I feel this won’t happen. If big subs continue to be dark too long, the reddit admins would simply remove the hostile volunteer mods and reopen the subs. The mods are used to being gods of their little domains. If they cross the line, they will be reminded that they own nothing. They can obey reddit or they can be replaced.
That is what I see in the future for any mods that try to hold subs hostage indefinitely.
A possible problem is that they would be forced to find new volunteers to run them. While I bet there’s many who want to be “gods” I bet it’s harder to find people who can do it well enough to run a 10+ million forum. Especially hamstringed by reddits lack of modtools.
So sure, Reddit can remove the mods and do it multiple times but it will continuously lead to a worse experience and sooner or later an unacceptable amount of spam, hate and CP will cause the advertisers to pull their ads.
This is absolutely true. There are often calls for ‘anyone want to mod’ on even smaller subs… and you know, it sounds fun to a lot of folks at first. But if you’ve ever actualy been a mod, even of a smaller community online? It loses its appeal very quickly.
I’ve seen plenty of communities where it’s clear that the mods only stop by from time to time and they get by just fine, spam and malicious posts will still be a small minority. Some set automod on a shoot first, ask questions later setting where all reported comments get deleted until the mod restores them.
I really don’t think finding new moderation will be an issue. As much as it would be nice for Reddit to be screwed over by the mods it’s going to be a non-issue for them, there’s already measures in place to prevent subreddit parking and plenty of willing volunteers.
I feel like ad revenue is not their top monetization priority personally. It’s speculation of course. But I think they are learning that the free content the users create will generate much more revenue from mega corps who want access to all of it to train emerging AIs. Data, specifically YOUR data is valuable. What posts do you look at? What do you upvote? What do you downvote? What subreddits do you subscribe to? There is a wealth of information they will monetize. This is why I think they don’t care that the little app devs can’t afford their new API pricing. They can’t give the app devs one price then think Microsoft, Google, Apple and other multi-billion dollar corps would pay a higher price.
Again, this is just my speculation. But the suddenness and the exorbitant price means they want to act now, and capitalize on this new market while it’s good. Their terms of service specifically say everything you post, you give them a license to use, sell, or sub-license without dispute, forever. This isn’t about ads.
I was thinking the same thing. Probably why the timeline is so fast too with only giving people a month’s notice of the API costs. And could also be true of twitter.
ChatGPT and other LLMs are gaining a lot of value from information freely available online and sites with large user generated text submissions like Reddit/twitter want a piece of the pie.
The problem with this line of thinking is “why is 3rd party API data any less valuable than 1st party API data for AI training?” While this may be true, I don’t see this particular move being motivated by AI. They still have all the API calls and interactions even if they aren’t being made by Reddit’s own apps.
I may be wrong here, but what I mean is they have ways to stop LLM companies from web scraping all of Reddit. The only other way the likes of chatGPT can get all the info is through to API which is currently free. So I think Reddit might be doing is saying this information isn’t free so pay X amount for access to our data.
Obviously 3rd party apps like Apollo won’t pay that, but Google and OpenAi probably will.
I’m not too sure what you mean by api data being worth less or more, it’s all the same data.
“TL;DR: Pushshift is in violation of our Data API Terms and has been unresponsive despite multiple outreach attempts on multiple platforms, and has not addressed their violations. Because of this, we are turning off Pushshift’s access to Reddit’s Data API, starting today”
I don’t disagree, that’s just the kind of response I expect will happen. I mean the CEO was trying to gaslight a major app dev, I feel like they will lie about and demonize mods that go against them.
A lot of subreddits will stay closed indefinitely.
I feel this won’t happen. If big subs continue to be dark too long, the reddit admins would simply remove the hostile volunteer mods and reopen the subs. The mods are used to being gods of their little domains. If they cross the line, they will be reminded that they own nothing. They can obey reddit or they can be replaced.
That is what I see in the future for any mods that try to hold subs hostage indefinitely.
A possible problem is that they would be forced to find new volunteers to run them. While I bet there’s many who want to be “gods” I bet it’s harder to find people who can do it well enough to run a 10+ million forum. Especially hamstringed by reddits lack of modtools.
So sure, Reddit can remove the mods and do it multiple times but it will continuously lead to a worse experience and sooner or later an unacceptable amount of spam, hate and CP will cause the advertisers to pull their ads.
This is absolutely true. There are often calls for ‘anyone want to mod’ on even smaller subs… and you know, it sounds fun to a lot of folks at first. But if you’ve ever actualy been a mod, even of a smaller community online? It loses its appeal very quickly.
Real-world equivalent: “You should run for HOA President!”
Or, union president. Or pto president. Or fucking president of anything else that’s volunteer run. Just. No. Some things aren’t worth it.
I am curious how much time you would say a mod spends a day modding
I’ve seen plenty of communities where it’s clear that the mods only stop by from time to time and they get by just fine, spam and malicious posts will still be a small minority. Some set automod on a shoot first, ask questions later setting where all reported comments get deleted until the mod restores them.
I really don’t think finding new moderation will be an issue. As much as it would be nice for Reddit to be screwed over by the mods it’s going to be a non-issue for them, there’s already measures in place to prevent subreddit parking and plenty of willing volunteers.
I feel like ad revenue is not their top monetization priority personally. It’s speculation of course. But I think they are learning that the free content the users create will generate much more revenue from mega corps who want access to all of it to train emerging AIs. Data, specifically YOUR data is valuable. What posts do you look at? What do you upvote? What do you downvote? What subreddits do you subscribe to? There is a wealth of information they will monetize. This is why I think they don’t care that the little app devs can’t afford their new API pricing. They can’t give the app devs one price then think Microsoft, Google, Apple and other multi-billion dollar corps would pay a higher price.
Again, this is just my speculation. But the suddenness and the exorbitant price means they want to act now, and capitalize on this new market while it’s good. Their terms of service specifically say everything you post, you give them a license to use, sell, or sub-license without dispute, forever. This isn’t about ads.
I was thinking the same thing. Probably why the timeline is so fast too with only giving people a month’s notice of the API costs. And could also be true of twitter.
ChatGPT and other LLMs are gaining a lot of value from information freely available online and sites with large user generated text submissions like Reddit/twitter want a piece of the pie.
The problem with this line of thinking is “why is 3rd party API data any less valuable than 1st party API data for AI training?” While this may be true, I don’t see this particular move being motivated by AI. They still have all the API calls and interactions even if they aren’t being made by Reddit’s own apps.
I may be wrong here, but what I mean is they have ways to stop LLM companies from web scraping all of Reddit. The only other way the likes of chatGPT can get all the info is through to API which is currently free. So I think Reddit might be doing is saying this information isn’t free so pay X amount for access to our data.
Obviously 3rd party apps like Apollo won’t pay that, but Google and OpenAi probably will.
I’m not too sure what you mean by api data being worth less or more, it’s all the same data.
Almost certainly true. Historically, you’d just grab the dumps from pushshift.
https://www.reddit.com/r/modnews/comments/134tjpe/reddit_data_api_update_changes_to_pushshift_access/
“TL;DR: Pushshift is in violation of our Data API Terms and has been unresponsive despite multiple outreach attempts on multiple platforms, and has not addressed their violations. Because of this, we are turning off Pushshift’s access to Reddit’s Data API, starting today”
This makes sense. I get the argument now. Thanks!
I’d like to see the mods wipe the subs on their way out if reddit continues to ignore the community.
That sends even more of a bad message to the users and I would expect even more people to leave due to all the fuckery going on.
I don’t disagree, that’s just the kind of response I expect will happen. I mean the CEO was trying to gaslight a major app dev, I feel like they will lie about and demonize mods that go against them.
Sadly, I think you’re right. And with the wild number of bootlickers out there, they be able to find new mods for those subs in a. Maatter of minutes.
these subs will turn into dumpsters due to lack of moderation and people hopefully will leave