zlacker

I don't want moderation OR censorship. I want proper labeling so that I can self-filter out things that I don't want to see. Just like the labels on cans of food tells you exactly what ingredients are found within the can so you can skip the ones with peanuts, gluten, or some other substance you don't want; I want labels (and levels) on data.

If I don't want to see profanity, I should be able to set my filter to exclude profane comments. If I don't want to see nudity, I can set that filter too. Just like movies get a certain rating (G, PG, R, etc.), we should be able to properly label data.

>>didget+(OP)
I'm with you on this. It also occurs to me, though, that by rigorously labeling all offensive content, you give ai a huge boost in creating it with perfect fidelity.

Which inspires another really weird, super uncomfortable thought. If the CSAM producers had cheap, reliable methods of creating their awful content without the use of real people, would that reduce the harm done?

I can't remember the last time I felt so conflicted just asking a question, but there it is.

>>pksebb+S8
I hear you. But maybe AI shouldn't even be in the equation. It's expecting a robot to solve a people problem. We try to do this a lot in the technology industry. But no matter how great technology is, it's a mistake to think that it can fully solve social and communication problems.

>>didget+(OP)
Labels themselves would become an ideological battleground. Imagine what reddit would look like if anyone could freely and publicly tag a post with something like "Libtard" or "Chud".

>>didget+(OP)
> I want proper labeling so that I can self-filter out things that I don't want to see.

This is basically what he describes in the article as a form of moderation: "If you wanted to get fancy, you could have a bunch of filters - harassing content, sexually explicit content, conspiracy theories - and let people toggle which ones they wanted to see vs. avoid."

>>didget+(OP)
Who is going to set the labels, and how will they do it in real-time?

>>Monkey+lm
Subjective labels being applied subjectively isn't as big of a problem though, since it'll probably be clear to most people just from the name that they're subjective. More objective labels will naturally be more useful, and therefore more popular. The only people who are going to use "Libtard" as a filter are those actively trying to create an echo chamber for themselves.

If your concern is with the labels themselves being used to convey a (possibly offensive) message, I think you could just have a way for people to hide specific labels and never see them again. Or maybe a way to label the labels as subjective, or just delete ones that are obvious flamebait.

>>zach_g+gJ
I would settle for some simple categories like 'Profanity', 'Sex', 'Nudity', and 'Violence' with a level 0 to 9 with specific criteria for each level (zero meaning none, 9 = full of it). Let each content author 'self-report' initially but let others up or down vote if the author 'mis-labels' something.

>>Ajedi3+JW
I was thinking along the lines of the second scenario where the labels are both the target and vector of abuse. Moderating what labels are allowed opens you up to claims of censorship again.

People love to misuse tools meant for good, on Reddit I've been on the receiving end of the "reddit cares" self-harm notification because of some barely spicy comments.

>>bryan0+9u
Who gets to choose what is labelled sexually explicit or a conspiracy? At the beginning of the war in Ukraine there were accusations of bot farms mass flagging people's streams as containing adult content to get them silenced.