6y ago
I've wondered this myself. Platforms could censor illegal material as required by law, then leave the gray areas to users. There's a pretty big technical challenge, though, because instead of training a single one-size-fits-all algorithm, you suddenly have to train N algorithms, where N is the number of users on your platform. In order to scale, this training may have to be done on the client. But in addition to the technical challenge, I doubt these platforms want to cede the enormous power they have over information distribution. Right now Twitter can flex on the president of a global hegemon, which is crazy power. Why would they give that up?
- travisoneill1 6y ago
  N is the number of content types (still a lot), so FB only needs one misinformation filter, and you the user could choose whether or not you want that one.
javagram 6y ago
This is not about user filters. I haven’t seen trump’s post on my Facebook because I don’t follow him. This is about employees who feel that Trump is so dangerous that his posts should be deleted or hidden to prevent his willing followers from seeing them. Needless to say, any open protocol Facebook alternative would still allow Trump to broadcast his evil messages. Just like he is able to send emails and texts with the messages to his supporters today. Is the next step AT&T and Verizon and TMobile employees demanding trump campaign SMSes be censored ? What is the meaningful difference between that and a tweet/post?