you are viewing a single comment's thread.

view the rest of the comments →

[–]DrStrangelove 3 insightful - 1 fun3 insightful - 0 fun4 insightful - 1 fun -  (12 children)

After a quick read, that's basically what the EO does. Sites that editorialize/censor/ban user generated content will no longer be afforded the liability protection offered under section 230 of the communications decency act, and can be sued for libel/defamation. To retain immunity, sites may not censor protected speech.

[–]kokolokoNightcrawler 4 insightful - 1 fun4 insightful - 0 fun5 insightful - 1 fun -  (10 children)

And I think this would suffice for now until they start relying only on shadow ban algos.

[–]LarrySwinger2Voluntaryist 1 insightful - 1 fun1 insightful - 0 fun2 insightful - 1 fun -  (9 children)

How would shadowbans protect them? The ones I'm familiar with can be detected simply by logging out and seeing if your post is still visible.

[–]kokolokoNightcrawler 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (8 children)

They are harder to prove than outright bans and censorship, plus, companies can resort to blaming buggy algos if sued.

[–]Canbot 3 insightful - 1 fun3 insightful - 0 fun4 insightful - 1 fun -  (7 children)

They would have to provide the algo for review during trial to prove that. If the algo is purposefully written for censorship it would be revealed.

[–]kokolokoNightcrawler 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (6 children)

You'd need a warrant for that...This is how sjw censorship started in the first place. They started with disappearing videos, comments, posts, making them quarantined, cordoned, unsearchable and so on and on...As time passed, and they've gotten almost zero pushback from the lawmakers, they got more and more brazen where now they stand fully behind their totalitarian decisions.

[–]Canbot 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (5 children)

You dont need a warrant, it's called discovery and is part of every lawsuit. If they want to use the algo as an excuse in court they have to prove it was the algo and the prosecution has the right to analyze it.

[–]kokolokoNightcrawler 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (4 children)

How does one prove a disappeared post without the access to a database and the servers...It's a complex issue. If you have thousands of lines of code responsible for keeping full functionality of comments for all users across the board, that's thousands of coggss in a machine that need inspecting to determine what actually happened. And, i'm sure with bigger socials, it's millions of lines.

[–]Canbot 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (0 children)

You prove it by having multiple computers with different accounts show that it does not come up when they look at the post.

If the defense wants to claim that it is a software issue and not intentional it is on them to prove it.

[–]magnora7 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (2 children)

Saidit has about 1.1 million lines of code, for example, using 8 different programming languages. So yeah...

[–]jamesK_3rd 1 insightful - 1 fun1 insightful - 0 fun2 insightful - 1 fun -  (0 children)

But as with any lawsuit they'll have a team of coders looking over the code.

Additionally, it isn't hard to look at certain macros, classes, or particular repetitive processes that would generally be used to censor or shadow certain content. It's the same when you grab the source from git, unless your doing a complete rework, lots of things are placed intuitively.

And ironically, knowing devs and the fact that most of the code is thought to be proprietary and never to be seen, they'd likely have some interesting comments on the code.

[–]magnora7 3 insightful - 1 fun3 insightful - 0 fun4 insightful - 1 fun -  (0 children)

To retain immunity, sites may not censor protected speech.

This is the key phrase.

Honestly I don't think it's that huge of a change, it's just codifying what's already in place: Websites that are "publishers" can get sued for libel/defamation, and websites that are just black boxes of information without reviewing the truth content of the information are immune.

Saidit is firmly in the latter category, so I think we're fine. (As long as the law isn't over-reaching and full of nasty loopholes, which is very possible)