Looking at this site-wide blackout planned (100M+ users affected), it’s clear that if reddit could halt the moderators from protesting the would.
If their entire business can be held hostage by a few power mods, then it’s in their best interest to reduce risk.
Reddit almost 2 decades worth flagged content for various reasons. I could see a future in which all comments are first checked by a LLM before being posted.
Using AI could handle the bulk of automation and would then allow moderation do be done entirely by reddit in-house or off-shore with a few low-paid workers as is done with meta and bytedance.
I don’t think it will happen. They can’t even fix already existing tools that are supposed to reduce the workload of the current moderators. Ironically they are also massively relying on 3rd party Bots for things they could have introduced years ago for all subreddits by default.
They also can’t use “a LLM”, they would have to give each subreddit a own LLM (or at least make small Subreddit groups). Each subreddit has different moderation requirements, small nuances that change how a Subreddit has to be Moderated. You can’t moderate all subs exactly the same. Also, it will be really hard to not decrease the satisfaction in the Community by only looking black and white at some things, Moderating is full of individual Decisions. You also can’t really outsource local subreddit to someone that hasn’t even a clue if the construction site that’s discussed exists and doesn’t know wordplays that are new and only used in that specific region.
If their entire business can be held hostage by a few power mods, then it’s in their best interest to reduce risk.
Reddit has only a low percentage of power Mods and they probably don’t involve themselves into this issue. The Majority of Moderators that want to prevent this change are people that do a ton of work for their subreddit and REALLY care about them.
As much as he probably wants, because he stopped caring a long ago, Steve will need to continue to involve voluntary Moderators. He also should start to make long term decisions because most of the recent changes may be short term improvements for the ad revenue but they made actually using the site a lot worse. (Ironically they are also destroying their own App while they want to force everyone from 3rd party to)
Each subreddit has different moderation requirements, small nuances that change how a Subreddit has to be Moderated.
Good point. I was mostly thinking in moderation in terms of abuse and breaking site-wide rules. One challenge would be if there’s a rule against posting spoilers without tags, how would the LLM know if they’re spoiling the movie without the script in the training data.
Reddit has only a low percentage of power Mods and they probably don’t involve themselves into this issue. https://www.reddit.com/r/ModCoord/comments/1401qw5/incomplete_and_growing_list_of_participating/
There’s over 100 million users in all of these subreddits, this can’t be ignored and mainstream news will cover this more
Long-term i believe human moderators for the largest subreddits will be a target for automation while trying to scale this to the entire site.
Sorry for the late reply, haven’t figured notification out here.
My subs don’t have a spoiler risk but for some things the Automoderator is already useful. For example we have a Script that makes a post nsfw if the comments mention that. It also reports the post for later review. Same could be done with spoilers, you can react as soon as someone complaints without knowing the Script.
It won’t be ignored but I’m certain that the Powermods don’t care. The few real power Mods I know of neither care about the site or the Subreddits. They just want to be mod in as much subs as possible.
Long time I expect some bots to become tightly integrated like the repostmasterbot, BotDefense and some others but the configuration and majority of work will still be with those that are responsible to shape their subreddit. The mods.
Here we come Dead Internet
I don’t understand the reason for the reference. I’m Moderating a few mid sized Subreddits and we outsource a ton of stuff to Bots like a lot of subs. Wouldn’t say it’s dead just because we try to automate as muchas possible. Moderating a community is a unpaid use of spare time so it helps to decrease the time I have to invest in repetitive stuff.
To my understanding, the concept of dead internet current day is conflated metrics due to non-human interaction. If there are 1mil HTTPS transactions and 400,000 of those are non-human, it skews things, making it look like it’s more highly used.
I’m not saying automation’s bad, hell it’s my job, just an interesting thing to think about that we’re reaching the point where most transactions (API, HTTP, etc, not financial) are handled by non-human entities.
So in this context, that we are literally trading out people for automation (via AI) is this concept by definition :)