RationalWiki talk:All things in moderation/Archive71

From RationalWiki
Jump to navigation Jump to search

This is an archive page, last updated 2 January 2024. Please do not make edits to this page.
Archives for this talk page: , (new)(back)

AI-generated text and images

I think think there should be a policy for use of AI-generated text and images on RationalWiki. I think that use on RW should be very restrictive. First, there are still open questions due to possible violation of legal rights, as noted here: ChatGPT#Legal issues. The second problem is with that of deception, AI-generated text and images can deceive that what it produces is factual. At this point, AI is well known to be a generator of truthiness but not truth. An example of what I would consider to be OK for inclusion on RW would be something that is used as a clear example of what generative AI does such as here: ChatGPT#Sample prompt. An example of what I would consider to be not OK for inclusion would be this: File:Murraycrossburning.png (@Pinkerite). The problems with this image are that it purports to be about a living person (Draft:Charles Murray) in an actual situation. The problem with this image is that even though Murray apparently did admit to burning a cross, it's unlikely that he did so in a scenario such as this. This image could conceivably be considered libelous since Murray expressed remorse for the admitted incident, and he could argue that this image implies that he did it a second time, which he did not claim that he did. More generally, the problem with such images are that they mislead the viewer, something that we should strive against. Bongolian (talk) 23:24, 17 September 2023 (UTC)

If I photoshop Ken Ham into a Flying Spaghetti Monster conference, or AI generate such a photo, that would be recognized as such obvious satire/parody that it couldn't be libelous. In this case there is no reason to keep the photo, I don't know why it was inserted in the first place. There is no obvious satirical or truth value to it. Chillpilled (talk) 00:01, 18 September 2023 (UTC)
I've deleted that image for the reasons that Bongolian and Chillpilled described, with which I wholeheartedly agree. Spud (talk) 00:14, 18 September 2023 (UTC)
I think we should be able to use Photoshop and AI generators for obvious joke images. Like I put a Mission Impossible joke into Putin's article some time ago. Otherwise, I agree with you. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 00:53, 18 September 2023 (UTC)
Wikipedia has essentially banned AI-generated images base on the above-mentioned legal issues:[1]
“”With new processes like Stable Diffusion that can generate high-quality art from a large database of art, the question of copyright remains open, and at the present time, such images should not be used on Wikipedia unless there is assurance that all images used in the AI system are from free or public domain sources.
Bongolian (talk) 05:06, 18 September 2023 (UTC)
Concur with this policy suggestion. There'll be legal battles fought over AI generated content. We don't need RW to be one of those battlegrounds for one. For second, the truthiness element makes it unusable for wiki content. Highly specific carveouts (ie. For the chatgpt examples section) can be considered on a "per-case" basis. -- Techpriest (talk) 07:35, 18 September 2023 (UTC)
I occasionally use ChatGPT here just to reword a sentence or two for smoothness and grammar (sometimes it's difficult to think of how best to do it). Basically just using it as a copyeditor; not routinely to random pages, but just when I hit a hard spot when writing my own new lines, to figure the best way to word something I couldn't think of. There should be no copyright issue here since I'm feeding it a unique (awkwardly written) sentence in the first place to work with. It might be wise to wait and see how these lawsuits turn out, true. There's no "there will be lawsuits" about it, these are already ongoing against Stable Diffusion and ChatGPT. I think they're going to end up showing that AI-generated content at least usually is public domain legally. But there are some facts of the cases which make me think it could go another way. So in the meantime the site shouldn't use AI generated content unless you're ready to defend it as fair use or whatnot regardless of its AI-generated status. Chillpilled (talk) 11:28, 18 September 2023 (UTC)
I strongly agree with Bongolian and Tech. While I probably one of the most optmistic users when it comes to the usage of this sort of novel technology they aren't really reliable yet, and the quality of their content does not meet our Standards IMO. There's nothing wrong about asking ChatGPT for sources and even for summaries of concepts and ideas you're working on, but truth is, you can't rely on them yet. GeeJayKWhere all evil dwells Where every lie is true 12:49, 18 September 2023 (UTC)
I agree on images; there could possibly be individual exceptions for images where fair use can apply and the image is genuinely important to the topic (e.g. it's among what the article describes) or adds great value, but as a general rule, no. On text, Wikipedia's proposed policyWikipedia seems pretty sensible. I've shortened it down to the below, which I agree with. I propose something similar to the below for RW:
  1. Don't publish original content from LLMs, even if heavily edited (except in special circumstances such as a quote from or description of LLMs).
  2. It's fine to use LLMs for copyediting, summarizing, and paraphrasing, with awareness of the risk that they may mangle something. They can also be used to suggest an outline or starting point, but it should not be kept in the edits made.
  3. Editors are fully responsible for any flaws and issues in edits they make using LLM text, and should only use it with due diligence and when they know enough about the topic to avoid messing up.
  4. Use of LLMs must be mentioned if it amounts to more than a little wording suggestion. Usually it should be mentioned in the edit summary.
  5. It's strictly forbidden to present LLM text as your own on talk pages and in edit summaries. RationalWiki is not a place for experimenting with LLM impersonation or use in general.
--ApooftGnegiol (talk) 15:36, 18 September 2023 (UTC)
That sounds good. Spud (talk) 16:35, 18 September 2023 (UTC)
It's OK by me. Bongolian (talk) 02:17, 20 September 2023 (UTC)
New tech is always difficult to regulate; we'll have to update this in the future, but I'll tentatively agree. CorruptUser 18:02, 20 September 2023 (UTC)

Vote on adding ApooftGnegiol's suggestion to CS (PLEASE READ)

As this is voting for a change to CS, a 2/3 supermajority is needed for anything to pass. The vote will be for the general regulations that are suggested, exact details can be worked out in the future if need be. If you have any suggestions for changes or additions to the proposal, feel free to bring them up. Plutocow (talk) 18:41, 20 September 2023 (UTC)

I think it's worthwhile to look at what WP decided on this topic. Personally, I think it's roughly the same rules we should have. KarmaPolice (talk) 15:31, 27 September 2023 (UTC)
This vote has been closed. Please do not add, remove or change votes.
The result of this vote was: All passed except point #2 (which may make #3 and #4 moot); whether the negation of #2 is to be adopted is discussed below this vote.

Point 1

"Don't publish original content from LLMs, even if heavily edited (except in special circumstances such as a quote from or description of LLMs)."

Yay
  1. Plutocow (talk) 18:41, 20 September 2023 (UTC)
  2. RW does not have the 'No original research' clause WP enjoys, so we do need protection from it's fairly woeful products. KarmaPolice (talk) 19:25, 20 September 2023 (UTC)
  3. Bongolian (talk) 19:52, 20 September 2023 (UTC)
  4. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 19:55, 20 September 2023 (UTC)
  5. Yes, of course. Spud (talk) 23:27, 20 September 2023 (UTC)
  6. Chillpilled (talk) 23:46, 20 September 2023 (UTC)
  7. --RWRW (talk) 13:13, 21 September 2023 (UTC)
  8. --Gang O' Shadow Wizards 🧙‍♂️Bore me 15:59, 21 September 2023 (UTC)
  9. —cosmikdebris talk stalk 16:15, 21 September 2023 (UTC)
  10. -- Techpriest (talk) 21:34, 21 September 2023 (UTC)
  11. 𝒮𝑒𝓇𝑒𝓃𝑒 talk 13:26, 22 September 2023 (UTC)
  12. Moon Sock (talk) 14:25, 22 September 2023 (UTC)
  13. As long as there is no significant increase in LLM content reliability. IluzasipalStone them! 14:21, 23 September 2023 (UTC)
  14. Monochroma (talk) 17:30, 23 September 2023 (UTC)
  15. GeeJayKWhere all evil dwells Where every lie is true 18:27, 25 September 2023 (UTC)
  16. Leaving aside the inaccuracies of LLMs, there is a risk of inadvertent plagiarism. MortgageBalls (talk) 18:51, 26 September 2023 (UTC)
  17. Bandwagoning! CorruptUser 07:40, 27 September 2023 (UTC)
  18. ←§ Reichtangle (talk) 16:05, 27 September 2023 (UTC)
Nay
Goat
  1. Before I vote on this one, can I get a clarification as to what "original content" means? (っ◔◡◔)っ ♥ Natsuki Marx ♥ (talk) 13:19, 21 September 2023 (UTC)
    I would assume it basically means 'LLM produced content is to be treated as untrustworthy unless proven otherwise by a human editor. Any 'original research' produced by LLMs carries no weight and are not to be included in RW mainspace articles unless it serves a particular purpose (such as an example of LLM output).' Now, LLMs are not really producing original research yet, but I suppose it's simple good sense to nail this down before they do. KarmaPolice (talk) 20:34, 21 September 2023 (UTC)
    On further consideration, this point could be read to mean 'LLMs are not to write final/direct mainspace text' [ie the words finally seen in RW articles]. I think it's important to clarify this [and let all above shift their votes if needed] before continuing. KarmaPolice (talk) 17:40, 22 September 2023 (UTC)
    I meant (and a clearer wording may be needed), whatever an LLM cooks up when given instructions to make something rather than to edit something fed into it. It's probably not possible to technically draw a completely clear-cut line between creating vs. editing a text, but I think it works well enough intuitively. So basically, the point is to forbid the former (creating). Then item #2 below would allow the latter (editing). --ApooftGnegiol (talk) 18:29, 22 September 2023 (UTC)
    A kinda academic distinction. I mean, you could technically get around this by getting LLMs to make massive 'edits' to a stub which makes it a large article. This is why I would widen 'creating' to include edits to existing mainspace too. #2 appears [at least to me] to be much more about the 'backroom' use of LLMs to aid the human editor producing mainspace content quicker. KarmaPolice (talk) 18:57, 22 September 2023 (UTC)
    I guess the more legal way to look at it would be authorship and derivative works. Are you still the (main) author? There is a distinction there which is commonly used, but IANAL and don't write it the best way. (Prompting "make me ..." to a generative AI doesn't count as the requester being an author for copyright purposes, but if what comes back is substantially something you're the author of and put in, then you still have copyright.) But that's also saying that I mentally link this very much to the plagiarism question. I'm not sure how practical this is, but maybe there's a way to make that come through clearer if you want to go with it. --ApooftGnegiol (talk) 19:10, 22 September 2023 (UTC)
    Ahh, so by "creating" or "original LLM content" I meant anything where the LLM is or becomes the "author" (also considered plagiarism here), and by "editing" I mean anything where the author remains the human editor and/or other RW writers (their article text being used). If you water down the authorship of an article text too much by repeatedly changing it, then it becomes, well, whatever it is then, and then it should be judged based on that, and not on exactly how it got that way. Probably, after these ideas are voted on, most of this needs to be rewritten to reflect whatever consensus we end up with on exactly what it should mean. --ApooftGnegiol (talk) 19:34, 22 September 2023 (UTC)
    I would personally term the LLM 'the author' if it's 'their' own words being put into the mainspace page - be it as a new article or an edit of an existing one. For even if LLMs are making minor edits, there's the risk over time the whole thing will become primarily LLM-written [which I think the consensus here is would be a Bad Thing]. Which if nothing else, I do think we need to avoid that situation due to the legal grey areas re: copyright etc. Plus, while I kinda agree with your position, I get the strong feeling the consensus here won't. Plus, we have no realistic method of policing an new article/edit of existing line. KarmaPolice (talk) 19:48, 22 September 2023 (UTC)
    It doesn't matter in my view either whether a page is newly made or already exists when LLM output is inserted, that's separate from the uses of the words create and edit I'm concerned with: the LLM tweaking text fed into it via the prompt ("editing") vs. making new text up ("creating"). The previous idea I had was in #2 below, to mention ways of having the LLM edit text from a prompt (the first sentence of the two sentences). Everything that does not contradict that could be forbidden and I'd say my idea is represented well enough. Of course, such actions as in #2 could still be done repeatedly, eventually resulting in that copyright gray area -- through much rephrasing, wording adjustments, shortening or lengthening, etc. --ApooftGnegiol (talk) 20:09, 22 September 2023 (UTC)
    Ah, I think I get the issue now. You're drawing the line between creating and editing content, while I thought you were talking about creating and editing articles. Personally, I would be okay with supervised and proofread by human LLM content editing [in fact, I would say this type of content falls more under 'Type #2 use of LLMs']. KarmaPolice (talk) 20:14, 22 September 2023 (UTC)
Ineligible
  1. Ψ (talk) 06:07, 21 September 2023 (UTC) Not here for long enough. Plutocow (talk) 06:41, 21 September 2023 (UTC)

Point 2

"It's fine to use LLMs for copyediting, summarizing, and paraphrasing, with awareness of the risk that they may mangle something. They can also be used to suggest an outline or starting point, but it should not be kept in the edits made."

Yay
  1. Plutocow (talk) 18:41, 20 September 2023 (UTC)
  2. Seems reasonable enough, suppose. KarmaPolice (talk) 19:26, 20 September 2023 (UTC)
  3. Yes, of course. Spud (talk) 23:27, 20 September 2023 (UTC)
  4. Chillpilled (talk) 23:46, 20 September 2023 (UTC)
  5. (っ◔◡◔)っ ♥ Natsuki Marx ♥ (talk) 13:19, 21 September 2023 (UTC)
  6. —cosmikdebris talk stalk 16:15, 21 September 2023 (UTC)
  7. IluzasipalStone them! 14:21, 23 September 2023 (UTC)
Nay
  1. Too risky. -- Techpriest (talk) 21:34, 21 September 2023 (UTC)
  2. Yeah people shouldn’t use these bots at all. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 06:52, 22 September 2023 (UTC)
    Yet how would you know they did? Does RW have any ability to detect use? In cases like this, I'd argue that simply 'banning' it will push it underground, both 'responsible' and 'irresponsible' users. KarmaPolice (talk) 09:44, 22 September 2023 (UTC)
    I just don’t want the CS to endorse the use of LLMs in any way. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 00:25, 26 September 2023 (UTC)
    I don't think 'trying to regulate' is 'endorsing', Duce. LLMs are here, whether we like it or not and they're not going away. RW can very well adopt a 'discouraged but permitted' tone regarding LLM usage, along with a coherent precis at the top of why we've reached this position. LLMs are a resource with potential as well as a risk, like any tech innovation. Lastly, if we go full-in for a total ban I fear we shall look increasingly outdated. What happens if search engines start using LLMs - will we have to boycott them for RW usage? Wait, how the hell did people do online research before a search engine? KarmaPolice (talk) 15:25, 27 September 2023 (UTC)
  3. Shouldn't be used. 𝒮𝑒𝓇𝑒𝓃𝑒 talk 13:26, 22 September 2023 (UTC)
  4. Moon Sock (talk) 14:27, 22 September 2023 (UTC)
  5. Looks like a back door.Ariel31459 (talk) 15:21, 25 September 2023 (UTC)
  6. Bongolian (talk) 19:52, 20 September 2023 (UTC) Changing my vote, I'm not thrilled with the idea. Bongolian (talk) 18:01, 25 September 2023 (UTC)
  7. --Gang O' Shadow Wizards 🧙‍♂️Bore me 18:10, 25 September 2023 (UTC)~
  8. Don't like this one. GeeJayKWhere all evil dwells Where every lie is true 18:27, 25 September 2023 (UTC)
  9. absolutely not ShabiDOO 09:56, 26 September 2023 (UTC)
  10. Far too far over the line of not being individuals' original work. Makes LLMs sound like glorified grammar checkers. MortgageBalls (talk) 18:51, 26 September 2023 (UTC)
    I suspect you can use them for that, Balls. It's not an either/or for LLMs, and their skill-sets will increase over time. Hell, I think an LLM would have better grammar than I do now, so using it for that would be an improvement for all. KarmaPolice (talk) 15:25, 27 September 2023 (UTC)
Goat
  • If this ends up used, I suggest adding a little reminder about plagiarism (inspired by a comment from KarmaPolice below) to the end of it:
"Text that you process using an LLM and then substantially keep must be original before using the LLM. If it would be plagiarism to use a text without LLM processing, then using it with LLM processing is also plagiarism (as is using content created rather than edited by LLMs)." --ApooftGnegiol (talk) 13:44, 22 September 2023 (UTC)
  • Or, if this part ends up rejected, then maybe item #1 above should be changed to a stricter forbidding of everything from LLMs. And of course, #3 and #4 below seem to become moot if this item is rejected, so would be abandoned then. But #5 would remain relevant. --ApooftGnegiol (talk) 18:35, 22 September 2023 (UTC)
  • Nay seems like it'll win (vote closing tomorrow). I thought of what nay to this point #2 means, and it looks like most people who say anything more about what they want really want all LLM use forbidden. As opposed to simply not wanting the rules to endorse any LLM use. If this is wrong, then we could have another vote on what to do next, whether we should just drop #2 or whether we should take this outcome to also mean that the rest should be simplified into a stricter no LLM use rule. But if people don't care for such an additional vote, then I think the above comments along with the nay votes are clear enough. --ApooftGnegiol (talk) 13:29, 26 September 2023 (UTC)
    Changes to policy (the "yes" vote here) require a 2/3rds majority to pass (RationalWiki:Community Standards#Common guidelines), so this point now is highly unlikely to win. Bongolian (talk) 02:09, 27 September 2023 (UTC)
  • While I'm on they Yay side, there should be restrictions on input text sources. I think it would be fine to allow an AI to condense a Wikipedia article down for RationalWiki's purpose. But not just any source should be allowed. The source should always be specified in the summary. --LiamM32 (talk) 23:43, 12 October 2023 (UTC)
Ineligible
  1. No. I don't think we should LLM in any circumstances. This practically defeats the purpose of RationalWiki editors if everything can be AI generated. Ψ (talk) 06:09, 21 September 2023 (UTC)

Point 3

"Editors are fully responsible for any flaws and issues in edits they make using LLM text, and should only use it with due diligence and when they know enough about the topic to avoid messing up."

Yay
  1. Plutocow (talk) 18:41, 20 September 2023 (UTC)
  2. Bongolian (talk) 19:52, 20 September 2023 (UTC)
  3. Yes, of course. Spud (talk) 23:27, 20 September 2023 (UTC)
  4. (っ◔◡◔)っ ♥ Natsuki Marx ♥ (talk) 13:19, 21 September 2023 (UTC)
  5. —cosmikdebris talk stalk 16:15, 21 September 2023 (UTC)
  6. Assuming point 2 passes, this is a requirement. -- Techpriest (talk) 21:34, 21 September 2023 (UTC)
  7. 𝒮𝑒𝓇𝑒𝓃𝑒 talk 13:26, 22 September 2023 (UTC)
  8. Moon Sock (talk) 14:28, 22 September 2023 (UTC)
  9. IluzasipalStone them! 14:21, 23 September 2023 (UTC)
  10. --Gang O' Shadow Wizards 🧙‍♂️Bore me 18:11, 25 September 2023 (UTC)
  11. GeeJayKWhere all evil dwells Where every lie is true 18:27, 25 September 2023 (UTC)
  12. Use at own risk... CorruptUser 07:40, 27 September 2023 (UTC)
  13. ←§ Reichtangle (talk) 16:17, 27 September 2023 (UTC)
Nay
  1. Editors are responsible for what they submit into RationalWiki anyway - and given that I don't think LLMs should be allowed at all, this feels redundant to add as well. MortgageBalls (talk) 18:51, 26 September 2023 (UTC)
Goat
  1. Lean to 'Yay', but officially Goat because I simply see it as a mere extension of the basic 'due diligence' rule in finding decent sources to base articles on etc. These things are simply tools to cut some of the slog in producing a basic draft, not something to cover for your lack of knowledge on a topic. KarmaPolice (talk) 19:29, 20 September 2023 (UTC)
    I mostly thought of this part as a reminder or note of caution, rather than additional restriction. Seemed good to include in a policy text on this, but technically it does little if anything. --ApooftGnegiol (talk) 23:20, 20 September 2023 (UTC)
    I just put that in because I believe a core principle on LLMs is that they are mere tools, and thus all responsibility for their fails etc is down to the editor not being conscientious enough. KarmaPolice (talk) 11:35, 21 September 2023 (UTC)
    In my experience there's a lot of overeager LLM fanboys who need this reminder explicitly. -- Techpriest (talk) 21:34, 21 September 2023 (UTC)
Ineligible
  1. If we have to use LLM, this point should be placed in.Ψ (talk) 06:10, 21 September 2023 (UTC)

Point 4

"Use of LLMs must be mentioned if it amounts to more than a little wording suggestion. Usually it should be mentioned in the edit summary."

Yay
  1. Plutocow (talk) 18:41, 20 September 2023 (UTC)
  2. Bongolian (talk) 19:52, 20 September 2023 (UTC)
  3. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 19:57, 20 September 2023 (UTC)
  4. That seems fair. Spud (talk) 23:27, 20 September 2023 (UTC)
  5. Chillpilled (talk) 23:46, 20 September 2023 (UTC)
  6. (っ◔◡◔)っ ♥ Natsuki Marx ♥ (talk) 13:19, 21 September 2023 (UTC)
  7. —cosmikdebris talk stalk 16:16, 21 September 2023 (UTC)
  8. Predicated on point 2 passing of course. -- Techpriest (talk) 21:34, 21 September 2023 (UTC)
  9. 𝒮𝑒𝓇𝑒𝓃𝑒 talk 13:26, 22 September 2023 (UTC)
  10. IluzasipalStone them! 14:21, 23 September 2023 (UTC)
  11. --Gang O' Shadow Wizards 🧙‍♂️Bore me 18:11, 25 September 2023 (UTC)~
  12. GeeJayKWhere all evil dwells Where every lie is true 18:27, 25 September 2023 (UTC)
  13. No lazybones!CorruptUser 07:41, 27 September 2023 (UTC)
Nay
Goat
  1. Yet... I'm not honour-bound to confess that [for example] I built a RW article using lots of bones swiped from it's WP equivalent, am I? [Or really, am I?] KarmaPolice (talk) 19:32, 20 September 2023 (UTC)
    These rules are not retroactive, but it would be honest of you at this point to state what the article was. Bongolian (talk) 19:52, 20 September 2023 (UTC)
    I'm trying to see if we already have precedent. As for 'fessing up, I don't do plagiarism - I only do articles which I already know more than the person who'd done the existing page and I don't like WP's tone/style at the best of times. But most of the RW articles I've done it can be said WP is an 'uncredited source'. But then again, much of the Kemal Atatürk article was based off 'Atatürk' book by Andrew Mango and I didn't credit that either. I'm floating this because while I've not used an LLM yet, I can forsee me asking one to produce a 'RW-style draft on [topic]' then whacking it into shape for the final edition [and thus, a distinct possibility that some direct LLM chunks might survive]. Basically, pointing out the relative contradiction between Points 2 and 4. KarmaPolice (talk) 12:27, 21 September 2023 (UTC)
    I'm not sure exactly what you think of, but if you can come up with some clear and short criteria, maybe a wording change suggestion for #2 and/or #4 could help. Or, you're simply saying it's not so worthwhile to label when some LLM snippets end up in an article? --ApooftGnegiol (talk) 13:02, 21 September 2023 (UTC)
    Unless we develop a general principle/system of crediting stuff not directly cited on pages but clearly 'had influence' in the creation [like the Mango book above] I am not sure how this would work. I would also argue that ultimately, does it really matter if LLMs end up having considerable input into RW articles - my beef would be if the content was incorrect, poor quality and/or not proof-read by a human first, not the origin/'methods of creation'. Plus, LLMs are going to be for some time a mere tool - I believe that any article produced majoritly by one would quickly flagged up by an editor, mod or simple user and 'dealt with'. However, there is one aspect not considered yet; that of copyright/IP in regards to LLM-generated content. That's going to be a grey area for a long time yet, and thus for that reason alone I would ban all LLM content from RW mainspaces. Basically, we extend the 'plagiarism' bit in the Community Standards to cover LLM generated content. So I would merge 2/4 into a single one which reads;
    'LLMs can be used at any stage of page creation (including drafting, copyediting, paraphrasing, proofreading and outlining) with the exception of the final product. Verbatim use (or token paraphrasing) of LLM produced content in mainspace articles is to be considered plagiarism, and should be kept to a minimum.' KarmaPolice (talk) 13:48, 21 September 2023 (UTC)
    Maybe you interpret "Use of" at the start of this point differently than I intended? I didn't mean it like "inspired by", but "including text processed by". A smaller change to clarify that, if so (and it may be good anyway), could be (bold part edited) "Inclusion of text processed by LLMs must be mentioned if it amounts to more than a little wording suggestion. Usually it should be mentioned in the edit summary." If there's no objections, I think that may be a better wording to use anyway.
    As for your proposed replacement for #2 and #4, I was a little confused by it. Your "with the exception of the final product" could be read as requiring throwing away and replacing all word sequences that came through an LLM (including from copyediting) before saving the page, as it's not clear how the final product is meant to differ from earlier work. Forbidding all use could be read into the later part too, but maybe you meant "original LLM produced content" (italic part added)? So it mentions things first as if allowing them, but then seems to forbid it all. --ApooftGnegiol (talk) 13:14, 22 September 2023 (UTC)
    On consideration, it all depends on how #1 is read - I've put my thought/issue in that one's Goat. #2 can be said to cover all the 'indirect' LLM usage - #4 is simply about attribution, right? Therefore, #4 can be said to be two questions; whether LLMs should be attributed if directly contributing [ie actual content ends up in mainspace] and when they are indirectly contributing [ie providing the services noted in #2]. KarmaPolice (talk) 17:48, 22 September 2023 (UTC)
  2. I do not think this makes sense in the context of LLMs being prohibited more generally. Anyone who tries to get around that prohibition is obviously not going to declare it, but anyone who adheres to it has nothing to declare. MortgageBalls (talk) 18:51, 26 September 2023 (UTC)
    Yes, as RW has no real ability to detect/enforce this [at least to my knowledge, at least, and at this time] it will basically be an honour system. In this case, I think if you're an editor who uses LLMs to do some of the 'back office' functions [if permitted] for RW mainspace you should declare you do on your bio page. KarmaPolice (talk) 15:27, 27 September 2023 (UTC)
Ineligible
  1. Correct. Ψ (talk) 06:11, 21 September 2023 (UTC)

Point 5

"It's strictly forbidden to present LLM text as your own on talk pages and in edit summaries. RationalWiki is not a place for experimenting with LLM impersonation or use in general."

Yay
  1. Plutocow (talk) 18:41, 20 September 2023 (UTC)
  2. Simple extension of the plagiarism rule, suppose. KarmaPolice (talk) 19:33, 20 September 2023 (UTC)
    There's another side to it too, that users can't chatbot their way through discussions (should someone try). Imagine some account posting signed comments on various topics, but then one of them happens to contain "As a large language model, I...". Wikipedia seems to care about risks of disruptive behavior and I thought why not include it. --ApooftGnegiol (talk) 23:20, 20 September 2023 (UTC)
  3. Bongolian (talk) 19:52, 20 September 2023 (UTC)
  4. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 19:57, 20 September 2023 (UTC)
  5. Definitely. Spud (talk) 23:27, 20 September 2023 (UTC)
  6. Chillpilled (talk) 23:46, 20 September 2023 (UTC)
  7. --RWRW (talk) 13:14, 21 September 2023 (UTC)
  8. (っ◔◡◔)っ ♥ Natsuki Marx ♥ (talk) 13:19, 21 September 2023 (UTC)
  9. —cosmikdebris talk stalk 16:16, 21 September 2023 (UTC)
  10. Users should put things in their own words. An easy concur. -- Techpriest (talk) 21:34, 21 September 2023 (UTC)
  11. 𝒮𝑒𝓇𝑒𝓃𝑒 talk 13:26, 22 September 2023 (UTC)
  12. IluzasipalStone them! 14:21, 23 September 2023 (UTC)
  13. --Gang O' Shadow Wizards 🧙‍♂️Bore me 18:12, 25 September 2023 (UTC)
  14. GeeJayKWhere all evil dwells Where every lie is true 18:27, 25 September 2023 (UTC)
  15. MortgageBalls (talk) 18:51, 26 September 2023 (UTC)
  16. Agreed CorruptUser 07:41, 27 September 2023 (UTC)
  17. ←§ Reichtangle (talk) 16:18, 27 September 2023 (UTC)
Nay
Goat


How to put this together? Full restriction or some leeway with LLMs?

So all passed except point #2 above. Various comments on #2 suggested people want its negation, not just its absence. But if so, then points #3 and #4 become completely pointless and #1 needs an edit to make clear that all LLM processing of content is banned. Looking at how almost everyone voted for #3 and #4 without comment, I think at minimum a discussion on this is now needed.

I'll put together two draft versions (wording can be further tweaked, I included some clarifications below) and ask which of them people want, for a start. If this is good enough as a choice to consider, that is, otherwise you can suggest something else. If you want a further vote and we go with just these two options, maybe a simple majority vote between the two could be used (it's already been voted with 2/3 majority that the CS should change, the question now is just exactly how). --ApooftGnegiol (talk) 18:17, 27 September 2023 (UTC)

Negating point 2

Don't publish content created or edited by LLMs, even if heavily hand-edited (except in special circumstances such as a quote from an LLM, or as part of describing it).

It's strictly forbidden to present LLM text as your own on talk pages and in edit summaries. RationalWiki is not a place for experimenting with LLM impersonation or use in general.

This option is for those who want a full ban on LLM use for making RW content. --ApooftGnegiol (talk) 18:17, 27 September 2023 (UTC)

Omitting point 2

Don't publish content created using LLMs, even if heavily hand-edited (except in special circumstances such as a quote from an LLM, or as part of describing it).

Editors are fully responsible for any flaws and issues in edits they make where LLMs have processed a text, and such should only be done with due diligence and when they know enough about the topic to avoid messing up.

Inclusion of text processed by LLMs must be mentioned if it amounts to more than a little wording suggestion. Usually it should be mentioned in the edit summary.

It's strictly forbidden to present LLM text as your own on talk pages and in edit summaries. RationalWiki is not a place for experimenting with LLM impersonation or use in general.

This option is for those who don't want LLM use for copyediting outright banned, just not endorsed. --ApooftGnegiol (talk) 18:17, 27 September 2023 (UTC)

Discussion

So, to just omit point #2, or to go further and negate it? Is there a clear consensus? Should we vote on it? --ApooftGnegiol (talk) 18:17, 27 September 2023 (UTC)

Neither. I say, we go back, chat about folks views about the particular issue, then try to re-write #2 in manner we can hope to get a supermajority in a re-run. You can't do the first suggestion because it's only at ~59% 'Ban' and the second one simply dodges the question. KarmaPolice (talk) 20:06, 27 September 2023 (UTC)
I think you're right, another vote would be needed to go with the full ban. And it's probably better to consider more ideas.
I'm largely out of ideas right now, I leave it to others to propose something more to fill the gap or change the approach. Also, while on the theme of generative AI it's worth having a vote on some approach to generative AI images too, not only text, but then suggestions are needed.
I only had one more thing, an idea for mentioning plagiarism if we were to allow some processing of text via a prompt to an LLM. "Text that you process using an LLM and then substantially keep must be original before using the LLM. If it would be plagiarism to use a text without LLM processing, then using it with LLM processing is also plagiarism (as is using content created rather than edited by LLMs)." --ApooftGnegiol (talk) 21:15, 28 September 2023 (UTC)
Yes to the additional plagiarism note.
How about an outright ban on all AI generated images involving living people so that they can't make legal threats? (The Jesus flipping over tables image is fun and he can't sue.) Spud (talk) 10:36, 30 September 2023 (UTC)
Strictly speaking, the vote was about LLMs, which cover text- but not image-generators such as DALL-E. I think the term 'generative AI' is more encompassing of both text and images. We had gotten to this vote from objections to inclusion of an AI-generated image. Bongolian (talk) 17:44, 30 September 2023 (UTC)
Okay, my idea is that we put together a sample of say, a dozen example uses of LLMs which is not them directly producing mainspace articles [which is already banned as per #1]. At least one of these can be regarding them producing imagery. Then we vote again for each example Yay/Nay/Goat so we can more accurately measure the 'state of mob opinion' and re-cut #2 to fit it. I suggest we leave a bit at the top of each vote so we can discuss the example etc. KarmaPolice (talk) 13:37, 4 October 2023 (UTC)
Sounds good to me. Spud (talk) 14:02, 4 October 2023 (UTC)
Yes. An example I think should be acceptable is to use an AI to condense a Wikipedia article down to only things relevant to RationalWiki. This would be a good way to get drafts started for subjects already covered on Wikipedia. They can then be manually edited to make them more suitable for our purpose. Any manual edits to long AI-condensed text (such as Wikipedia articles) should happen in a separate edit, though I don't think this is necessary if one condenses their own writing. I will add Condensing to the list below. --LiamM32 (talk) 23:10, 14 October 2023 (UTC)

Okay. I've taken on-board Bongolian's point and will say in this case, image-generators shall be banned under Rule 1. Below are the six 'main' ways I can see them being used in 'back office' manner for RW mainspace, in the order as they came into my head. If anyone thinks of a new category, feel free to add them. If you have any thoughts/complaints/questions, shove them under the appropiate group. KarmaPolice (talk) 23:34, 12 October 2023 (UTC)

A: Outlining; 'Generative AI produces a plan on how a RW article on [subject] is laid out, what topics need mentioning and so on for a human editor to use as a guide.'
B: Researching; 'Generative AI gathers the raw materials for a RW article on [subject] (such as sources) which are then verified and (perhaps) used by a human editor.'
C: Drafting; 'Generative AI produces an article on [subject] in the style of RW, which is then gets a heavy re-write by a human editor before use.'
D: Copyediting; 'Generative AI reviews existing RW human-made content, which then provides suggestions for improvement which is then performed by a human editor.'
E: Proofreading; 'Generative AI reviews human-made RW content before publishing for such things as word usage, grammar, sentence structure, working links and so on.'
F: Translating; 'Generative AI converts raw material into more usable forms by human editor. Most commonly language, but could also be speech-to-text or data to graphs.'
G: Condensing; 'Generative AI condenses text from an acceptable source without copyright protection; removes unnecessary parts without adding more. An example would be to condense a Wikipedia article to remove anything irrelevant to RationalWiki.'

AI Images

I know I'm late to this party, but I really feel the need to point out some horrifying things that people on Reddit have found. This is not only disgusting imagery, it also unfairly drags the names of animation studios with the good sense to avoid right-wing bullshit, like Disney and Pixar, through the mud. --Luigifan18 (talk) 15:55, 12 October 2023 (UTC)

The Mod-Gods deleted it. What was it about, exactly? KarmaPolice (talk) 23:01, 12 October 2023 (UTC)
Basically, fake movie posters glorifying the Holocaust. --Luigifan18 (talk) 00:23, 13 October 2023 (UTC)

Second vote on AI

To clarify the vote on AI above, I'm calling for a second round of voting:

This vote has been closed. Please do not add, remove or change votes.
The result of this vote was: Point 2: "Negate" passes unanimously (7-0).

Negating point 2 vs. omitting point 2:

Negate

Don't publish content created or edited by LLMs, even if heavily hand-edited (except in special circumstances such as a quote from an LLM, or as part of describing it).

It's strictly forbidden to present LLM text as your own on talk pages and in edit summaries. RationalWiki is not a place for experimenting with LLM impersonation or use in general.

  1. I guess this is my old man yelling at cloud moment. So far, I haven't seen LLMs do anything good for online communities. With bot-generated ads, and even regular posts (this one from my experience at one of the places I used to work at) being so obviously low quality (as in, just using a lot of big words that may or may not be related to the subject) and generated from ChatGPT, I'm not convinced that they bring anything good at all yet. LLMs are also incapable of thinking like a human does, they don't understand what they say on a fundamental level. They just "predict" what words should follow what words. Some of our best articles contain certain humor that these bots are incapable of understanding. Endorsing their use, no matter the purpose, is likely to degrade quality and make us look more like Wikipedia, with the dry tone they use (Wikipedia is likely a big part of their training material). I may not be able to escape ChatGPT at any workplace nowadays, but other than that, I don't want it anywhere near me. Rabbitseatcarrots (talk) 08:33, 13 October 2023 (UTC)
    Quit shaking your fist for a moment. My word processor is 'incapable of thinking like a human does', but that does not make it useless - I compose RW mainspaces in it so I get the best checks for spelling and so on. LLMs are merely tools - it is on the onus of the RW'ian to use them correctly and to choose reliable ones for the desired task. So you cannot really see much use of LLM tools right now for RW? Fair enough; but this tech is moving very fast. My list of six examples above show areas I do think they could become useful for us - perhaps not now, but two-five years from now.
    I will also point out that LLMs have already been banned for directly editing mainspace. The question is now how much 'backoffice' functions LLMs should be allowed to perform at the behest and under the supervision of a human editor. KarmaPolice (talk) 10:46, 13 October 2023 (UTC)
  2. This is my personal preference; again, I think the legal risks are too great for the time being. I'll also admit to some degree of a personal preference here; LLM processed text (for ie. copyediting) tends to get you very "samey" results and it would result in the loss of our RW:SPOV. For smaller articles especially, the snark is a big part of the identity of the site, we're not diet wikipedia. -- Techpriest (talk) 12:26, 15 October 2023 (UTC)
  3. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 12:56, 15 October 2023 (UTC)
  4. -Flandres (talk) 13:27, 15 October 2023 (UTC)
  5. Bongolian (talk) 16:55, 15 October 2023 (UTC)
  6. If LLMs get much better in the future, we can always revise our policy in that future. I don't think it's clear that they will just make rapid progress, though. https://www.tomshardware.com/news/chatgpt-response-quality-decline Vomitorium (talk) 05:36, 22 October 2023 (UTC)
  7. 𝒮𝑒𝓇𝑒𝓃𝑒 talk 21:58, 22 October 2023 (UTC)
Omit

Don't publish content created using LLMs, even if heavily hand-edited (except in special circumstances such as a quote from an LLM, or as part of describing it).

Editors are fully responsible for any flaws and issues in edits they make where LLMs have processed a text, and such should only be done with due diligence and when they know enough about the topic to avoid messing up.

Inclusion of text processed by LLMs must be mentioned if it amounts to more than a little wording suggestion. Usually it should be mentioned in the edit summary.

It's strictly forbidden to present LLM text as your own on talk pages and in edit summaries. RationalWiki is not a place for experimenting with LLM impersonation or use in general.

Goat
  1. I don't think you can simply re-run without really changing the question. KarmaPolice (talk) 10:46, 13 October 2023 (UTC)
  2. This isn't useful, and is gonna drag out the discussion further. Votes on changing the community standards should be explicit, e.g., "Add the following text to the community standards: LLMs should not be used for..." Vote proposals shouldn't be this vague. —cosmikdebris talk stalk 14:34, 13 October 2023 (UTC)
    Which is why I did my 'six back-office uses of LLMs' list above, to better guage where a supermajority line of ban is [just above 'AI Images' sub-title]. Then, and only then can #2 be re-crafted and put up for a vote. KarmaPolice (talk) 15:15, 13 October 2023 (UTC)
  3. I’m generally not a fan of adding rules that ban something that will be difficult to detect and enforce and which will open up a lot of potential “wikilawyering”/cooping based on more or less good faith/credible allegations of improper AI use and/or declarations of AI use. That said, I agree that RW definitely shouldn’t allow the AI equivalent of “churnalism” (whether in text or imagery), but I find it hard to figure out how a ban on back office uses (as per KarmaPolice above) could work in practice. ScepticWombat (talk) 22:43, 22 October 2023 (UTC)


This vote has been closed. Please do not add, remove or change votes.
The result of this vote was: Total ban on AI-generated images: Passes 8-2 (80%).

AI-generated images on RationalWiki:

Total ban
  1. No. Rabbitseatcarrots (talk) 08:42, 13 October 2023 (UTC)
  2. The use of AI art to illustrate pages that don't have illustrations is IMO largely useless. Not every page needs an associated image. The only real use to put an image on every page is because Google likes it, and I don't think this site should be chasing Google's SEO rankings. In addition, any concerns about copyright are IMO elevated tenfold for image generation AIs. See other arguments under Goat. -- Techpriest (talk) 12:26, 15 October 2023 (UTC)
  3. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 12:55, 15 October 2023 (UTC)
  4. -Flandres (talk) 13:28, 15 October 2023 (UTC)
  5. Bongolian (talk) 16:55, 15 October 2023 (UTC)
  6. I really don't trust this technology. GeeJayKWhere all evil dwells Where every lie is true 22:12, 15 October 2023 (UTC)
  7. Don't like 'em Vomitorium (talk) 05:36, 22 October 2023 (UTC)
  8. 𝒮𝑒𝓇𝑒𝓃𝑒 talk 21:58, 22 October 2023 (UTC)
No ban, except for living people
  1. I think that's the best way of avoiding having any potentially libelous content at all. Spud (talk) 11:17, 15 October 2023 (UTC)
  2. AI art could be of help in providing illustrations to articles where no (suitable) images exist or are hard to find. Carthage (talk) 11:57, 15 October 2023 (UTC)
No ban, except for libelous content
Goat
  1. See my comment under "goat" for "Negating point 2 vs. omitting point 2" above. This kind of election booth is just going to keep this debate dragging on and on and on. We should be voting on explicit changes to the community standards and not vague concepts. —cosmikdebris talk stalk 14:36, 13 October 2023 (UTC)
  2. Would it be possible to have something between "no ban, except for living people" and "total ban"? AI art certainly has its uses, and it's great for allowing those who don't have the talent or patience to develop conventional art skills to be creative, but I'd rather not allow us to be made to look like we think the Holocaust was a jolly good time or something terrible like that. (And yes, the file I uploaded as a demonstration is a real Facebook screenshot that a Reddit user took as an example of just how… depraved some people can get using AI tools. I would have linked to the original Reddit post rather than uploading this garbage to RationalWiki, but the post was deleted.) --Luigifan18 (talk) 16:14, 13 October 2023 (UTC)
  3. I simply don't see the point for AI art on our site (besides possibly a confined example to a single article); if you want to experiment with it on your own time, be my guest, but I don't think RW should get in the business of making fabricated images when a part of our goal is to debunk falsehoods. Generating our own images with AI would be very contrary to the sites mission. One of our goals is debunking cranks and opening up room for AI generated images means that we leave room for moonbats (who tend to ideologically align with us on the surface, only to diverge when they begin pushing their particular brand of crankery) to generate fake images in order to push cranky behavior. Personally I would rather see a policy proposal allowing AI-generated images/videos (media?) only in the context of dealing with fake news peddlers, should they ever choose to use them to spread fake news (which already has happened quite a bit). Anyway otherwise, I echo Cosmikdebris; vague votes like this will just result in leaving room to drag the votes on for much longer than they're supposed to. Assuming this ends with a negative vote in favor of the proposer (not saying it will), this will just result in them drafting yet another vote with slightly different wording/arguing over the minutae. -- Techpriest (talk) 12:26, 15 October 2023 (UTC)
  4. RationalWiki should be a wiki that challenges the status quo. We can't do that if we are relying on AIs that precisely enforce the status quo by mimicking the linguistic bias, cultural bias, gender bias, etc. of current writings. As one paper puts it: [2]

    ...the groups capable of building these models may have excessive power and influence, many end users might suffer from any biases present in these models, and it will be difficult to tell if models are safe to use in particular contexts because the models and their training data are so large.

    MayGodSaveUsAll (talk) 12:51, 15 October 2023 (UTC)
  5. I don’t see how AI generated images could possibly be allowed given the general ban on AI/LLM generated original content in point 1 above, so why is there a separate vote for image generation? It seems contradictory to me to ban AI generated original text (especially if even copy/editing is banned as well), but to allow AI generated images. ScepticWombat (talk) 22:32, 22 October 2023 (UTC)


Retroactively striking votes

Sometimes, after a vote has concluded, some votes can be found to be invalid, and in particularly close votes, this can change the outcome. A recent example is the Carthage vote that was removed from the YouTube vote, which isn't too big a deal as it was fixed only a couple of days after the conclusion of the vote. But this is not the only time this has happened: this vote was decided by Dutchbag, who turned out to be a sockpuppet of the banned user Rob Smith. If his vote was struck, it would have changed the outcome of the vote (which is moot, considering the user being voted on is now permanned, but still). There are also cases where User:Judge Dredd and User:Ze voted in the same vote despite being the same person (although I won't dig through the archives to see if any votes were split because of this). My question is, are there any procedures for if a vote is retroactively found to be invalid for one reason or another? Plutocow (talk) 07:21, 21 September 2023 (UTC)

Dredd and Ze weren't the same user. It's just that Dredd was so confusing about the whole affair that half the site is now convinced they're the same person. Having interacted with both, they're very different people. Dredd generally behaved out of aloof spite. Ze was too caught up in being buds with Raven, while Dredd never gave a damn. No real comment on the rest of your suggestion; an outcome on this would probably be helpful. -- Techpriest (talk) 10:38, 21 September 2023 (UTC)

Why here?

Why are we holding these discussions here?

They should be on the community standards talk page and the results should be updated to CS.Bob"Life is short and (insert adjective)" 09:49, 21 September 2023 (UTC)

Nobody reads the CS talk page. Plutocow (talk) 16:55, 21 September 2023 (UTC)
I'm reasonably sure that the current site-wide notice would send them there.Bob"Life is short and (insert adjective)" 20:00, 21 September 2023 (UTC)

Closing out the AI/LLM/Community Standards discussion

@Bongolian @CorruptUser @Flandres @LeftyGreenMario @Spud @GeeJayK

This discussion hasn't seen a post since October 4 and has fizzled out. I propose coming to closure and archiving it. There are some unresolved issues and I propose restarting this, if desired, on the Community Standards talk page.

To close this, I propose the following changes to the Community Standards:

  • In section "Page Maintenance" add a new subsection "AI-generated text and images" between the "Notability" and "Talk pages" subsections.
  • In this new paragraph, add the following text:

Do not publish original content from large language models (LLMs) even if heavily edited, except in special circumstances such as a quote from or description of LLMs.

This is the community-approved "Point 1" from above. Since Point 2 failed, Points 3 and 4 become moot.

I'll leave this open for comment for a day or so before making any changes. —cosmikdebris talk stalk 17:07, 12 October 2023 (UTC)

Looks good to me. But since it was a faked AI generated image which was supposed to depict a living person that sparked this debate in the first place, I still think the use of such images needs to be addressed. Spud (talk) 00:27, 13 October 2023 (UTC)
As much as I don't want to oppose AI art, the recent sight of a Hitler Youth boy skateboarding over a pile of burning Jew skulls has me rather on edge. No, I'm not bloody joking. I wish I was… --Luigifan18 (talk) 00:35, 13 October 2023 (UTC)
I think we need to vote on clarifying the rejection of point 2, as noted above. Also, I think we should have a vote on AI-generated art. It would be useful to keep the above text here until a vote on these two issues is completed. Bongolian (talk) 01:05, 13 October 2023 (UTC)
Since this discussion has come alive again, I'll defer on making any changes to the CS or archiving discussions. —cosmikdebris talk stalk 14:37, 13 October 2023 (UTC)

I think I'm ready for my mop back.

Hey, I'd like to request a restoration of my sysop tools. I went over the complaints from the last ATIM case involving me, and it all seems to boil down to me being too trigger-happy with the sysop tools (especially revision deletion) and making some very questionable-at-best remarks in the block and deletion logs, and I've been actively working to curtail my outrage at troll behavior. I've actually been holding off on making this request for a while because I wasn't sure if I was ready for the responsibility, but I've been more active in the RationalWiki community lately (to the point where I'm actually working on a mascot for the site) and I feel that I could contribute more to the site if I had my sysop tools back. --Luigifan18 (talk) 16:04, 12 October 2023 (UTC)

I hate to say butI think this might not be a good idea. Luigifan18 is clearly productive and well-meant, but I still think they kinda lack the maturity for being a sysop. Maybe in the future. That being said, that's just my opinion, and I'd like to see what others have to say as well. GeeJayKWhere all evil dwells Where every lie is true 23:07, 15 October 2023 (UTC)
Judging by the lack of engagement here, I think everyone else was just waiting for someone to say it first. So thanks for that. Needless to say, I agree. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 23:20, 15 October 2023 (UTC)
Lacking maturity is definitely right. The main reason I voted for their mop to be taken away from them less than a year ago was because of their making death threats against trolls while blocking them. Then a month and a half ago, they posted in the Saloon Bar that a song which basically says everybody on the political right should "hurry up and die" and "stop wasting air" should be RationalWiki's theme song. That's twice in the past year they've made me say, "We need to be better than that." I'm also sure nobody else gives a shit about RationalWiki having a Pokémon style mascot. If we had an actual image of it instead of just a lengthy text description, that might be a little bit different. I will give Luigifan18 credit for coming up with RationalWiki: The Musical, however. I've had fun with that. Spud (talk) 23:38, 15 October 2023 (UTC)
This is actually something I've been struggling with for my whole life. I legitimately get enraged when I see innocent people suffering, especially when it's due to the malice of others. I despise cruelty and hatred, and I can't stand narcissism, sociopathy, or bigotry. So I tend to get hot under the collar when I see people expressing toxic and harmful viewpoints such as white supremacy or antisemitism. I despise the Republican Party because it's descended into nothing but a fascist mockery of concepts like "patriotism" and "liberty" that revels in causing pain and despair to everyone who isn't part of its "master race". As for killing conservatives… I have no illusions about that being a viable long-term solution to the problems that conservatives are causing. It isn't. The long-term solution is to improve education so that people are informed enough to not fall for woo or conservative bullshit, and to improve mental health care and welfare so that nobody's life sucks so badly that they'll grasp on to any ideology, no matter how depraved or idiotic, to gain even a faint veneer of hope or self-esteem. I just have trouble thinking of a good short-term solution that will prevent existing conservatives (who are too neck-deep in their depraved ideals to even consider changing their minds) from undermining the improvements that society needs to gain a population where conservatism and fascism are universally recognized as bad ideas. My vitriol towards conservatives and fascists comes not from a desire to see them suffer, but a desire to prevent them from causing others to suffer. I understand that wishing death upon them is far from ideal, I just don't know how else to truly stop them from hurting people.
As for Azureality… I've had the idea of commissioning someone to draw it on the back burner for a while (I'm not much of an artist myself). I do have a decent idea of what I want it to look like — a blue donkey-goat hybrid creature decorated with symbols of rationalism and science, such as having fur that resembles a lab coat and horns that resemble test tubes. (Pokémon as a franchise tends to have its version mascots be red and blue by default, and it makes sense for Azureality and its opposite number, Crimsolusion, to take the colors of the Democratic and Republican Parties. Azureality's shiny color could be gray, to be more clearly representative of RationalWiki.) Luigifan18 (talk) 14:58, 16 October 2023 (UTC)
I ah, don't think that "maturity" is itself pretty fair. We have sysops who regularly hurl ableist slurs and others who often resort to disingenuous debating tactics (not naming names). So that by itself is not a fair metric, I think. What matters is whether or not a contributor is disruptive to the project, or at least that's what should matter. Also, I see the Pokemon project as a harmless side-project. That a user is passionate enough to engage in such a project shouldn't be a mark against them. That being said, I do think we should have a higher standard for our sysops, and wishing death on people, regardless of who they are, doesn't really reflect well on the project. Some of the essays added by Luigifan18 have been more or less questionable too, but we have sysops who also add questionable content to mainspace, so... Carthage (talk) 15:05, 16 October 2023 (UTC)

Iban between A Somebody and Carthage

Now, to be fair with Carthage, I don't recall them being uncivil with A Somebody, it's mostly a one-sided problem. A Somebody has trolled discussions with Carthage for months, including Carthage's talk page, and I don't think anything good will come with their interaction. I'd like to propose an iban between the two of them. GeeJayKWhere all evil dwells Where every lie is true 20:57, 25 October 2023 (UTC)

Oh. You already posted something. I'd support an iban. I know we haven't always gotten along well but I am genuinely thankful you've brought this up. Carthage (talk) 20:57, 25 October 2023 (UTC)
I wasn't trolling her when I said 'hey timothy Synder is a moron, maybe we should remove that section'? But than when I actually do remove it (and she agreed with me that Timothy Synder is, infact a moron) she reverted it for no reason. I get angry and she knows that, she has been actively pissing me off by faking her agreement. A somebody. (talk) 21:03, 25 October 2023 (UTC)
Uhm, no I didn't. I paraphrased you. Carthage (talk) 21:04, 25 October 2023 (UTC)
I think your problem with the police forces of America is that they are not of the religious variety. A somebody. (talk) 21:11, 25 October 2023 (UTC)
(EC) Relevant. [3] That being said, please, don't squabble ITT. GeeJayKWhere all evil dwells Where every lie is true 21:12, 25 October 2023 (UTC)
(EC) I fully support this. A somebody is clearly obsessed. I was considering a proposal of this type but thought at the time that A) an Iban would make it look like both sides are at fault and B) it seemed weird to propose something implicitly on behalf of carthage without consulting them.-Flandres (talk) 21:16, 25 October 2023 (UTC)
(EC) Carthage again actively fucking messes with me, she fake agrees with me so she can continue to further her interest of messing with me. But noooooo she would never do that, just ignore the French Revolution talk page and the dictatorship talk page. A somebody. (talk) 21:19, 25 October 2023 (UTC)
While Carthage has been egging A somebody. on, I don't see her as the aggressor here and A somebody. has a rap sheet of debauchery at this point. I would support further sanctions on A somebody. beyond just an iban. Plutocow (talk) 00:25, 26 October 2023 (UTC)
Carthage is clearly the fucking aggressor, when ever she sees me get into an argument she comes in and pisses me off with her shitty arguments and fake agreement. A somebody. (talk) 00:47, 26 October 2023 (UTC)
If you hate her so much you should support this iban so you'll never have to interact with each other anymore. GeeJayKWhere all evil dwells Where every lie is true 01:10, 26 October 2023 (UTC)
I support it. A somebody. (talk) 01:13, 26 October 2023 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ So, with both users supporting the iban, candidate for speedy resolution? 𝒮𝑒𝓇𝑒𝓃𝑒 talk 02:09, 26 October 2023 (UTC)

Close for the iban, sure, but we should still vote for sanctions on A somebody., they've been sanctioned before and I'm pretty sure we're all tired of their crap. Plutocow (talk) 06:28, 26 October 2023 (UTC)

A somebody again

A somebody reverted a talk page edit of mine twice in a row. I know that some people here think I have a "feud" with A somebody but I have done my damnedest not to feed the troll since that spat in July. This is getting tiresome. I make a comment about institutions and law enforcement and get called a "monarchist" and the user in question still has that genocidal exhortation on their userpage. They also have resorted to Godwin against other users too. And their edits to mainspace are generally disruptive. I can't be the only person tired of this, right? Carthage (talk) 20:57, 25 October 2023 (UTC)

I gotta be honest, I never thought this was a feud. This seems more like A Somebody holds a grudge against you that prejudices their viewpoints and shapes the way they interact with the site as a whole. I'm not a fan of iBans but I'm open to one in this case, if not some other mod action. Pumpkin pixel glow.gifChef Moosolini’s Ristorante ItalianoTry our pumpkin spice lasagna! 22:09, 25 October 2023 (UTC)
Cartage actively messes with me but noooooo, I'M the one with the a grudge. A somebody. (talk) 22:15, 25 October 2023 (UTC)
Interaction ban between Carthage and A Somebody is what I support, maybe with a short ban for A Somebody so they can calm down. The Krusty Krab (talk) 22:15, 25 October 2023 (UTC)
Apart from any question of an interaction ban, messing with other people's talk page comments is, in itself, an offense. Probably a mod should make this clear to "A somebody".Bob"Life is short and (insert adjective)" 05:45, 26 October 2023 (UTC)
An interaction ban is fine, but A Somebody is already under sanction for past misbehavior, so a block of a short length might be also warranted. Bongolian (talk) 07:16, 26 October 2023 (UTC)

Voting on sanctions for A somebody.

This vote has been closed. Please do not add, remove or change votes.
The result of this vote was: By a unanimous vote, A somebody. is banned for pi months and the interaction ban between A somebody. and Carthage passes.

Pi weeks ban

Aye
  1. The bare minimum. Plutocow (talk) 21:09, 26 October 2023 (UTC)
  2. Ditto. Carthage (talk) 21:43, 26 October 2023 (UTC)
  3. Would be fine with this, although it's obviously already been applied before. -- Techpriest (talk) 15:19, 27 October 2023 (UTC)
  4. I like saying aye. Ditto to Plutocow's reason. The Krusty Krab (talk) 18:12, 28 October 2023 (UTC)
Nay
  1. Arcadium Trancefer (talk) 10:33, 27 October 2023 (UTC)
Goat

Pi months ban

Aye
  1. Pi weeks didn't work last time, and I don't think he'll cool down now either. GeeJayKWhere all evil dwells Where every lie is true 21:18, 26 October 2023 (UTC)
  2. If permaban doesn't pass. Carthage (talk) 21:30, 26 October 2023 (UTC)
    Since you are voting multiple times I think you should be permanently banned from this site and promptly sealed in an underground bunker. A somebody. (talk) 00:42, 27 October 2023 (UTC)
    Now I'm convinced that you're not an adult with this shit. Arcadium Trancefer (talk) 23:08, 27 October 2023 (UTC)
  3. Since pi weeks didn't improve matters. Bongolian (talk) 23:32, 26 October 2023 (UTC)
  4. Scream!! (talk) 00:42, 27 October 2023 (UTC)
  5. It's been a few months and I haven't seen improvement.-Flandres (talk) 01:43, 27 October 2023 (UTC)
  6. Pumpkin pixel glow.gifChef Moosolini’s Ristorante ItalianoTry our pumpkin spice lasagna! 01:44, 27 October 2023 (UTC)
  7. It'll be a nice pi months. Plutocow (talk) 04:47, 27 October 2023 (UTC)
  8. Next time in trouble permaban ←§ Reichtangle (talk) 06:12, 27 October 2023 (UTC)
  9. Bob"Life is short and (insert adjective)" 06:55, 27 October 2023 (UTC)
  10. Bongolian (talk) 07:01, 27 October 2023 (UTC)
  11. Arcadium Trancefer (talk) 10:30, 27 October 2023 (UTC)
  12. 𝒮𝑒𝓇𝑒𝓃𝑒 talk 11:07, 27 October 2023 (UTC)
  13. ShabiDOO 14:07, 27 October 2023 (UTC)
  14. Per GJK. -- Techpriest (talk) 15:19, 27 October 2023 (UTC)
  15. Per GJK.The Krusty Krab (talk) 18:18, 28 October 2023 (UTC)
  16. Bunker, to be unsealed in pi months. Chillpilled (talk) 23:22, 28 October 2023 (UTC)
  17. Spud (talk) 15:40, 29 October 2023 (UTC)
  18. --Gang O' Shadow Wizards 🧙‍♂️Bore me 15:23, 30 October 2023 (UTC)
Nay
Goat

Permaban

Aye
  1. Kinda getting tired of this user's shit. Weeks didn't work last time, they came back with the same shit as usual. I haven't been given the impression that a Pi months ban would work either. And remember, I'm not the only user they've trolled. Carthage (talk) 21:32, 26 October 2023 (UTC)
    'Why does the guy whom I actively piss off get angry at me?'-You A somebody. (talk) 21:35, 26 October 2023 (UTC)
    Didn't he also accuse Geejay for being a "Libertaryan" several times? Arcadium Trancefer (talk) 10:52, 27 October 2023 (UTC)
    Yeah. Carthage (talk) 18:41, 27 October 2023 (UTC)
  2. ShabiDOO 14:07, 27 October 2023 (UTC)
  3. I know this won't pass, but I'm almost certain we will revisit this in a few months.-Flandres (talk) 18:37, 27 October 2023 (UTC)
    Yeah, I'm willing to bet it'll be like this. Unless he completely forgets about the site, because he found something else to spend his free time on... Arcadium Trancefer (talk) 23:05, 27 October 2023 (UTC)
  4. Aye, someone with such a block record and problematic history such as this is not worth going round and round with pi month bans. Lavalizard101 (talk) 20:20, 30 October 2023 (UTC)
Nay
  1. If pi months proves insufficient. 𝒮𝑒𝓇𝑒𝓃𝑒 talk 11:07, 27 October 2023 (UTC)
  2. Excessive. -- Techpriest (talk) 15:19, 27 October 2023 (UTC)
  3. A tad excessive. The Krusty Krab (talk) 18:18, 28 October 2023 (UTC)
  4. I’m like 80% sure we’ll end up here eventually, but I don’t think we should jump the gun. Pumpkin pixel glow.gifChef Moosolini’s Ristorante ItalianoTry our pumpkin spice lasagna! 22:14, 28 October 2023 (UTC)
  5. What Duce said. Chillpilled (talk) 23:10, 28 October 2023 (UTC)
Goat

Interaction ban between Carthage and A somebody.

Aye
  1. They both want it, so… Plutocow (talk) 21:09, 26 October 2023 (UTC)
  2. GeeJayKWhere all evil dwells Where every lie is true 21:18, 26 October 2023 (UTC)
  3. Carthage (talk) 21:30, 26 October 2023 (UTC)
  4. If both users want it, who can argue against it?-Flandres (talk) 21:59, 26 October 2023 (UTC)
  5. Bongolian (talk) 23:31, 26 October 2023 (UTC)
  6. Spud (talk) 00:09, 27 October 2023 (UTC)
  7. Pumpkin pixel glow.gifChef Moosolini’s Ristorante ItalianoTry our pumpkin spice lasagna! 01:44, 27 October 2023 (UTC)
  8. Bob"Life is short and (insert adjective)" 06:56, 27 October 2023 (UTC)
  9. Arcadium Trancefer (talk) 10:31, 27 October 2023 (UTC)
  10. ShabiDOO 14:07, 27 October 2023 (UTC)
  11. Both parties want it and for the sake of keeping the peace we probably should. -- Techpriest (talk) 15:18, 27 October 2023 (UTC)
  12. Why are we even voting on this? The Krusty Krab (talk) 18:19, 28 October 2023 (UTC)
  13. --They wanted this anyways Gang O' Shadow Wizards 🧙‍♂️Bore me 15:24, 30 October 2023 (UTC)
  14. Hopefully this will keep things out of ATIM CorruptUser 20:10, 30 October 2023 (UTC)
Nay
Goat
  1. Are there any topic bans that might help? Plutocow (talk) 21:09, 26 October 2023 (UTC)
    Don't think so. He stopped the edit war on the French Revolution page. GeeJayKWhere all evil dwells Where every lie is true 21:18, 26 October 2023 (UTC)
  2. Also, I know we said that we can treat the iban as passed, but since we're voting on everything else we might as well vote to affirm this. It prevents rule lawyering and all that. Plutocow (talk) 22:01, 26 October 2023 (UTC)
  3. Despite that we all love being officious busybodies, when two parties agree to an interaction ban a vote is a peculiar requirement given that the parties concede that such a ban would be preferable to continued interaction. It should be sufficient for the parties involved be willing to accept the ban, that is, be unwilling to appeal the ban once it has been proposed by moderators. UncleKrampus (talk) 16:05, 27 October 2023 (UTC)
    Yeah I feel like this should have been enacted by mod fiat since both users agreed to it. Pumpkin pixel glow.gifChef Moosolini’s Ristorante ItalianoTry our pumpkin spice lasagna! 22:15, 28 October 2023 (UTC)
  4. I have nothing against Krusty but do they have the edit amount required to vote here? I forgot what the amount required is. Chillpilled (talk) 23:08, 28 October 2023 (UTC)
    It's 75, I believe. But you can tell they qualify because they are in the usergroup "eligible user", which can be determined by simply hovering over the link to their userpage. Plutocow (talk) 03:32, 29 October 2023 (UTC)
    Would anyone care if I gave Krusty autopatrolled? The stupid edit that got them promoted seems to me to be more the mark of a newbie wiki editor than anything genuinely malicious, and I'm tired of patrolling clearly good faith edits. Carthage (talk) 03:34, 29 October 2023 (UTC)
    I wouldn't mind it. Chillpilled (talk) 16:09, 30 October 2023 (UTC)

Umaru16 (yet again)

They're now displaying a WWII propaganda poster with a racist caricature on it with text below it reading in bold "China and Japan are very evil garbage countries!!!!!" This is entering the realm of bigotry, and given that this is not the first time they've made bigoted comments (see the first time I sent them to ATIM) and they're a self-admitted nationalist, I don't see this problem going away. What do you suggest we do here? Plutocow (talk) 04:31, 9 November 2023 (UTC)

I didn't attack ethnic Chinese or ethnic Japanese. It is just an image brought to denounce Chinese imperialism and Japanese imperialism. There is room for consideration in that it is the propaganda of the Allies during World War II. (I just removed it from the user page.) This image is basically a poster denouncing Japanese imperialism, depicting Hideki Tojo as a monkey. Simply put, it's like portraying Adolf Hitler as a monkey. I have never attacked Japanese users racially. Umaru16 (talk) 05:06, 9 November 2023 (UTC)
The problem with your claim here is that the poster had nothing to do with Japanese foreign policy. It was US wartime propaganda warning against wastefulness. Pizza SLICE.gifChef Moosolini’s Ristorante ItalianoMake a Reservation 05:23, 9 November 2023 (UTC)
I think Plutocow is overly sensitive to anti-Japanese issues. By his standards, most South Korean liberals and progressives will be anti-Japanese racists. Plut lacks an understanding of anti-imperialism and post-colonialism. In South Korea, the group with strong anti-Japanese sentiment is rather an active supporter of the 차별금지법(anti-discrimination law), and the right-wing group with little anti-Japanese sentiment is anti-Chinese/anti-Islam/anti-multiethnic/anti-North Korean/anti-LGBT/anti-feminist and anti-차별금지법 fascist. (The anti-discrimination law is a law that outlawes racial discrimination in actual workplace, residence, or legal matters. South Korea's leftist anti-Japanese nationalists, denounced by rightists like Shin Ki-wook, support the law, and many pro-Japanese rightists in South Korea are opponents of it.) Umaru16 (talk) 05:43, 9 November 2023 (UTC)
Yes, opposing racist characatures made by the US during WWII is being "too sensitive". You do know what happened to Japanese-Americans during the war right? But Umaru is a nationalist, they'll say anything is justified as long as Korea does it or if it harms Japan or China. Plutocow (talk) 05:49, 9 November 2023 (UTC)
I didn't post the image on my user page with that intention. And I removed the image from the user page as soon as you took issue with it. Is this the case that will go all the way to All Things in Modulation? I just posted on my user page what is known as a representative image of anti-Japanese anti-fascist (like anti-German anti-fascist) propaganda during World War II. From my nickname in the first place, it was named after a girl in Japanese animation (Umaru Doma), I like J-Pop, and I don't hate conscientious Japanese people who face the history of the past. Umaru16 (talk) 08:36, 9 November 2023 (UTC)
I don't actually think I'm a nationalist. However, anti-colonial nationalism cannot be put on the same line as imperial nationalism. Gandhi's anti-British nationalism cannot be equated with Hitler's anti-Semitic nationalism. Rather, imperialism would still have been maintained without anti-colonial resistance. (Likewise, anti-Korean nationalism by Japanese conservatives and anti-Japanese nationalism by South Korean liberals are not on the same line.) Anti-colonial nationalism is just one identity politics. And I'll add one more thing. I never attack post-colonial countries. China is an exception because it is an imperialist country. Umaru16 (talk) 05:11, 9 November 2023 (UTC)
Generally, we're rather lenient about what a person can put on their own home page, and that image is from Wikimedia Commons. It is historical racist poster (based on the stereotyped caricature) from the US during World War II. Unless there's evidence that Umaru16 is pushing a nationalist agenda on mainspace pages, I don't see anything that should be done at this point. Bongolian (talk) 05:35, 9 November 2023 (UTC)
Umaru16 has been promoting a Korean nationalist agenda in the mainspace, which resulted in them being taken to ATIM at least three previous times. Like, their whole current crusade is trying to promote the idea that Korean nationalism isn't racist, which I have largely avoided as dealing with this user is incredibly frustrating. This is just a symptom of a larger issue. Also, we do have exceptions to what we allow in userspace, and blatant racism should be one of them; Umaru16 clearly wasn't trying to give historical context or anything to the poster, so it was just racism, although since they removed it it's a moot point. Plutocow (talk) 05:49, 9 November 2023 (UTC)
No, I didn't edit the mainspace article from a Korean nationalist agenda perspective. Are you talking about the Category talk:Irredentism? I was only referring to Irish and Korean nationalism in that irredentism is a nationalist ideology but not necessarily imperialism. Umaru16 (talk) 05:58, 9 November 2023 (UTC)
I don't think Korean nationalism is racist when it is an issue related to post-colonialism or national identity about Japan, the United States, and broadly China. But I think it is racism when Korean nationalism is anti-Muslim, anti-immigrant, and racial discrimination against black people, similar to the European far right. Brian Myers and Shin Ki-wook attack anti-Japanese nationalism, which has rarely led to hate crimes and is mostly a post-colonial issue, but they rarely condemn South Korea's anti-Chinese nationalism, which has actually led to hate crimes. (Of course, postcolonial anti-Chinese sentiment and anti-Chinese racism must be distinguished.) American liberals express anti-Russian sentiment very often even before the Russia-Ukraine war, but I don't think it's racist. Japan to South Korean liberals is Russia to American liberals. Umaru16 (talk) 06:10, 9 November 2023 (UTC)
I find your nationalism to be very obnoxious most people on this site don't even agree with you and how is north korea anti-imperialist? And why are you trying push Korean nationalism on this site? ←§ Reichtangle (talk) 06:17, 9 November 2023 (UTC)
I do not deny North Korean crime and racism and totalitarian monopoly. However, I just wrote that translating the word 민족 as "Korean race" without context is a clear mistranslation and reflects anti-Korean prejudice. This is not a view of Korean nationalism, but against anti-Korean prejudice (including both South and North Korea). South Korea's main left-wing and center-left parties often include 민족 in their party platform. Umaru16 (talk) 06:23, 9 November 2023 (UTC)
I think North Korea is anti-imperialistic. Anti-imperialism does not mean good or bad. Anti-imperialism is just an ideology. Ukraine supported Azerbaijan when Azerbaijan invaded Armenia. However, Ukraine is anti-imperialistic in that it stands up to Russian imperialism. North Korea is anti-imperialistic in that it stands up to U.S., Japanese and Chinese imperialism. North Korea cooperates with Russia for survival. It is practically difficult for a particular country to oppose all imperialism. I am not a Korean nationalist. However, the right wing of South Korea is a Japanese nationalist, similar to the Russian nationalist among Ukrainians. I also support the anti-Chinese and anti-imperialism of Taiwan nationalism, but that doesn't mean I automatically become a Taiwanese nationalist. I am not a Korean nationalist, but I support Korean nationalism in that it is anti-imperialism. Umaru16 (talk) 06:32, 9 November 2023 (UTC)
For your information, most Korean nationalists are against Taiwanese nationalists. Taiwanese nationalists are very pro-Japanese, anti-Chinese and hate Chiang Kai-shek. But Chiang Kai-shek is also a hero of the Korean independence movement. I am not a Korean nationalist. I just support anti-imperialist solidarity. (Conservatism in many parts of Asia is similar to that of Europe in the early 19th century. In the early 19th century, nationalism was a progressive ideology.) Umaru16 (talk) 06:35, 9 November 2023 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ So @Plutocow why is dealing with this user very frustrating ←§ Reichtangle (talk) 06:48, 9 November 2023 (UTC)

I'm not qualified to analize Umaru's mainspace edits, I find some of them to be strange (see the "anti-Korea" section on our Wikipedia article, for instance), but I think that their userpage is fine now that they removed some content. That being said, I don't like when people say they hate a whole country. I'm pretty sure most of us hate the LDP and the CCP almost as much as Umaru, but we shouldn't conflate the people of these countries with their horrible governments, even when the said people seems to support these parties. My advice to Umaru is, don't hate other countries, focus more on ideas. GeeJayKWhere all evil dwells Where every lie is true 14:08, 9 November 2023 (UTC)
In a strict sense, I didn't create the "anti-Korean" session. # Although the dictionary translates the word Minjok into English as people, nation, ethnicity, race, etc., it never means "race" in the sense that Americans think. Minjok is a term that is also included in the South and North Korean Constitution, and no Koreans use Minjok as the meaning of the English word "race." However, either Myers or Shin or Kelly often refer to Minjok as "Korean race" and accuse the South Korean left of being "race-based nationalists." The left in South Korea is by no means anti-Japanese/anti-American racist. 민족주의 is 'ethnic nationalism' in literal translation, but it is not ethnic nationalism in the United States or Europe. For example, "ethnic nationalist" in English reminds people of Hitler or the Nazis in a negative sense, but "민족주의" in Korean refers to anti-Japanese independence activists or national regular supporters based on resistant nationalism in a positive sense. Umaru16 (talk) 00:03, 10 November 2023 (UTC)
so how do we deal with him? ←§ Reichtangle (talk) 14:20, 9 November 2023 (UTC)

Personally, as of right now I'll file this under "extremely reprehensible yet not actionable," especially because he removed the content in question quickly when called out.-Flandres (talk) 16:14, 9 November 2023 (UTC)

Umaru16 is mainly frustrating because they're a nationalist, and whenever I try to argue against them they just repeat the same two or three arguments ad nauseum without ever engaging with my arguments. I understand that there is a language barrier here, but I feel the big problem here is Umaru16's nationalist mindset. If nothing comes of this case, I suppose I'd be fine with it, but I find this very annoying as this user has been a repeat issue. Plutocow (talk) 17:19, 9 November 2023 (UTC)
They are very annoying when I try to debate with them they overwhelm me with arguements. Their arguement goes like this " Japan bad and very imperialist and so is china. America bad and good at the same time. North Anti-Imperialist. Myers bad and anybody who doesn't like korean nationalism is anti-korea " . They are the type of person to use Anti-Japanese U.S Propaganda posters ( I don't condone Japan's war crimes; Infact I want an article about Japan's war crime denial). He also claims the western Alt-right is Anti-korean the alt-right at worst cares about korean as much as it cares about puppies ←§ Reichtangle (talk) 17:38, 9 November 2023 (UTC)
While I agree with the Gish Gallop definition of the user's debate responses [the few I've spotted in the wild tally] this is not prima facie evidence of 'wrongdoing'. KarmaPolice (talk) 18:57, 9 November 2023 (UTC)
It also tends to be more of an issue in real-life debates than on wikis where you can slowly go point for point more easily if you wish. Chillpilled (talk) 20:08, 9 November 2023 (UTC)
Personally, think it's the other way about; a bigger problem online because it's much easier to shove out 'walls of text' with a dozen different points within [I shall admit to being a bit guilty of this myself] than in a physical debate [where either the moderator/opponent will interrupt the ramble and/or simply cut their mic and/or will have time limits for responses]. What's more, a live debate/discussion will feature much more ad-libbing than an online one because you can't be [usually] trawling Google etc for stats/quotes etc to back up your points and while you can usually predict most of the questions there shall always be some curve-balls you couldn't prepare for. KarmaPolice (talk) 20:27, 9 November 2023 (UTC)
@Reichtangle Brian Myers is really SK right-wing. He has often referred to South Korean liberals as ethnic nationalists in a negative sense and has interacted with South Korean conservatives. Therefore, it is worth noting that when he is quoted in English Wikipedia or in the media, his views may be biased. Korean "민족주의" and English "ethnic nationalism" have very different meanings. Can it be considered "race-based nationalism" for South Korea to celebrate 2333 BC without not celebrating 1948? The mainstream political forces in South Korea, except for the radical pro-U.S. right wing, disagree at all. It is artificial to celebrate 1948. No matter how much I googled, I couldn't find his interviews with South Korean liberal media. But he frequently interviews right-wing media in South Korea. Umaru16 (talk) 20:53, 9 November 2023 (UTC)

I'm somewhat skeptical of Umaru but they've justified certain of their dubious-looking edits in mainspace with edit summaries even if I can't verify how valid they are as I don't speak Korean (so really have no specific grounds to revert them). The racist-offensive connotations of the WWII poster might be lost on somebody from Korea. That specific discussion about the poster probably should've taken place on their talk page before here. Chillpilled (talk) 18:45, 9 November 2023 (UTC)

There's another problem with some of Umaru's arguments, generalization of national opinions without citation. Sure, there's good reason for Koreans to dislike the Japanese because of the colonial history, but at the same time, there's high-level international cooperation between the two countries and there are many Koreans who live in Japan (though there's also discrimination against them). Has this sort of unsubstantiated generalization crept into mainspace pages? I don't know. Bongolian (talk) 18:57, 9 November 2023 (UTC)
Umaru's opposition to the work of Brian Reynolds Myers could hold validity, in my view, I don't know that Myers is a "right-wing" scholar in specific but his claims don't align with the harsh denunciations of fascist racialism in works by the North Korean state I've seen. Specifically, in the works of Kim Jong-il (such as here). It's a creepy police state but I'm not under the impression it's racially supremacist. Some other academics have criticized Myers's writing on this as well. Chillpilled (talk) 19:19, 9 November 2023 (UTC)
My view of Umaru16 is generally ambivalent. They can tone down the nationalism a bit, but I think their overall content has been pretty informative from the Korean information sphere. That said, I'm not sure I agree with everything they write (irredentism can be anti-imperialist?), but regarding Korea I don't know enough about Korea to dispute whatever they write. Carthage (talk) 20:55, 9 November 2023 (UTC)
They said in one of their edit summaries here they're using a translator to help communicate. Though I don't doubt that's what they actually meant, but just to be clear, misunderstandings are abound. If one for example considers an Israeli-Palestine two-state solution to be somewhat irredentist, which, it could be construed that way by one definition. You may begin to see what they might mean. Chillpilled (talk) 21:58, 9 November 2023 (UTC)
They've been warned in the past for this behavior if memory serves me right. I understand that east Asia's relationship with Japan is... At best strained, but that doesn't mean that putting old racist WW2 propaganda posters on your user profile is the right decision. Granted, they've been removed and that's hopefully the end of that. I will note that this isn't the first time Umaru has been dragged to ATIM for similar behavior (the previous time the reporter being Plutocow). I'm very unsure what to think of all this; my tacit suggestion would be for Umaru to perhaps stick to writing about Korea rather than Japan; they clearly have an ideological blind spot when it comes to Japan. -- Techpriest (talk) 23:57, 9 November 2023 (UTC)

Johns and Oliver Smith

@Spud and I have received communications alleging that @Johns is receiving suggestions from banned user Oliver Smith, and acting upon the communications (on RW). I don't think that this situation is against any specific RW policy since Johns has not been alleged to be a sock. This is a chance for Johns to speak in his own defense, and for anyone else weigh-in on policy matters. Bongolian (talk) 02:14, 13 December 2023 (UTC)

Smith emailed me in the past about this Kane drama, I have not spoken to him in over 6 months. I deleted my email and no longer use it. It is incorrect to say I have received communications about this. Until today, I had never heard of Justapedia before. I am guessing these communications came from Jonathan Kane who posts as Shuichi Tezuka. It's probably because I left a comment here a few hours ago [4] about his recent activity at Justapedia. There is recent talk on wikipediocracy [5] that he is Shuichi Tezuka of Quillette fame. The users over there laugh at his alt-identity. Everyone that has looked into these matters knows they are the same person. In the past I objected to including that information on his article because we had no reference but now I am not so sure.
Kane doesn't want anyone editing his RationalWiki article and he will try and get users banned here for editing it. I don't think we should be helping this person, just like we shouldn't be helping Abd Lomax or any of the other people pissed at us. Johns (talk) 03:03, 13 December 2023 (UTC)
I don't like Kane's views, but I wouldn't lump him in with Lomax because 1) he's not banned, 2) he had some reasonable complaints about RW pages that were addressed to his satisfaction. That does not however allow him to censor reasonable edits to the page about him. Bongolian (talk) 03:16, 13 December 2023 (UTC)
I think acting on the suggestions of a banned user is a roundabout way of ban evasion, and is thus actionable. That said, I don't believe we've been given any reason to trust this alleged source. Plutocow (talk) 04:27, 13 December 2023 (UTC)
After I read Jonathan Kane's email, in which he kind of alleged that Oliver D. Smith was using Johns as a meatpuppet but didn't offer anything to back that up, I decided no further action needed to be taken. As Bongolian said, no specific rules at RationalWiki were being broken. Spud (talk) 09:03, 13 December 2023 (UTC)

I would have rather not been named in public as the person who originally raised this issue, but I guess some things can't be avoided. Now that I have been, I may as well explain the reason for my suspicion.

As background, RationalWiki's article about me was originally created in 2020 by a Smith sockpuppet, and subsequently deleted as well as oversighted. The discussion about its deletion is here, confirming that Smith was the author. A new version of the article was created by Johns in July of this year, and while his version of the article had some differences from Smith's deleted article, many parts of his article used wording that was identical to the wording Smith had used. (I am referring here to Johns' originally recreated version of the article, not to the current version which was rewritten by BobJohnson.)

Here is how Smith's original, deleted article from 2020 described my background:

Kane describes himself as an “armchair palaeontologist”. Although having no degree in palaeontology, he worked as an intern at the New Jersey State Museum under natural-history curator David C. Parris from 2001 to 2005.

Here is how Johns' recreated version of the article from July described my background:

Kane describes himself as an “armchair palaeontologist”. Although having no degree in palaeontology, he worked as an intern at the New Jersey State Museum under natural-history curator David C. Parris from 2001 to 2005.

Here is how Smith's original article described my political views:

He considers his political views to be libertarian but given his controversial views on race, paleolibertarianism is more accurate.

Here is how my political views were described in Johns' recreated version. The only change was to remove one of the links that had been in Smith's sentence.

He considers his political views to be libertarian but given his controversial views on race, paleolibertarianism is more accurate.

Here is how Smith's original article described how my writings have been received by creationists:

Unsurprisingly, a negative review was published in the Journal of Creation by John Woodmorappe. Kane has responded to these criticisms.

Here is how the same thing was described in John's recreated version. He removed a link and added a source, but otherwise made no changes to Smith's wording.

Unsurprisingly, a negative review was published in the Journal of Creation by John Woodmorappe.[18] Kane has responded to these criticisms.

I could keep going, but these three examples ought to be enough to demonstrate the pattern. Around half of Johns' recreated version of the article was using material that had been originally written by a Smith sockpuppet. Apparently one of two things was the case: either Johns was copying from a three-year-old deleted (and oversighted) article written by a banned user, or he had Smith's help in recreating the article.

To be clear, I'm not arguing for any specific sanction against Johns. However, if he has in fact been proxy-editing for Smith, I don't think that issue should be totally ignored either. Tetrapteryx (talk) 10:29, 13 December 2023 (UTC)

This was not proxy-editing for Smith, I was not even aware who originally wrote the article. I wrote about this before, I just recreated the article from archive. I basically copied the old version that was on archive.is [6]. What happened here, is that years ago your article was created but you had Arthur Kerensa a friend of yours and his alt identity to delete it. Kerensa who posted as D and EK both admitted to being a friend of yours (you all post on Wikipediocracy). D posted that the reason he deleted your article is because he knows you and you are a "good" friend of his. He incorrectly deleted the article without any admin support apart from his other account EK his other sockpuppet.
D/EK have since been banned from RationalWiki. They incorrectly deleted your article for "doxing". D/EK also banned Smith for "doxing". I was also warned by this user in the past not to use his name as it is private. However, please check his very own userpage on this website ! [7] "I am Arthur Kerensa". He was very open about his identity and wrote his own name many times. He also confessed to sending users dick pictures, "You're the banned user Arthur on Uncyclopedia, who after getting banned, relentlessly spammed our wiki, harassed and sent dozens of unsolicited dick pictures to our Discord and wiki users" [8]. This guy was a lunatic that is banned from this website and plenty of others. After being banned from here he struct up a relationship with Nathan Larson. I basically undid his deletion which was against policy. Your article has since been re-written by another user BobJohnson [9]. What is the real issue here? You just do not want users editing your article. Johns (talk) 11:02, 13 December 2023 (UTC)
Echoing PC here insofar that open meatpuppetry would be actionable. The connection between Johns and Smith was never particularly a secret either. As for Smiths ban; keep in mind that he was rebanned alongside Abd in the coop when it became clear Dysk enacted both bans outside procedure (Smith doxxed someone after being banned, Abd attempted to raid RW). Insofar as the rest goes; it seems to pertain to edits that have been struck from an article by BobJohnson during a rewrite. I'm gonna concur with Spud on this one; there's simply not enough here, nor any real active harm to be actionable about. Also, and I say this purely from personal frustration - speculation and arguing about Dysks identity, motives and actions is not gonna be worth it. He's banned from RW due to him claiming that he mishandled the PII of editors. Were not a gossip rag for his skullduggery. Anything to do with Dysk has been relitigated so many times that it feels like kicking a dead horse. The twitching of the leg kinda stopped being interesting when it was beginning to snap off due to the rot. -- Techpriest (talk) 11:41, 13 December 2023 (UTC)

Since the new user is new and also currently blocked: "Bizarre claims of plagiarism"

This discussion was moved here from RationalWiki:Technical_support.

My account 'Box' was blocked for 3 days for alleged plagiarism for having created the Natal Conference article.

There was no plagiarism on the page - I mentioned Austin chronicle and cited the article as a source in a quote.

Bizarrely, I was blocked for plagiarism with someone saying I did not attribute the Austin chronicle as a source.

https://i.ibb.co/PTLBLvp/Austin-Chronicle.png - clearly shows I did not plagiarize. 2A00:23C6:C022:C701:601C:6D8A:4C76:502E (talk) 09:29, 28 December 2023 (UTC)

Hi Mods! Won't ping you all, cause... uh... I can't be arsed? Anyway, here is the relevant thingy from our new Box. Y'all can have a look at the deleted revisions as to what you think. Then, here is Bongolian telling Box that plagiarism in unwelcome on the site. Bongo nukes the page and blocks Boxy for pi-days. Afterwards, a BoN posts on the Tech Board claiming they are Box. They can't post here, so I bring the discussion here. So yeah, that's the timeline. Now for some PERSONAL COMMENTARY:
Firstly, I despise this page (and the coop as well, f@ck that place) and frankly if I could have not posted here I would have liked that because (and I say this as someone who un-ironically has great affection for all you silly bastards on this shithole of a website, like genuinely, I missed you all and it's great to be back) you all get your panties in a twist sometimes. Which leads to my second point: No one f@cking died, mkay? And I'm not playing Wiki Phoenix Wright with my diffs, I'm putting those there so you all see the timeline. No one f@cking died, Bongo is a good mod and a good editor who is an all-round nice guy, who on this occasion jumped the shark, and Boxy is new so we can't be mad at Boxy. So NOTHING should happen. No punishments, or any of that other moronic shit you all clog my glorious recent changes with when you all decide that someone insulted your Samurai sense of honour. Regards. - Rairyu75 (Talk) 16:02, 28 December 2023 (UTC)
Also, this page is protected? So how is Box supposed to... uh, whatever, never mind, I don't care, I don't care, I don't care, I don't care, I don't care, I don't care, I don't care, I don't care. - Rairyu75 (Talk) 16:15, 28 December 2023 (UTC)
I just restored the page. I got confused by the poor formatting (i.e., no quotes around quoted material). Citations were given, so not technically plagiarism, I suppose. Bongolian (talk) 17:47, 28 December 2023 (UTC)

Group of long-term trolls (likely EmilOWK ban evasion)

Apologies if this is already known. I wanted to at least document for the record. And so people know what to look out for.

After scrutinizing the edits of the accounts in question, and the timeline of the site's page creation log, I think the banned user Oliver Smith has largely-correctly flagged a group of long-term trolls who attempted to both imitate him and "parody" (vandalize) RationalWiki for not supporting racism. Assuming one generally distrusts Smith, let's even call it a stopped clock moment (the fact he was the one who initially made the connection is really not important, aside from the fact the trolls are apparently trying to imitate him — let's call them "pseudo-Smiths").

@Tetrapteryx (confirmed by himself (1), GeeJayK (2), and @Bongolian (1) to be Jonathan Kane) gives a testimony here that suggests the RW-banned user Emil Kirkegaard (User:EmilOWK) was behind this trolling:

  • "I also want to provide some more detail about the low-level trolling that's been going on, both here and at Wikipedia, of which the Richard Haier article [created by Kfotfo] was one example. I received an email from Emil Kirkegaard asking me to help him rewrite [a Wikipedia page], and offering to share the WP account's login details with me. But I was unwilling to help Kirkegaard turn that article into a parody, and he and I argued for more than a week about whether his doing that was a useful strategy. Kirkegaard and I have actively avoided one another ever since then."

Smith has claimed off-site that "parody" articles are specifically made about individuals who Kirkegaard personally has a bone to pick with.

Allegedly, this activity extended to Wikipedia as well, but I'll present on-RW evidence for this apparent abuse.

Timeline and accounts:

The consistent timing from one account to another, combined with the fact that they all edited exclusively on race-related topics, is compelling.

There could be more pseudo-Smiths. Smith suspects Silky, too:

But while there's similarity between Silky and the other accounts (some quirks in writing style between most of the accounts are identifiable), I consider the on-wiki evidence too inconclusive in this specific case. The one page they created had to be rewritten, which is in line with the other accounts.

Some other links between the accounts:

Some misconduct:

This is condensed. There are more examples, especially of bad additions or removals to misc. articles, but these ones are telling. Even if Lusaz weren't connected to the others, I would say the attacks on someone for mental health (which RW:BLOCK lists as unacceptable to discriminate against) could have been interpreted as excessively offensive. Even if CBH weren't connected to the others, I would say the exchange with the IP could have implied a doxing attempt. But I think the proof suggests they are a ban-evading EmilOWK which would also be enough on its own.

Based on everything above, I blocked CBH and Jean Lusaz.

Ways to mitigate this kind of thing

  • Verify strong claims. If a subject is labeled a racialist pseudoscientist, make sure they have actually generated papers attempting to support that position. If they simply defended racialists, or were associated with them, that doesn't in-itself make them a pseudoscientist but makes them someone who defended such a person, associated with such a person, boosted that person's claims, etc.
  • Sysops can spot new edits via recent changes and verify inserted statements. If it is a lot added, one can select a few random claims from the additions to attempt to verify. Failed verifications should be a red flag.
  • Consider the notability of figures who new pages are made about. Consider they may have had a significant impact you don't know about, though.
  • Ensure content is not mean-spirited. I mean, that's in the guidelines somewhere, I remember. If someone is silly that's one thing, but attacking them for physical appearance or mental health struggles or whatever is mean-spirited. If sarcastically pointing out hypocrisy as with "Haggard's law", ensure it's in a more lighthearted not-too-judgemental way.

Chillpilled (talk) 11:00, 30 December 2023 (UTC)

Emil Kirkegaard was indeed behind these parody accounts, it was always obvious he was CBH. The agenda was to discredit RationalWiki articles on race and intelligence research by including made up or stupid claims. Examples of this included the now deleted Richard Haier article. The original Eric Turkheimer by CBH that has now been re-written was an obvious parody as was the original Robert Plomin article. Smith has written a 65 page public document about how Kirkegaard and his associates created fake articles on RationalWiki [10]. See page 29 of his paper. Heiner Rindermann in September 2023 authored a book chapter criticizing RationalWiki citing these hoax articles as evidence against the website [11], he is very deceptive here because he cites old version of these articles before they were corrected. His book chapter was published by Springer a good publisher.
The examples that Rindermann cites are the Robert Plomin and Richard Haier articles. If you scan Twitter for these articles they were cited a few years ago by Emil Kirkegaard and Noah Carl. This was obviously an organized attack against RationalWiki, to create fake articles and then blog about them in an attempt to discredit the website. They have succeeded here because they have managed to get one of their associates Heiner Rindermann to publish about these hoaxes and he has obviously fallen for it hook line and sinker. In total there were only 5 hoax articles and they have now all been deleted or re-written so I consider this a case closed but it is disturbing about the deceptions that were going on. Johns (talk) 15:55, 30 December 2023 (UTC)
Should we put an ATIM notification on CBH's talk page? --Luigifan18 (talk) 17:17, 30 December 2023 (UTC)
He's banned. Carthage (talk) 17:30, 30 December 2023 (UTC)
Yes, but last I checked, that account isn't "disallowed from using their own talk page" banned. Is that meant to be giving them one last chance to defend themselves, or is it simply an oversight? Luigifan18 (talk) 18:00, 30 December 2023 (UTC)
There have been doubts about CBH before, as shown on their talk page. I've put up a notice anyway. Bongolian (talk) 18:38, 30 December 2023 (UTC)
Leaving talk page access was intentional, but if the proof is considered strong enough by others it can be blocked. Kirkegaard's talk page access was revoked for something about what was interpreted as attempted doxing and further disruption attempts from I guess whatever he continued to post there (I wasn't around, and didn't bother to dig up archives on his specific case other than his own talk page, all I knew was he was banned before and for this). Chillpilled (talk) 18:43, 30 December 2023 (UTC)
RE: the Smith paper Johns posted, Smith was apparently wrong about Dysk and Ze being the same person, based on statements I read from people who talked to both. They are linked however. So don't go citing that paper as proof the two are the exact same person is all I'll say. Chillpilled (talk) 19:18, 30 December 2023 (UTC)
The link they posted looked like doxing, so I revoked talkpage access. As for Dysk, I believe a Wikipedia investigation revealed the two to be the same person but that is a dead horse as far as I'm aware. Plutocow (talk) 21:04, 30 December 2023 (UTC)
Dysk posted on his own user-page that EK is his own sock-puppet, a link to his RationalWiki userpage here [12]. He also said in conversation to Abd Lomax on Reddit that EK was a figment of his imagination depending on his mood. Helena Bonham Carter was able to work out Dysk real identity but never posted it publicly, this was in the archives [13]. He is much older than he claimed to be. The name and photo he was using were both fake. I agree it doesn't matter because this user is no longer active here. Johns (talk) 00:30, 31 December 2023 (UTC)
Retrieved from "https://rationalwiki.org/w/index.php?title=RationalWiki_talk:All_things_in_moderation/Archive71&oldid=2613337"