Timeline for Moderation Strike: Stack Overflow, Inc. cannot consistently ignore, mistreat, and malign its volunteers
Current License: CC BY-SA 4.0
20 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 6, 2023 at 0:36 | comment | added | Ryan M | "a day of review that starts on a holiday in the US, Canada, UK, and other countries doesn't seem to be consistent with the spirit of the agreement" Especially when said day of "review" is after the policy is effective, which occurred simultaneously with it being given to moderators for said "review". The moderator agreement requires a "preview for review"; this was not that, in either letter or spirit. | |
| Jun 5, 2023 at 19:43 | comment | added | starball | This statement now posted to MSE at "Statement from SO: Moderator Action today". On the bright side, this is a step up from the silence on some other similar scenarios. Ex. Why hasn't there been an official community announcement or blogpost about charging for usage of subscriber content in the training of LLMs? | |
| Jun 5, 2023 at 16:28 | comment | added | Stargateur | the lack of transparency is alone enough to justify the strike anyway. | |
| Jun 5, 2023 at 15:43 | comment | added | Bryan Krause | Really, it makes the CEO's claim that it is suspensions for AI content that have driven people away quite foolish, because the number affected is so small, while emphasizing the importance of moderating this content in the first place: it's what keeps the amount of AI content manageable. | |
| Jun 5, 2023 at 15:41 | comment | added | Bryan Krause | @Trilarion The amount of AI content was initially quite high, especially on SO; then, in consultation with staff, moderators took away the "carrot" for posting AI content: by deleting such content and suspending accounts when necessary, we made it so the value of posting AI-generated content here was low, which reduced submissions of that content. The same concept applies to all moderation: you aren't just acting on certain content, but discouraging it in the first place. | |
| Jun 5, 2023 at 15:30 | comment | added | Thomas Owens | @Trilarion A lot of the heuristics mods used, including our assessments of various detectors, was in mod only spaces to prevent people from being able to make the most minimal of changes to avoid detection. We don't discuss other moderator tools or techniques in public for similar reasons. | |
| Jun 5, 2023 at 15:29 | comment | added | NoDataDumpNoContribution | From the details it looks as if the number of suspensions for posting AI generated content wasn't very large and every single case could be vetted carefully. This sounds good but is a bit in contrast with the warnings that AI generators would make it so easy for people to post such content. I would have expected an actually really large number of suspensions then. Maybe some stats about how many accounts were suspended for this (especially on SO) would be helpful. | |
| Jun 5, 2023 at 15:17 | comment | added | NoDataDumpNoContribution | "is something that has been discussed" My impression is that publicly not so much was discussed. For example see Makyen on How can we determine whether an answer used ChatGPT? "The general question of how to detect these posts is something we've been avoiding publicly answering..." So basically a no comment. I would have liked to discuss it more. | |
| Jun 5, 2023 at 15:13 | comment | added | Bryan Krause | On Academia.SE, the percentage of striking moderators is 100%. One of us already resigned due to the policy and the manner it was passed down. The policy is against academic ethics of attribution and honesty, and potentially dangerous to many of the askers on our site who are looking for professional guidance from humans with experience in Academia, as they confront conflicts that will determine the course of their careers. | |
| Jun 5, 2023 at 14:47 | comment | added | Zoe - Save the data dump | @E_net4isonstrike Minor clarification: 11% was correct network-wide at the time, but not on SO in particular. Note how the CEO fails to call it the Stack Exchange network, but says "Stack Overflow network" (not just Stack Overflow). If it's specifically SO, then yes, 11% is a blatant lie. SO (the site) has had >50% sign the open letter since at least yesterday. However, this morning, 11% network-wide would've been approximately correct. It's currently at 14%+ | |
| Jun 5, 2023 at 14:47 | history | edited | Thomas Owens | CC BY-SA 4.0 |
added 2042 characters in body
|
| Jun 5, 2023 at 14:40 | comment | added | Silvio Mayolo | An unfortunate comment but an unsurprising one. A company's natural reaction to a strike is to discount the threat. But there's a reason we're actually doing this and not just talking about it: Because the longer we hold out, the more they have to accept that those of us making a point here are a vital and difficult-to-replace part of the ecosystem. | |
| Jun 5, 2023 at 14:40 | comment | added | E_net4 | @ThomasOwens Sure, I wasn't implying that you stood by this percentage, just added some more context. By the way, network-wide it has already reached 14%. | |
| Jun 5, 2023 at 14:38 | comment | added | Thomas Owens | @E_net4isonstrike I didn't make a claim about 11% being accurate or not. Even if it's a small fraction of total moderators, it's the most active moderators on the most active sites. It's also some of the most active curators on the most active sites. It's going to be a huge strain to have these people not participating. | |
| Jun 5, 2023 at 14:34 | comment | added | David Roberts | "Usage of these tools correlated to a dramatic upswing in suspensions of users with little or no prior content contributions; " -- because both correlate to the rise of people posting ChatGPT spam. It's not rocket science. And correlation is not causation, obviously. | |
| Jun 5, 2023 at 14:34 | comment | added | E_net4 | Zoe (who was interviewed by Dev Class) claims that the 11% was accurate at the time, but the percentage on Stack Overflow in specific is much higher now: 15/27 (according to the list of moderators on Stack Overflow) makes 55% of Stack Overflow elected moderators on strike. | |
| Jun 5, 2023 at 14:33 | comment | added | Lamak | also, the CEO here is using private communications with the mods, that's just so completely wrong that I'm kind of amazed at it | |
| Jun 5, 2023 at 14:31 | comment | added | Andreas moved to Codidact | "These unnecessary suspensions and their outsize impact on new users run counter to our mission and have a negative impact on our community." This is simply untrue. We are the community, and we are sending SE a message. They don't get to put words in our mouths. Accept the words we're giving, instead. | |
| Jun 5, 2023 at 14:30 | comment | added | Andreas moved to Codidact | This is not just "not a good start". It's the opposite of what we're requiring. This only reinforces the strike. | |
| Jun 5, 2023 at 14:29 | history | answered | Thomas Owens | CC BY-SA 4.0 |