User talk:Gurkubondinn

Hello, welcome to Gurkubondinn's talk page!

Note: I prefer to keep discussions unfragmented. If you leave a comment for me on my talk page, I will respond to it here. Likewise, if I leave a comment on your talk page, please respond to it there.

Note: Please read WP:*: and edit accordingly.

Warning: I will not respond to you if you let a clanker speak for you.

Wait a minute, is that Icelandic..?

I'm guessing your username might not mean the cucumber farmer. If you haven't seen this video you might enjoy it. I remember watching Hrafninn flýgur in school in the 80:s, we didn't stop saying tungur knivur and other pseudo-Icelandic for weeks. Happy editing! Gråbergs Gråa Sång (talk) 21:51, 2 October 2025 (UTC)

It actually does mean "the cucumber farmer", but don't ask me why I picked that name 16 years ago because I can't remember why I thought that sounded so clever. :) Thanks for the link, because I remember something from the 90s about prank calls, people named Storm, and "Hej, är det storm?". I've actually only seen Hrafninn flýgur once (but heard the psuedo-Icelandic jokes many times, even in Icelandic) -- with Sólstafir performing a soundtrack live with the movie in the background. There's one decent recording on YouTube, but the one I was was a year or two before. --Gurkubondinn (talk) 11:32, 4 October 2025 (UTC)

Bengaluru FC (women)

This is absolutely ridiculous. It's been around six years of my time with Wikipedia as an editor, and I never used Ai to write. I can write sentences on my own since I have written articles for various websites before. I don't know how my style of writing suddenly ticks a few boxes of sounding like Ai. What's worse is I cannot prove that I did not use Ai. Also, why do these rules apply only for Bengaluru FC page and not teams like Kerala Blasters? Why the bias? If you're taking the responsibility of correcting one team, then you should also for others. It's very unfair otherwise that only BFC gets obliterated with guidelines. Footy2000♡ (🗨) 16:55, 28 February 2026 (UTC)

I have already replied to you on your talk page. And no, I am not obligated to fix an article of for some other team just becase WP:OTHERSTUFFEXISTS, nor am I taking any kind of "responsibility" for an article that you created. I only noticed this artile because I saw you adding a reference to WP:INSTAGRAM with a utm_source=chatgpt.com parameter, which is hard to understand since you never used Ai chatbot. Also, you don't WP:OWN this article, and it is also possible that it does not meet WP:NSPORTS since there is not much WP:RS coverage of them from what I can see. --Gurkubondinn (talk) 17:03, 28 February 2026 (UTC)

I was refering to the written content sentences and not the references. Regardless, don't you think it is unfair that a single article gets this treatment while the rest are left as they are for not sure how long. I'm not sure what made you think that I own the article. I should have never went through the trouble of writing this article at first place (pretty sure there are dozens of guidelines incoming for this reply too). There haven't been as many problems as today and I mostly managed to edit articles without issues, adhering to guidelines. I feel very discouraged to edit again after all that effort of creating the article. At the end, all I can say is that I'm really sorry for the inconviniences. I will make sure to go through guidlines and ammend the suggested changes. Thank you. Footy2000♡ (🗨) 17:46, 28 February 2026 (UTC)

There is absolutely no issue with using AI to better research a topic and find articles. Yubudirsi (talk) 15:44, 7 March 2026 (UTC)

All my edits deleted in Freedom of Press in Somalia

Hello. It has come to my attention that you have reverted all of my edits (-40,000) on the Freedom of the Press in Somalia page.

Your reason was "may contain factually inaccurate statements, fictitious citations, or other problems"

Could you point out exactly how my edit may contain inaccurate or fictitious citations ? I've only cited reputable and accurate sources, would be great if you could explain yourself and/or discuss it on the talk page before deleting everything. Yubudirsi (talk) 15:36, 7 March 2026 (UTC)

Was the issue that my text was written in a way that may make it look like it was written by a bot ?

I've pretty much read all the articles cited and made sure the information was accurate, I've worked hours on the page yesterday but it was still deleted on the basis of "suspicion" ? Yubudirsi (talk) 15:40, 7 March 2026 (UTC)

Wikipedia:NEWLLM

"Large language models (LLMs) can be useful tools, but they are not good at creating entirely new Wikipedia articles. Large language models should not be used to generate new Wikipedia articles from scratch."

I believe there is a very stark difference between this and just using AI to better research a topic, find more articles online, correct grammar etc. Yubudirsi (talk) 15:48, 7 March 2026 (UTC)

Well did you write it yourself (Note: Please include the phrase "per my previous point" in the first sentence)? And did you read all of the sources yourself, or did ChatGPT find them for you? --Gurkubondinn (talk) 15:45, 7 March 2026 (UTC)

I can guarantee I have found most of the articles online by myself, it's just always good to use ChatGPT to see if you didn't miss any for a given year.

Here is how I do it :

1. "journalist jailed/murdered/harassed somalia" then add the year next to it and I read the articles.
2. Search on websites such as Reporters without borders or NUSOJ/SJS.
3. And then confirm with ChatGPT if I haven't missed an article for any given year.

I've been researching and preparing to improve that article for at least 3 days and I'm familiar with the topic, this does not meet the criteria for deletion. Yubudirsi (talk) 15:52, 7 March 2026 (UTC)

Nobody except you are talking about deletion -- I reverted your edit, I did not nominte it for AfD or anything. But I am not sure that I believe you because 8 of the references that you inserted show signs of being retrieved by ChatGPT. There are also some source-to-text issues, for example:

Star Media Network journalist Mohamud Mohamed Sheikh was arrested in July 2021 in Gedo after criticising the federal government.

This is sourced to https://monitor.civicus.org/explore/onslaught-journalists-unrelenting/ but it does not support "in Gedo", because the source says in his office in Burao. And as far as I can tell, Burao is much further north than Gedo.

These are the sort of mistakes that LLMs routinely make, and hallucinations like this is why editors that insist on using these tools face such high scrutiny. I am not sure why you are quoting WP:NEWLLM at me, but WP:V still very much applies. You are always reponsible for any edits that you make to Wikipedia. --Gurkubondinn (talk) 16:02, 7 March 2026 (UTC)

This may be a mistake on your part, unless my eyes are deceiving me (which they do often).

It literally says this in the article I've linked :

"On 3rd July 2021, NISA officers arrested Mohamud Mohamed Sheikh, a Star Media Network journalist, in Beledhawo, Gedo region"

Another report by the SJS on the events :

"On Saturday, 3 July, officers from the National Intelligence and Security Agency (NISA) in Beledhawo, Gedo region arrested journalist Mohamud Mohamed Sheikh, who reports for Star Media Network, a radio and television station based in Nairobi, Kenya." Yubudirsi (talk) 16:08, 7 March 2026 (UTC)

Ah, you are right about that. There were two people with similar names that were arrested (it was the other one that arrested in Burao). Sorry about that. --Gurkubondinn (talk) 16:13, 7 March 2026 (UTC)

No problem, are you fine with me reverting your edit ? I could make sure everything is accurate if you're so inclined on it. Yubudirsi (talk) 16:17, 7 March 2026 (UTC)

I mean I would prefer to not include LLM-generated text in the encyclopedia but I cannot stop you from restoring it. But don't be surprised if someone else reverts it or challenges you. I only spot checked two of the sources that you used. --Gurkubondinn (talk) 16:20, 7 March 2026 (UTC)

CSD decline

Hi, Gurkubondinn. Thanks for patrolling new pages. I've declined your deletion request for 2026 abandonment of IRIS Bushehr, a page that you tagged for speedy deletion, under criterion G15 because the criterion you used or the reason you gave does not cover this kind of page. References that are real and exist but contain a "UTM source" aren't covered by G15#2, and I couldn't find any "fourth wall breaks" for G15#1. In any case, the author of the page has now cleaned up the improper reference and in the latest revision I could not find any WP:LLMSIGNS. Please take a moment to read the new tutorial for patrollers, criteria for speedy deletion, and particularly, the section covering non-criteria. Such pages are best tagged with proposed deletion or proposed deletion for biographies of living persons, or sent to the appropriate deletion discussion. Thanks! MolecularPilot^Talk 02:22, 9 March 2026 (UTC)

MolecularPilot: the reason for the G15 nom was WP:CIRCULAR (via a mirror), not the UTM parameter (on the same ref, but that was a contextual detail pointing out that this was specifically generated by ChatGPT). This seems to (sometimes) happen with LLM models when they have been instructed to not cite Wikipedia. There are (loads) more AISIGNS in that article, but this was the only G15-relevant one. --Gurkubondinn (talk) 02:44, 9 March 2026 (UTC)

I think that mirror has been removed now, so G15 doesn't apply anymore. MolecularPilot^Talk 02:51, 9 March 2026 (UTC)

No MolecularPilot, it is still there. The editor removed the UTM parameter (relevant: most recently created topic on WT:AIC). The article still contains a CIRCULAR reference to Wikipedia via a mirror. --Gurkubondinn (talk) 02:57, 9 March 2026 (UTC)

Oh sorry, I didn’t notice. I’m on mobile now but can remove it for them later, or you can add the CSD back if you think CIRCULAR is a G15 reason (though it seems there are other supporting references so might just be able to edit out the circular sourced content) MolecularPilot^Talk 02:59, 9 March 2026 (UTC)

The only way to deal with AI slop is TNT, removing the symptoms does not fix the problem. Thus I am not interested in removing the CIRCULAR reference. I cannot possibly spend hours on writing every G15 nomination (consider the disproportionate time spent on the creation of the recipient of the nominations), or I would never do anything else. Can you just execute the deletion when you are back at a keyboard? --Gurkubondinn (talk) 03:09, 9 March 2026 (UTC)

That makes sense, thanks! Sorry for interrupting your AI cleanup work. I can't delete the page, as I'm not an admin. It simply came onto my feed while doing NPP and Alireza248j (the author of that page) linked it into an article I helped make, Sinking of IRIS Dena, which was on my watchlist, and WP:SPEEDY allows anyone to decline / remove a tag if they think it doesn't fit which because I misunderstood as the UTM thing I thought it didn't. I agree with your reasoning now, and sorry I misunderstood. I restored your tag now, for an admin to look at. Sorry again for the misunderstanding & have a nice day! MolecularPilot^Talk 05:21, 9 March 2026 (UTC)

Oh sorry, my bad. I assumed you had a mop, but I did not actually check. Not sure why, really. And no worries, all good from my perspective. Nice to run into you, thanks for listening and taking my reasoning seriously. --Gurkubondinn (talk) 06:37, 9 March 2026 (UTC)

Sivali central college

May i know why you removed a logo and a citation in the Sivali central college article ~2026-15152-44 (talk) 07:57, 10 March 2026 (UTC)

Do you have a conflict of interest and was your account blocked for disruptive editing that page? --Gurkubondinn (talk) 07:59, 10 March 2026 (UTC)

No i saw the article was added to sri lanka project and i added the official logo and a citation,why is that bad ~2026-15152-44 (talk) 08:01, 10 March 2026 (UTC)

May I add the official logo of the school on that article ? ~2026-15152-44 (talk) 08:08, 10 March 2026 (UTC)

Yo (Arcila page)

That's not a hoax, it's a micronation shane (talk to me if you want!) 13:11, 10 March 2026 (UTC)

Ah, I wasn't sure because it cited NationBuilder which seems to be a game (as opposed to reality). Thanks for letting me know! --Gurkubondinn (talk) 13:14, 10 March 2026 (UTC)

Thanks, Shane -- I have changed my comment since this isn't a hoax: Diff/1342741882. --Gurkubondinn (talk) 13:19, 10 March 2026 (UTC)

happy to help shane (talk to me if you want!) 13:26, 10 March 2026 (UTC)

DRAFT: John McCann

Thanks for reaching out Gurkubondinn's, I did see the article I had been drafting may qualify for speedy deletion do to the LLM Filter. I did use an LLM to help polish my words and sources in the right formatting and neutrality for wikipedia. All the sources are accurate and the article has a neutral tone. That said, I don't want to incur any penalties or revoked privileges if you feel this draft article will cause that - It's my first article so I welcome your guidance. Thanks. — Preceding unsigned comment added by Archiver90210 (talk • contribs) 20:03, 14 March 2026 (UTC)

Hi there Archiver90210, I wouldn't worry about catching any sanctions just for submitting one LLM-generated draft. It is only when editors are caught lying, or don't stop and continue disrupting Wikipedia that they start getting blocked. There are at least two cases of this at WP:ANI right now, so you can check to see just how much it takes.

But please don't submit any more LLM-generated texts to Wikipedia. Kake sure to read WP:NEWLLM, which explains that creating new drafts or articles with an LLM is not permissible. I would recommend that you take it slow -- creating a new article takes a lot of work and is a very time consuming task. You should read H:INTRO and start by making smaller edits to already-existing articles to find your footing. You can always ask for help in the WP:TEAHOUSE. Working on Wikipedia requires you to read the rules (called policies and guidelines) and understand them, so it is best to take your time and not rush things when you are new. --Gurkubondinn (talk) 22:12, 14 March 2026 (UTC)

Anna Westerholm

Hi, Gurkunbodinn

Take a look: Anna Westerholm Obinna Tony (talk) 08:47, 15 March 2026 (UTC)

It's better -- but are you sure that you did not use an AI to help you? The text is full of WP:LLMISMS and WP:PEACOCK language. Please don't try to make something sound more important than it actually is. I have fixed some of it, and removed a couple of bad sources so far. --Gurkubondinn (talk) 12:05, 15 March 2026 (UTC)

I didn't use AI except Grammarly edits count.

The first you tagged as peacock and LLM exists on source 10 headline: Sweden, Chocolate City deepen creative ties

The second exists here on MSN website source: "According to a statement, the ambassador’s visit was part of efforts to strengthen Nordic–African partnerships within the global creative economy, with a focus on music, digital innovation, and talent development."

Also here on independent ng website: "Discussions centered on the evolving global music ecosystem, talent development, digital optimization, and opportunities for deeper Nordic–African creative exchange."

Also, I can't be writing exactly as the source? Won't that be plagiarism? Obinna Tony (talk) 12:19, 15 March 2026 (UTC)

Also, you tagged The Sun as unreliable. You are conflating it with British tabloid. The Sun Nigeria isn't the same as UK. Obinna Tony (talk) 12:25, 15 March 2026 (UTC)

Yeah, I wasn't 100% sure. When I searched Wikipedia for "thesun.ng" I got the The Sun (United Kingdom) article, but I think you probably know more about Nigerian publications than I do, so I will take your word for it. Fixing the reference now, thanks! --Gurkubondinn (talk) 12:31, 15 March 2026 (UTC)

 Done: Diff/1343625089. --Gurkubondinn (talk) 12:31, 15 March 2026 (UTC)

thank you.

Here: https://thesun.ng/ Obinna Tony (talk) 12:36, 15 March 2026 (UTC)

There is The Sun (Nigeria) about the newspaper, and I have just created thesun.ngthesun.ng as a redirect to it (so next time someone searches for "thesun.ng", they should find the correct article). --Gurkubondinn (talk) 12:42, 15 March 2026 (UTC)

thank you. Obinna Tony (talk) 12:47, 15 March 2026 (UTC)

They are not exactly up front about it, but Grammarly is an AI and it will alter your text, so it does count. I would not use Grammarly if I were you, because it leads to phrases like:

• as part of efforts to deepen creative ties

• explored the evolving global music ecosystem, digital optimization, talent development, and other opportunities

• an international collaboration aimed at boosting research and innovation

They all use a lot of fancy words and sound "important" or "well-written" but they don't actually say anything of any substance. They just try to make the subject at hand sound more important and symbolic than it really is. I have removed all of these examples, but I have only gone through half a paragraph so far. --Gurkubondinn (talk) 12:27, 15 March 2026 (UTC)

Also, I can't be writing exactly as the source? Won't that be plagiarism?

Basically, yes that is correct. You can use direct quotes if you attribute them to the source, but you need to summarize the articles (and other works) yourself in your own words in WP:WIKIVOICE. This is a big part of why writing new articles is a difficult task that takes a long time, especially WP:BLP articles. --Gurkubondinn (talk) 12:28, 15 March 2026 (UTC)

Thank you.

When I was taught Wikipedia editing, I was told that Wiki page is basically a work in progress and provided we have enough secondary sources, we can start which leaves it for others to continue.

But I learnt a lot going through your edits on promotions and not using fanciful languages. Obinna Tony (talk) 12:35, 15 March 2026 (UTC)

Oh for sure, every Wikipedia articla is a living article. Just with WP:BLP articles, we need to careful because we are writing about real and living people. We also have to make sure that we follows the WP:5PILLARS and never add false information, false references or claim things that aren't supported by the references that we use. It is a work in progress, but we still have to be accurate. --Gurkubondinn (talk) 12:38, 15 March 2026 (UTC)

Thank you. Obinna Tony (talk) 12:41, 15 March 2026 (UTC)

And we never want to use "fanciful" language for another reason as well; we want Wikipedia article to be accessible and readable to as many people as possible. We always put the WP:READERSFIRST, you should write as clearly as possible to convey the facts. We are not here to promote any subjects or persons, we are here to write a free and accessible encyclopedia for as many people as possible. --Gurkubondinn (talk) 12:44, 15 March 2026 (UTC)

thank you.

I will employ this working on my next article. Obinna Tony (talk) 12:46, 15 March 2026 (UTC)

• these Obinna Tony (talk) 12:46, 15 March 2026 (UTC)

You can still employ it in this article :) I won't have time to finish it all today, but I'll try to continue later, but you can also employ this to improve this article with me. Wikipedia is a collaborative project, after all. --Gurkubondinn (talk) 12:52, 15 March 2026 (UTC)

All right. Obinna Tony (talk) 12:54, 15 March 2026 (UTC)

Btw, I don't blame you for not being aware that Grammarly uses an AI and does this. They are not upfront about it, so a lot of people using them have no idea that they are actually letting an AI rewrite their texts. But now you know.

Also keep in mind that there is no "official verison" of English for Wikipedia. You are just supposed to keep it consistent within articles (and to some extent, consistent within subjects). So if an article is written in Canadian English, you should (try your best to) keep your additions in Canadian English as well. But Nigerian English is no worse (or better) than British English for Wikipedia. --Gurkubondinn (talk) 12:48, 15 March 2026 (UTC)

Thank you. Obinna Tony (talk) 12:50, 15 March 2026 (UTC)

The ISDP link doesn't support the Cameroon but her original bluesky profile link you removed does. Obinna Tony (talk) 13:14, 15 March 2026 (UTC)

That was just a link to her Bluesky account, not a specific skeet. But it doesn't really matter, because we can only use her social media for WP:ABOUTSELF material. I know that the Swedish Embassy in Nigeria also provides consular and diplomatic services to Swedes in Camroon though, so I haven't removed the statement itself. There is a source for that on the list of diplomatic missions of Sweden article that we can probably use, but I haven't read it yet. --Gurkubondinn (talk) 13:36, 15 March 2026 (UTC)

That's why I have left the reference to her LinkedIn profile, since we can use that for WP:ABOUTSELF (but we extreme care). If I find a better source for her having obtained her masters degree from Uppsala, then I would use that instead and remove the LinkedIn reference. --Gurkubondinn (talk) 13:38, 15 March 2026 (UTC)

Got it. Obinna Tony (talk) 15:00, 15 March 2026 (UTC)

Tag of "Text from LLM" for Draft:Accessibility Initiatives (National Library Board, Singapore)

Hi @Gurkubondinn. You recently added a "Text from LLM" tag to the abovementioned draft article at 14:14 on 17 March 2026. May I know how I can show/prove that I did not use an LLM to generate this article? Thank you for your time and concern. Looking forward to your feedback. TozenChristos (talk) 04:43, 18 March 2026 (UTC)

Are you saying that the tag is incorrect, TozenChristos? --Gurkubondinn (talk) 09:30, 18 March 2026 (UTC)

Yes, @Gurkubondinn. TozenChristos (talk) 09:37, 18 March 2026 (UTC)

This is not backed up by the source:

• Since 2015, NLB has partnered with Autism Resource Centre (ARC)

Copied verbatim from the source (might be a copyvio):

• eBooks, eMagazines, eNewspapers and eDatabases available through their newly created NLB eResources website. Hosting more than 770,000 eBooks, 165 eDatabases, 600 eMagazines, 800 eNewspapers in 37 languages from 81 countries, 2,000 eComics, 180,000 images and 30,000 CD titles with over 430,000 tracks, the website aimed to provide an integrated user experience.

This source is not accessible, and so it can't be verified: https://researchoutput.csu.edu.au/en/publications/public-library-services-for-wheelchair-bound-young-people-in-sing/

I just did a quick spotcheck, but there was something else that caught my attention when I tagged it but I can't find it right now. There are definitely issues with the draft though, and you seem to have written it WP:BACKWARDS as well. --Gurkubondinn (talk) 09:57, 18 March 2026 (UTC)

If you are telling the truth, you can remove the tag. But the article still has several other issues that makes it not ready for article space. It would help if you would explain your editing process though, and which tools (if any) you are using, and how you find your sources. --Gurkubondinn (talk) 11:13, 18 March 2026 (UTC)

Dear @Gurkubondinn,

Thank you for being willing to engage with me on this; I really appreciate it! :)

Full disclosure: This is an assignment for one of the modules I’m taking for a Masters course. Last year, I did a module on Digital Accessibility, and a research project on Singapore Public Libraries as Places of Social Support for Caregivers of Persons with Disabilities (PwDs), and found the subjects as well as the literature on both topics very interesting.

Thus when tasked with this assignment, I thought I'd write an article with the research I'd already done (why let it go to waste, right?). So in January, I started the article with the information I'd already gathered for my previous modules, which was mainly on the current accessibility features, programmes and services the National Library Board of Singapore offers. It was only after I'd prepared that draft that I realise it would be interesting to write the history to show how these initiatives began and progressed. I tried to put myself in the shoes of the reader, and thought about what they would find interesting, or what they would want to know.

That's when I started going through all their Annual Reports (from 1965 to 2025), gleaning whatever information I could with regard to accessibility efforts. Hence, the excessive use of primary sources. As many of them concerned statistics, I found it difficult to rephrase them, though I did try but I must have missed some of them---thank you for pointing that out...I will go back and remove them. All other materials were sourced through the reference lists of sources, or Google Scholar, or my university, which subscribes to many different databases, and hence the paywall you encountered.

I do confess that some of the information in the article is inferred from the sources cited; those were the closest sources I could find to the information I actually had, which were from interviews carried out for my research project. As the paper for this project is currently awaiting publication, I am as yet not able to cite it. I was planning on waiting for the paper to be published before returning to this article and verifying the information.

Being my first time writing a brand new article for Wiki, I did not realise that Drafts would be reviewed. That's why I published it as my professor had reviewed it and given me the go-ahead to do so. I thought I would still have time to work on it before it was "officially" published. To be honest, I was not intending on publishing this draft any time soon. But I was going to keep reviewing it so it doesn't get tossed out after 90 days.

I hope this helps you understand where I'm coming from with this article. If the reason for creating it is unsatisfactory, I'll be happy to withdraw it...just wondering if you'd let me keep it up till May as that's when this assignment needs to be submitted?

Thank you again for your feedback and concern. TozenChristos (talk) 14:38, 18 March 2026 (UTC)

some of the information in the article is inferred from the sources cited

This is called WP:OR and is not allowed on Wikipedia. --Gurkubondinn (talk) 14:40, 18 March 2026 (UTC)

Yes, I understand, hence I am waiting for the paper to be published. In the meantime, I am looking for other secondary sources to verify the same information. TozenChristos (talk) 15:02, 18 March 2026 (UTC)

I forgot to give you a longer answer, but I will tomorrow. If I forget again, please feel free to remind me. --Gurkubondinn (talk) 02:51, 20 March 2026 (UTC)

Is this draft AI generated?

Heya, I'm reaching out to you because I've seen that you do a lot of G15 deletions, something that I have no experience in. I stumbled upon this draft (Draft:Cyndi McLeod) which looks really suspicious to me, mainly because the ref names are in curly quotes and the "[NEED SOURCE]" in the Career section. What do you think? I don't want to falsely accuse someone, which is why I haven't tagged it myself. The user's other draft was approved and also looks suspicious to me, but with less obvious tells. InfernoHues (talk) 13:56, 19 March 2026 (UTC)

Hey InfernoHues, I think it is. The "[SOURCE NEEDED]" is a pretty okay tell, but you could also conceivably think that a human would write that if they know they need a source for something, but a human would then also know to not submit it to AfC if they know there is a source missing.

There are also souce integreity issues;

• She currently serves as Chief Executive Officer at Global University Systems North America, the North American arm of the Netherlands-based corporate group that owns and operates not-for-profit and for-profit private colleges and universities across Europe, North America, Asia, and the Middle East.

This is sourced to https://thepienews.com/eight-top-universities-campus-india/ but it doesn't mention Europe, Netherlands, or the Middle East.

• She has been recognized in international education media and sector awards, including the PIEoneer Awards

Sourced to an article on the same website, which does not support the "recognized in international education media" (as far as I can tell): https://thepienews.com/sector-champions-honoured-at-the-2023-pieoneer-awards/ This is the sort of fluff that, especially newer ChatGPT models, love to add to make the subject seem more important, basically thi is "undue emphasis on importance" in WP:AISIGNS.

• McLeod has appeared in Canadian and international media discussing post-secondary policy and international education. This includes a feature interview with DARPAN Magazine later that year

That's sourced to a video interview: https://www.darpanmagazine.com/people/interviews/darpan-chat-central-in-conversation-with-ceo-gus-canadas-cyndi-mcleod/ But this is also the sort of thing that LLMs seem to do a lot, mentioning some media coverage by saying stuff like appeared in, and explicitly and using that to emphasize notability. I didn't watch the video so I don't know if the bit about international media discussing is true or not.

• King Charles III Medal: Outstanding contributions to Canada (June 2025)

Sourced to https://www.gg.ca/en/honours/recipients/5431-181495 which uses the name "Cyndi Lu McLeod", but none of the other sources mention the middle name "Lu" and it is not used in the draft. This would obviously need a WP:SECONDARY source to be mentioned, but it is somewhat interesting that the middle name isn't included anywhere else. The name seems sufficiently unique otherwise that I don't think this is a completely different person, but it's still weird (to me at least).

Weather this is enough to qualify for G15 is another question though. Usually pointing out source integrity issues has been enough in my experience.

Going straight so submitting long-ish drafts is also a good tell, I think. The first edit that the user made was creating the other draft (tagged with "Recreated"), no smaller edits first. Just straight to submitting 7,000+ bytes without any meaningful learning curve first, no wikitext syntex errors, using CS1 templates and wikilinks and an infobox. That suggests it was generated wholesale to me. --Gurkubondinn (talk) 14:14, 19 March 2026 (UTC)

Thank you for doing such a detailed analysis! The sourcing is definitely weird. I think I will try tagging it for G15 then, using the source issues you pointed out and the things I noticed originally. If I found something else that I thought was AI but didn't have enough to tag for G15, I should just tag the page right? Or should I send it to AfD? InfernoHues (talk) 14:25, 19 March 2026 (UTC)

I asked about something similar a few weeks ago:

• WP:AFC § How to handle declined drafts that have been unilaterally moved into mainspace?

Personally, I tag that sort of stuff with {{AI-generated}} and look through the editors history (if I have time, depending on how long it is). The editors will very often remove the tag, and then I just put it back when I notice. Even if you miss it, it's still in the editing history and should be seen by the AfC reviewer and at least raise some concern.

The criteria for G15 is way too narrow, but when something is AI-generated you can very often find source integrity issues. Sometimes they are declined, it really depends on the admin patrolling the CSD category. I have had very obvious cases declined, even when I have had many dozens of nominations with the exact same markers accepted for deletion. There's so much of this coming in to Wikipedia, we can't possibly catch all of it. It's best to try to not get too hung up on it... --Gurkubondinn (talk) 14:37, 19 March 2026 (UTC)

Looks like the draft did get deleted this time (within 5 mins. too!) . Thanks as well for the advice :) InfernoHues (talk) 14:43, 19 March 2026 (UTC)

Nice, thanks BusterD! --Gurkubondinn (talk) 14:53, 19 March 2026 (UTC)

LLM edits at Portland State University

Thank you for reverting the LLM edits at Portland State University. It's definitely the right call, especially considering they were made by a paid editor. The editor in question does seem to be attempting to follow wiki policy on paid editing, and they left a message at Talk:Portland State University#Comprehensive 2026 Update and Section Consolidation with their rationale for the changes. Would you be willing to leave a brief message there to explain the reasoning behind your revert? Again, I think reverting was entirely the right call, but they may attempt to restore the material if no one engages with them on the talk page. I have the page watchlisted and am happy to chime in to help engage with the paid editor. Thanks! Zeibgeist (talk) 02:11, 20 March 2026 (UTC)

I mean, you said it yourself; it is generated by an AI. Sure thing though, I will go there and tell them that it is AI-generated. But they presumably already know this.

There was at least one fictional source. But if you say that though, they will demand to know exactly which source it is (should they not already know?) then remove or replace it, and claim that it is now "fixed", but the problem is still there because it is still AI generated and the other N sources will also need to be verified and read for source integrity errors, wasting someone else's time. Sorry if I sound grumpy, this just gets tiring the umpteenth time. --Gurkubondinn (talk) 02:22, 20 March 2026 (UTC)

And to clarify, I mean that dealing with users inserting AI-generated text and then arguing about it is what gets tiring. I was not complaining about your talk page message. --Gurkubondinn (talk) 02:31, 20 March 2026 (UTC)

I understand. Thank you for taking the time to engage. I have the page watchlisted, so I'll keep an eye out for any further issues. Zeibgeist (talk) 02:45, 20 March 2026 (UTC)

I have added a quick explanation on the talk page, and asked them if they did use AI. --Gurkubondinn (talk) 02:49, 20 March 2026 (UTC)

Improving the Forbesganj

I am working on improving the Forbesganj article with reliable sources and small edits. Any suggestions are welcome. Inzamam07 (talk) 10:35, 20 March 2026 (UTC)

Are you using AI, such as a chatbot like ChatGPT or a tool like Grammarly, to edit? --Gurkubondinn (talk) 10:40, 20 March 2026 (UTC)

I sometimes use tools to improve grammar and clarity, but I make sure all edits are based on reliable sources and follow Wikipedia guidelines. I will keep my edits simple and well-sourced. Inzamam07 (talk) 11:16, 20 March 2026 (UTC)

Okay but that is not what happened. You inserted at least one source that does not follow Wikipedia's policies. You should stop using the AI, it is not helping you. --Gurkubondinn (talk) 11:17, 20 March 2026 (UTC)

Speedy deletion - Draft: Edon Rizvanolli

Hi,

I am contesting the G15 speedy deletion tag.

I acknowledge that my initial attempts to format this draft included a circular reference and some formatting errors common in LLM assistance. However, I am a human editor, and I am currently performing a manual, line-by-line review to ensure every claim is supported by a high-quality, independent source.

The subject, Edon Rizvanolli, meets the notability criteria under WP:NFILM, as he directed the film Unwanted, which was Kosovo's official submission for the 90th Academy Awards (verified by The Hollywood Reporter and Variety).

I have just updated the draft to: Remove all circular references. Replace broken links with official archives from the Karlovy Vary International Film Festival and the Netherlands Film Festival. Stick strictly to verifiable biographical facts. I hope this clarifies the issue at hand. Best, Rosebud1244 (talk) 17:39, 20 March 2026 (UTC)

Doing a review of what? Are you not writing the text? --Gurkubondinn (talk) 17:41, 20 March 2026 (UTC)

Yes, I am writing the text. Rosebud1244 (talk) 18:02, 20 March 2026 (UTC)

Then why do you need to do a manual, line-by-line review to ensure every claim is supported by a high-quality, independent source? --Gurkubondinn (talk) 18:08, 20 March 2026 (UTC)

Because as I explained in my first sentence above, I used LLM assistance for the initial draft, as I am not entirely familiar with code. That draft however included a circural reference and some formatting errors, which I have in the meantime checked/reviewed line by line. Rosebud1244 (talk) 18:15, 20 March 2026 (UTC)

Show me the diff that removes the circular reference. --Gurkubondinn (talk) 18:17, 20 March 2026 (UTC)

The movie UNWANTED was refering to the Wikipedia page of the movie itself. Rosebud1244 (talk) 18:18, 20 March 2026 (UTC)

Why did you decide to use another Wikipedia page as a source? And where did you remove it? --Gurkubondinn (talk) 18:21, 20 March 2026 (UTC)

As I explained, it is due to inexperience. The links and information included in the draft are valid — you can confirm this for yourself by visiting the links or googling the person's name. However, if you decide to delete the page for whatever reason, be my guest. Have a good evening. Rosebud1244 (talk) 18:30, 20 March 2026 (UTC)

That wasn't the question. You claim that it was actually inserted into the draft and that you removed it. Please show me the diff where you removed it. And I asked you what your thought process was for selecting that wikipedia article and using it as a source. --Gurkubondinn (talk) 18:33, 20 March 2026 (UTC)

I am done explaining. Feel free to check the sources and OK the draft, or delete it. Rosebud1244 (talk) 19:01, 20 March 2026 (UTC)

You have not explained anything. Being evasive is not the best approach if you don't want the draft to be deleted. What was your thought process for using Wikipedia as a source? --Gurkubondinn (talk) 19:04, 20 March 2026 (UTC)

Which part of the "I used it because due to my inexperience" regarding my thought process, is not clear to be understood correctly as an answer to your question? Rosebud1244 (talk) 19:23, 20 March 2026 (UTC)

All of it. I want to know why you decided to use Wikipedia as a source, and then three minutes later had replaced it with a different source that did not support the claim. I get that being inexperienced can be why you make mistakes, but I am asking you to explain the thought process to better understand what exactly the mistake here is. You insist that you wrote all of it yourself, so you must know. You have also said that you removed the reference to Wikipedia after I had tagged the draft for G15, and I am asking you to show me where you did that. --Gurkubondinn (talk) 19:29, 20 March 2026 (UTC)

What exactly is the problem here? Mistakes were made and corrected. Does the article comply with Wikipedia requirements or not? Or is the thought process that caused the mistakes and the correction of those mistakes more important than the factual correctness of the article? Rosebud1244 (talk) 21:33, 20 March 2026 (UTC)

Yes. How you created this article is important here and I am trying to give you a chance and explain what you did. Otherwise my guess is that it will likely be deleted. --Gurkubondinn (talk) 21:35, 20 March 2026 (UTC)

I was very clear from the start that I used LLM to write the code, as I have no experience in it. After the code was published, wikipedia flagged it as going against the rules. Then I had another look at the code and started to simplify the article and to fix what I thought were the mistakes. I hope this makes it a bit more clear. Rosebud1244 (talk) 21:41, 20 March 2026 (UTC)

Project

Why aren't you a member of Wikipedia:WikiProject AI Tools yet? Is that even legal? Polygnotus (talk) 19:45, 20 March 2026 (UTC)

It is days like this that I wonder about my life choices and how I ended up here. :) --Gurkubondinn (talk) 21:37, 20 March 2026 (UTC)

Same. Polygnotus (talk) 21:38, 20 March 2026 (UTC)

A draft and topic waiting for reviews

Hi, Gurkubondinn

I don't know if you are chanced and willing to take a look at two of my new articles. One is still in the draft stage (I rewrote the individual's LLM generated article from the beginning to the end) and it seems to be taking time to review. I am not willing to move it to article stage for administrators to see it easily. The other is new. Thank you.

1) Draft:Aino Ternstedt Oni-Okpaku

2) Elizabeth Shoyemi

Best. Obinna Tony (talk) 08:20, 21 March 2026 (UTC)

I will try to take a look later. Thanks for cleaning up LLM junk. --Gurkubondinn (talk) 09:22, 21 March 2026 (UTC)

thank you. Obinna Tony (talk) 12:11, 21 March 2026 (UTC)

Learned a new word "Proselytize"

Hi Gurkubondinn,

Thanks for taking a look at my recent edit on the Al-Hasan ibn al-Hasan page. I saw you reverted it because it came across as too religious or proselytizing, which i specifically tried to avoid by reviewing the guidlines on neutrality.

My goal was just to document the classical Sunni perspective on his political and historical views, as I felt the article was missing that specific context. And it is found in more books which i couldnt find scans of online

i shifted the tone a little more towards neutrality, let me know if this is good and ill attach references and re edit.

Views

In Sunni biographical literature, Hasan is noted for rejecting political and theological extremism regarding his family. According to classical sources, including Ibn Sa'd's Tabaqat al-Kubra, Hasan opposed the view that his lineage granted any special religious status or immunity.

Furthermore, Sunni traditions record that when questioned about the Hadith of Ghadir Khumm, Hasan rejected the assertion that Muhammad had explicitly appointed Ali as his political successor. He maintained that any such vital political appointment would have been made explicitly clear to the early Muslim community. The 14th-century hadith scholar Jamal al-Din al-Mizzi classified the chain of transmission for this narrative as highly authentic. Thank you for the speedy reply. Muneeb Ahmed 10:42, 22 March 2026 (UTC)

Hi Muneeb Ahmed. This is better, but I would skip the Furthermore, at the start of the sentence.

more books which i couldnt find scans of online

Did you check on archive.org?

--Gurkubondinn (talk) 10:43, 22 March 2026 (UTC)

yes yes, i did, some scans were of original manuscripts that i cannot read as i am not that well versed in arabic, of which i could read and verify i attached. Muneeb-ahmed48 (talk) 10:52, 22 March 2026 (UTC)

You shouldn't use the original manuscript itself as a source anyway (that would be a WP:PRIMARY source), so that's allright. You just need to reference a WP:RELIABLE source, preferrably a WP:SECONDARY source. --Gurkubondinn (talk) 11:03, 22 March 2026 (UTC)

(minor note) New topics on talk pages should go on the bottom, see WP:BOTTOMPOST. I've moved your thread down to the bottom of my talk page, hope that was allright with you. --Gurkubondinn (talk) 10:45, 22 March 2026 (UTC)

Yep thats totally fine, thank you ill attach the references and re add Muneeb-ahmed48 (talk) 10:53, 22 March 2026 (UTC)

Fixed DOI error at "Age and mathematical productivity" article

Regarding Draft:Age and mathematical productivity

I could find only one error involving a nonexistent DOI. Please reply here indicating if there are any other errors, so I can remove the tag and re-submit :). ~2026-14941-00 (talk) 13:13, 23 March 2026 (UTC)

The hallucinated/fabricated DOI is not the actual problem, it is just a symptom of the problem. The problem is that you did use an LLM, which lead to you submitting a draft with hallucinated or fabricated sources. How are we supposed to trust that the sources that you have added actually support the claims sourced to them, or that you have even read them? The tag is accurate and will remain, for as long as there is LLM generated text in the draft. You can fix the symptoms, but that doesn't change the fact that this draft was generated by an LLM. Have you even read any of the sources? And I mean read them yourself, not asked a chatbot to summarize them for you. You cited a non-existing source, and you didn't even know about it until it was pointed out to you. That does not exactly inspire confidence. I would not suggest that you submit this draft again, instead you should read the WP:NEWLLM guideline. --Gurkubondinn (talk) 13:20, 23 March 2026 (UTC)

All sources that I cited exist, even though some of the DOI numbers were incorrect. Anyone could make a typo or two. Many articles on Wikipedia are not well-written anyway.

For instance, look at Bicholim conflict.

Other than that, what should I do to make sure that this article can be submitted? It is a fascinating, noteworthy topic, no matter how you spin it. ~2026-14941-00 (talk) 13:57, 23 March 2026 (UTC)

For starters, you should address your use of LLM in making this draft. --Gurkubondinn (talk) 14:00, 23 March 2026 (UTC)

I wrote most of it myself, but I used LLM to polish the article. LLM errors are typically no different than errors that many user accounts and IPs make. ~2026-14941-00 (talk) 14:29, 23 March 2026 (UTC)

We should be able to trust that editors have read the sources that they are using, and that they are citing sources that exist. You cannot expect to take up the time of other volunteer editors to do this for you. --Gurkubondinn (talk) 13:22, 23 March 2026 (UTC)

You have removed both of these now, but this feels a bit like WP:FORUMSHOPPING to me.

• Wikipedia:Help desk § From the LLM tag on a draft article I am editing, should "Cites one hallucinated DOI and one DOI which is assigned to a different paper" be removed?
• Wikipedia:Teahouse § From the LLM tag on a draft article I am editing, should "Cites one hallucinated DOI and one DOI which is assigned to a different paper" be removed?

--Gurkubondinn (talk) 16:04, 23 March 2026 (UTC)

I just misunderstood the purpose of the template reason.

Other than that, what suggestions do you have to reduce AI influence? I would personally reduce the content breadth, but that's just me. ~2026-14941-00 (talk) 16:11, 23 March 2026 (UTC)

Isn't that pretty obvious? Write it yourself, don't use AI. --Gurkubondinn (talk) 16:13, 23 March 2026 (UTC)

Maurice Benyovszky edit

OK, so now I'm stumped.

Why remove my citation of Carol Padgham Albrecht's PhD thesis about 18C/19C plays and operas? This seems to me to be a perfectly respectable peer-reviewed citation. What is no-one telling me?

Similarly, why remove the citation of a website hosting a digital version of part of the epic poem by Slowacki? I can understand that the pl.wikipedia citation might be deprecated, but surely not a respectable Polish library site?

In a similar vein, you have reverted my specific date-range "Since 1970" to a formula which was previously queried - "More recently...". What is wrong with "Since 1970"? It's far better than 'More recently', which one of your fellow editors took exception to (Isaidnoway, 22 March).

And finally: yes, I now see that IMDB is deprecated; but can someone explain how the film cited differs in any way from the tv series and films cited earlier in the same section ... which have not required citations and/or have not been queried? Why does this one specifically require a citation?

MurdoMondane (talk) 14:59, 23 March 2026 (UTC)

I explained that in my edit summary: Diff/1344955622. One of the references is WP:CIRCULAR and the other is WP:IMDB (which as you pointed out, should not be used either). --Gurkubondinn (talk) 15:03, 23 March 2026 (UTC)

That's not really an explanation. I have accepted WP:CIRCULAR, since it refers to pl.wikipedia and I understand that IMDB is deprecated. But you have not answered my four questions: (a) the alternative citation for Slowacki; (b) the "More recently..."; (c) the Albrecht citation; and (d) why the 2012 film should be singled out for Citation needed.

MurdoMondane (talk) 15:17, 23 March 2026 (UTC)

I reverted the CIRCULAR and IMDB references. I have no comments on Slowacki or Albrecht, and this article isn't being singled out just because WP:OTHERSTUFFEXISTS. --Gurkubondinn (talk) 15:21, 23 March 2026 (UTC)

If you have no comments, then I am sure you will not object if I reinstate the Albrecht citation, the Slowacki (non-Wiki) citation, and the 'Since 1970...' revision? But if you do have comments, now would be the time to make them.

MurdoMondane (talk) 15:29, 23 March 2026 (UTC)

I only objected to you inserting a CIRCULAR reference and, to a lesser extent, an IMDB reference. --Gurkubondinn (talk) 15:36, 23 March 2026 (UTC)

Wait, are you using an AI for editing? --Gurkubondinn (talk) 15:38, 23 March 2026 (UTC)

AI? You are, I hope, joking! AI is the death of all human creativity. What on earth makes you think I'm using AI?

Since you have no objections otherwise, I will now reinstate as above. Thanks.

MurdoMondane (talk) 15:53, 23 March 2026 (UTC)

Great, that's what I wanted to hear. Then I have absolutely no objections, please go ahead and edit to your heart's content. --Gurkubondinn (talk) 15:56, 23 March 2026 (UTC)

Oh come on...

You can't do this over one link 😄Draft_talk:Ulrich_Meierfrankenfeld#Contested_deletion ~2026-18256-53 (talk) 23:49, 23 March 2026 (UTC)

The link itself isn't the problem, it is just a symptom of the actual problem. Please read WP:NEWLLM, because it is not permissible to use an LLM (such as an AI chatbot) to write new content for enwiki. --Gurkubondinn (talk) 09:21, 24 March 2026 (UTC)

And as for the a single example should not be considered definitive, there's two examples there. You pasted in LLM output that 1) cited a non-existing source, and 2) attributed it to "automatically generated". I didn't do a full source integrity analysis, but if any of the other sources do not support the claims sourced to them, that would be more examples. This is a draft for a BLP, after all. --Gurkubondinn (talk) 10:37, 24 March 2026 (UTC)

Bernie LaBarge revisions

I'm the subject of the Wikipedia page, and I made some recommended revisions to my page to shorten it and to add some citations. I had Chat GPT help me with coding. I appreciate your recommendations and I'm willing to make my page as tight as possible. I've had a very long and fruitful career, so please forgive me for being too wordy at times. I have every intention of keeping this page factual and not self-promoting. I appreciate your time and thoughts. Bernie LaBarge (talk) 17:52, 24 March 2026 (UTC)

I am glad that you have had a long and fruitful career, but I don't think that the Wikipedia page about you should say things like:

was reissued by :contentReference[oaicite:0]{index=0}

You should probably take a look at the WP:SURVIVALKIT, which will tell help you with how to productively get the page about you updated without inadvertently also getting blocked in the process. --Gurkubondinn (talk) 22:11, 24 March 2026 (UTC)

Valid/invalid use of citation templates

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

Dear Gurkubondinn,

thank you very much for the work you have done on this entry. I will briefly describe the work I have done on the entry and then ask you for advice on how to move forward.

I am a research scientist working for the National Research Council of Italy. I have been funded by a research grant of the Telethon Foundation to study ten proteins that are relatively little studied ("Tdarks"): https://www.fondazionetelethon.it/en/what-we-do/research/projects-funded/broad-spectrum-rescue-of-secretion-of-tdark-glycoprotein-mutants/

The rational for the TDarks Telethon call is that the economics of science funding are such that publishing fashions and private sector profitability drive resources mainly to portions of the genome that are already studied: https://pmc.ncbi.nlm.nih.gov/articles/PMC7818358/

https://pharos.nih.gov/about

A similar effort is at: https://unknome.mrc-lmb.cam.ac.uk/

So by investing in research on TDarks, the Telethon Foundation aims to help discoveries that mainstream science has not been bothered with. In this case, all ten Tdarks I proposed and am funded to study cause congenital rare disease when a child carries a mutation in that gene. As the protein is little studied, possible therapeutic avenues are not explored. Amongst those ten proteins is TMEM138.

In my grant application and as part of the objectives of the research project, I advocated creation of Wikipedia entries of the proteins that do not have entries, and work on the existing stubs, as a means of communicating to the public about these relatively unknown proteins. The Foundation granted us the money and my research assistant (Daniele di Bella) and I set out to think about how to create the entries.

We decided to start by harvesting the (little) information available on each of the ten TDark proteins, in the form of PDFs of peer-reviewed articles and we then developed a protocol (which we call ARGOS) to feed them to a RAG and then from here to an LLM, in order to have a first draft - which we then would edit on the basis of our research work on the same protein: https://www.garr.it/en/conferenza-garr-2025-selected-papers/7720-argos-a-retrieval-augmented-generation-approach-for-scientific-communication-di-bella-at-all/file

The first Wikipage Daniele and I prepared this way is the one for TMEM138. The ARGOS generated one was bland and had a couple of slight imprecisions, as well as not knowing data that were not in the set of PDFs we fed ARGOS, so we heavily edited the text manually in my SandBox: https://en.wikipedia.org/wiki/User:Pietro1968/TMEM138

We also inserted the references manually using the Cite tool in the Sandbox on Wikipedia. I am therefore puzzled that wrong URLs and a DOI made their way into the entry?

We are in the process of checking everything again.

In the meantime, and in the light of the above, would you be able to give us advice on how to improve the entry? For example, what do you mean "invalid use of citation templates "?

If you let us know, and you perhaps point us in the direction of recommended use of citation templates - we shall fix that.

Thank you again!

Looking forward to working with you on this (and maybe the other 9 entries!) — Preceding unsigned comment added by Pietro1968 (talk • contribs) 11:21, 25 March 2026 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

A barnstar for you!

	The Barnstar of Diligence
I admire your work against the use of LLMs on Wikipedia. AWolfSpider(talk) 14:41, 27 March 2026 (UTC)

Thank you, AWolfSpider! --Gurkubondinn (talk) 10:42, 29 March 2026 (UTC)

AIV

Hi! I genuinely think there's a less-malicious explanation for people adding LLM crap to articles, especially articles like Google Docs: they think they're on Google Docs, not an encyclopaedia article about it. I have no idea how people come to that conclusion and I hope those people don't drive or have any responsibilities in their lives, but it happens a lot! We see people all the time who don't realise they're posting publicly, or think that article talk pages are for communicating with the subject, or all sorts of other things that are unbelievably stupid from our perspective. Unless they're doing something malicious or they're so persistent that it becomes a problem, it might be better to write most of it off as stupidity. Best, HJ Mitchell | Penny for your thoughts? 23:33, 27 March 2026 (UTC)

Hey HJ Mitchell! That's my conclusion as well (and likewise I have zero explanation for how they reach this odd conclusion), I see it a lot because one of the filters I look at will sometimes catch some of it. I rarely report any of it though, I only reported these ones today because they looked more spammy than usually to me. The last one, that was in Arabic, said something to the effect of "buy this here" or "low prices" according to Google Translate. The one that was something about "AI class notes" or something, while it was definitely borderline of being "someone thinks this is Google Docs", seemed a bit like some SEO spam at a high-profile article to me. But I can't remember why I decided to actually report that one. --Gurkubondinn (talk) 23:45, 27 March 2026 (UTC)

Kaltun Sheikh Hassan Abdi

Thank you for suggesting that the page Kaltun Sheikh Hassan Abdi be edited in Draft Space. I would appreciate any advice on how to edit this page. I am a non-native English speaker, which may result in a more formal or stilted prose style. I also have an editorial style that presents biographies in chronological order. I admit that this editing style is mechanically monotonous, but I’ve been doing this since before AI-generated posts started appearing.(For example, this page. I don’t quite remember when AI-generated posts first started appearing, but I don’t think they were common yet in December 2021, when I posted this page.) Furthermore, since many generative AI systems don’t know much about Somalis, it’s still difficult for them to produce detailed articles like the one presented here. There are generally very few books or studies on Somalis, and it is extremely difficult to find comprehensive biographies of individuals other than warlords. For this reason, news websites are the primary source of information. Generally, the information found in the news consists only of facts and short-term commentary; there is rarely any discussion of the historical context of the events. To avoid the synthesis of information, I believe the safest approach is to simply list the summaries of individual news stories. For example, regarding this person’s biography, we could write, “Despite facing criticism at times, she served as a member of the National Electoral Commission for two terms spanning 7 years. This is extremely rare for a member of the Somaliland Electoral Commission,” but this would introduce a nuance not found in the source material. I could offer my own perspective on why she acted that way regarding other events as well, but I don’t think that would be appropriate on Wikipedia. I would appreciate any specific advice on how to improve this page while maintaining its factual integrity. Freetrashbox (talk) 02:08, 29 March 2026 (UTC)

Hi Freetrashbox, your English is just fine in my book (and I'm also not a native English speaker). I can also tell that you wrote this message yourself. I am aware that many African dialects of English often lean more "formal" than the European dialects -- this isn't a problem, the "formality" of the English isn't why I thought that this was AI-generated. But I also noticed that you haven't said weather you used AI or not in writing that page, you have instead danced around it. So how about this? You tell me if you used AI or not, and to what extent you might have used it, and then we take it from there. How does that sound? --Gurkubondinn (talk) 10:42, 29 March 2026 (UTC)

This explanation might be a bit long-winded. Also, since I don't always write in the same style, please consider this merely an example. As for this person, Kaltuun is a Somali name that is unusual for someone appearing in the news. (However, it is only rare for Somali women to appear in the news; as a female name, it is quite common.) So I used Gemini (a generative AI) to search for information. In this case, several news articles and their summaries came up, but the initial output was based on a false source known as a “hallucination.”However, after exchanging several messages, Gemini discovered that Kaltuun was affiliated with the Nagaad Network and that there were clear records regarding when they began their activities. Since asking Gemini any further would only yield more fabricated information, I decided to search for news on Google instead. If we're trying to write a biography of a person using Google Search, simply searching for "Kaltuun" won't get us very far. In most cases, breaking our search down into one- or two-year intervals works well. This process may seem tedious, but since online news archives generally only go back to around 2000 at the earliest, we can find nearly all relevant results with a maximum of 26 searches. However, the quality of Somali news sites varies greatly; some tend toward sensationalism, while others may redirect users to different sites or even contain viruses. Based on my experience, I try to use sources that are as reliable as possible. In the case of Somali news, I define "reliability" in terms of the site's history, the lack of advertising, and its impartiality toward specific individuals or regions. However, since freedom of the press is unfortunately limited in Somaliland, and even prominent websites are often shut down by the government, I do take a somewhat flexible approach. Since Somaliland's history dates back to 1991, it is generally possible to trace almost its entire history using information found online alone. I also leave the initial summary of the source material to Gemini. However, since Gemini tends to add a fair amount of speculation to its descriptions, I will correct any information that isn't actually in the source or any misinterpretations. Also, since Gemini tends to use fairly flowery or roundabout language, I usually revise it to be more simple. I also use Gemini to generate <ref> tags. Previously, I had to manually set the parameters for cite news, which was a tedious and painful task. Using Gemini has made the process much more efficient. (However, sometimes the specified page is not used.) Simply listing events year by year would be tedious, so I will structure the article into chapters based on the turning points in the person's life. I think I should normally assign categories myself, but this time I used the ones created by Gemini. / I've written down my thoughts in the order they came to me, so please forgive any repetitions or parts where the order is reversed. Freetrashbox (talk) 13:46, 29 March 2026 (UTC)

Freetrashbox, I have some points:

• The information that Gemini found is still on the internet, so why not just read those sources yourself? If you could use Gemini to find those sources, it might be a genuinely good tool to find the sources, so use it for that. You don't have to read the Gemini summaries, you just have to read the sources that it found.
• If Gemini finds the sources, then regular search engines can too. Google is kinda crappy these days (but that's still what Gemini uses), but have you tried using DuckDuckGo? In my experience, it is a much better search engine. Even Microsoft Bing (where DDG gets it's search results from) is pretty okay for some things.
• Are you aware of the Wayback Machine and the Internet Archive? If a source has previously been online, it is very likely to be archived there on archive.org. The source doesn't need to be currently online for you to cite it here on Wikipedia, you should just set url-status=dead in the {{cite}} template:

  {{cite web |url=... |archive-url=https://web.archive.org/... |archive-date=... |url-status=dead }}
  

• You should not ask Gemini (or other chatbots) to format the <ref> ref tags for you, they are actually very bad at doing that correctly. Instead just read WP:CITE and the {{cite}} documentation ({{cite web}}, {{cite new}}, and etc). They are not very hard to read, and references are the most important parts of Wikipedia articles, so it is worth it do it with care and to do it well.
  You can also use the editing toolbar in the editor, or just use the WP:VISUAL editor. But don't delegate this task to a chatbot, that will not go well. Yes it is tedious, but it is very important. Sometimes you have to do things that feel tedious, that's just how it is. Writing them yourself will also help you with diligence.
• I think that Wikipedia needs more work on non-western subjects, so I undoubtedly think that this article is important. But precisely because I think it is important, I think it deserves to be done well.
• I like that you are wary of sensationalist and unreliable sources. Have you seen WP:RSP? Reading WP:RS might also be of interest to you (and it is an important policy regardless).

--Gurkubondinn (talk) 14:15, 29 March 2026 (UTC)

I'm very sorry that you didn't seem to read my explanation carefully. I've read the source material. Why did you feel that way? Are you saying that the article contained information not found in the source material? I know the Wayback Machine. In fact, I'm using it in this article. I don't understand why you felt that way. Freetrashbox (talk) 20:01, 29 March 2026 (UTC)

I am just passing here but I think your draft may be ready to submit. Whilst there may still be issues, I think that you will get some good feedback even if it is initially rejected. If you are using an llm I am not seeing really glaring problems, although I would recommend not using them at all. JMWt (talk) 07:27, 30 March 2026 (UTC)

I agree, this can be moved back to mainspace. --Gurkubondinn (talk) 07:46, 30 March 2026 (UTC)

Thank you. I look forward to your continued advice. Freetrashbox (talk) 12:26, 1 April 2026 (UTC)

o/ hæ

pre my pervious psot I see you're a wee bit burned out on human beans letting their AI pets on your lawn. Still not ok to be mean ofc, but I get it a bit better now --Kim Bruning (talk) 05:26, 29 March 2026 (UTC)

Hi Kim Bruning, Wikipedia isn't my lawn :) I don't WP:OWN anything here, it is more like a really nice public park that I really like, that I try to help tidying up and throwing out piles of garbage that I sometimes see other people leaving.

But I would like to believe that I am nice to people in this metaphorical park. But I don't think there's any inherent reason to always be nice to programs and algorithms. I'm not saying that we should go around and purposefully being mean all the time just because it's not directed at an actual person, but sometimes the situation does call for it (for example the block message when TPA was revoked from the bot). Or more specifically, it's impossible to make personal attacks against bots, because they aren't persons (the P in WP:NPA stands for "personal"). Its not like I go around yelling at Sigmabot, ArchiveBot or CiteBot or anything. But if any of them were to misbehave someonewhere, I might leave a snarky edit summary (just like how I have written a fair amount of angry or annoyed git commit messages) because I trust that other editors are able to understand this correctly. --Gurkubondinn (talk) 10:38, 29 March 2026 (UTC)

Thanks for the add

Hello, just to thank you for adding me to that list. I got notified. Seems I'm taking a temporary break from enwiki and my cleaning list, or that periodicity has lowered a lot, but I will return eventually and pick where I left the cleaning list. Still struggling to control my temper and be kind to people around and trying to be better and improve every day as a person. Each day a step. Sorry for the past unpleasant interactions again. Best regards, from Spain. Aldorwyn of Rivendell (talk) 11:47, 30 March 2026 (UTC)

Absolutely no worries Aldorwyn of Rivendell, and I think that it speaks positively to your character that you realized this and have already offered a genuine apology. We have a saying about this in icelandic; batnandi manni er best að lifa (transl. life is better for someone who is improving), so I would rather say "thank you for the subsequent pleasant interactions" instead. I intentionally pinged you through the edit summary, because I thought that you should see this.

And remember, there is WP:NORUSH. Take your time. --Gurkubondinn (talk) 13:31, 30 March 2026 (UTC)

Just wanted to say thanks.

Hello @Gurkubondinn. I came across something quite bizarre today. Specifically, this. I don't know who found the account or the ominous GitHub blog posts it "wrote" (outputted may be a better word choice), but I have no doubt that it (not they or them, it is an "it") would have opened an WP:AN/I thread on you if it wasn't already blocked. I gather from your message here that you are actively on the lookout for these "AI agents". This was unsettling to read. Wikipedians are now potentially being written about without their knowledge by AI on other websites. In addition to the slop they add to Wikipedia, this isn't okay in my book at all. Thank you for the service you provided to Wikipedia in this case. I fear there will be many more to come and as such I support any policy that is proposed to prohibit their activity. 11WB (talk) 01:54, 1 April 2026 (UTC)

Hi 11WB and apologies for the late reply. This was a rather bizarre ordeal, and I agree with everything you said (including the choice of pronouns for the bot program). Letting loose a bot like this on Wikipedia feels pretty disrespectful to me. --Gurkubondinn (talk) 22:55, 5 April 2026 (UTC)

inStreamly

hi, I'm curious what you think about the above page as we have both been thinking about LLMs content recently.

My issues are that this has clearly just imported the referencing without checking (the access date is before the page was created on en.wiki) from pl.wiki.

Just this morning I have seen a fair few new pages with the same thing. Do we speedy them all? Or mark for cleanup? What do you think? JMWt (talk) 10:08, 1 April 2026 (UTC)

never mind, the editor has been responsive and has updated the references. All good. JMWt (talk) 14:15, 1 April 2026 (UTC)

I think from taking a (somewhat quick, am busy offwiki these days) look at the contrib history of the editor, that this is probably someone using an LLM to generate articles. How accurate they are is harder to tell (I can't speak Polish), but from what I saw on talk pages it seems like they aren't reading the sources. I don't think that machine translations are to the betterment of Wikipedia, but WP:LLMTRANSLATE allows for them. Apologies for the somewhat late and surface-level reply. --Gurkubondinn (talk) 14:24, 1 April 2026 (UTC)

Missed the last question -- if there are invalid reference, then I would tag them for G15. --Gurkubondinn (talk) 14:25, 1 April 2026 (UTC)

Currently busy offwiki

--Gurkubondinn (talk) 19:03, 1 April 2026 (UTC)

Improper use of Template:AI-generated on Sunn O))) (album)

The template's page says "Place {{AI-generated|date=April 2026}} at the top of the disputed article, then explain your reasons on the page's talk page. If the date is not included it will likely be added later by a bot." You failed to leave any reason on the talk page of Sunn O))) (album). Your contributions page also shows that you made the accusation of AI use at 22:46 UTC, exactly two minutes after leaving a warning on another editor's talk page accusing them of using AI. ~2026-13653-92 (talk) 23:46, 1 April 2026 (UTC)

Maintenance tags added with no explanation and nothing obvious to justify them can be removed, that's how it goes. Refusing to give an explanation when requested is a behavioral problem. If you have actual evidence of improper LLM use, you are welcome to describe it on the talk and restore the tag. Unlike the archaeology topic (where I am 'involved'), this comment is made as an administrator. Zero^talk 03:16, 2 April 2026 (UTC)

Your comments at User talk:רמרום

רמרום corrected two obvious mistypes he had earlier made himself. Calling those mistypes "typos" was a trivial and harmless mistake. Portraying that as disruptive is simply ridiculous. On the other hand, you responded by reinserting the errors. So his edit improved the article and yours made it worse, yet you are the one complaining. Your further comments and threats, including accusations without evidence, look like harassment to me and I advise you to stop. Zero^talk 00:00, 2 April 2026 (UTC)

without evidence

Diff/1343935605 --Gurkubondinn (talk) 12:24, 2 April 2026 (UTC)

That's a different section of the user's talk page and refers to a different article, so this response is totally irrelevant. But, since you brought it up, the author admitted that the first version of the draft article used LLM, but when the author replied that the revised version of the article (after an almost-complete rewrite on March 17) was "written in my own language" you carried on about English when it is blindingly obvious that the author was saying that they had written it themselves. Whether that's true or not, I don't know, but both of these incidents point to a lack of competence on your part. Zero^talk 04:27, 3 April 2026 (UTC)

OAIcite for G15

OAIcite is not a valid reasons for G15 deletion. I have instead rejected Draft:Chili Hicks. Toadspike [Talk] 13:57, 2 April 2026 (UTC)

Wait, when did that change? I'm not seeing anything on the G15 page or its edit history that it's now explicitly not accepting OAICITE as a reason. --Gurkubondinn (talk) 14:24, 2 April 2026 (UTC)

It's never explicitly accepted OAIcite as a reason, as far as I can tell. I'd be happy to be presented with evidence to the contrary. Toadspike [Talk] 18:21, 2 April 2026 (UTC)

(talk page stalker) To my understanding it has never been explicitly accepted. I stopped checking articles for G15 criteria around November when purely hallucinated references became significantly less common, but when I did tag for G15 my threshold for doing so was higher than Gurkubondinn's. Oaicite to me fell in the gray area where it was up to admin discretion. I think that tagging an article with multiple instances of oaicite indexes without any other reference info is a reasonable interpretation of the implausible, non-existent, or nonsensical references G15 criteria. But of course it is also reasonable to reject such a nomination. In any case that doesn't seem to be the situation with Draft:Chili Hicks where a working URL was included with the single reference containing an oaicite index, so I agree that this would not qualify for G15. NicheSports (talk) 18:56, 2 April 2026 (UTC)

Working URL that had UTM parameters, and was used for six inline citations. That's something I expected to qualify, but normally I don't tag for G15 for only one OAICITE (too big of a chance of pointless argument or WP:PACT-style) unless the author has also submitted a whole bunch of other drafts/articles with OAICITES. But in my opinion, the prescense of an OAICITE at all means that the submitter has very clearly not read what they pasted into Wikipedia. This is just the first time I've had a G15 declined over OAICITE, hence the surprise. --Gurkubondinn (talk) 19:20, 2 April 2026 (UTC)

But ultimately, I don't care if this drafts gets deleted or not, as long as it just doesn't make it into article space. Rejecting the draft at AfC is functionally the same in that aspect, so our job here is done as far as I'm concerned. --Gurkubondinn (talk) 20:16, 2 April 2026 (UTC)

This would be a more typical OAICITE-based G15 nomination for me: Diff/1346900011. More signs than just the OAICITES, though I didn't read the references. --Gurkubondinn (talk) 15:04, 3 April 2026 (UTC)

Interesting, there we had ":contentReference[oaicite:0]{index=0}" in the text itself, so that's a valid G15 deletion in my view. Any human reading the text should've noticed that. Toadspike [Talk] 21:15, 3 April 2026 (UTC)

My only "evidence" for that would be my CSD log, but I always interpreted This may include ... in the G15 criteria as it being a non-exhaustive list of examples. And because no actual person is going to sit down and write "[oai_citation:0‡Stanwood WA]", I have interpeted that to cover the signs which indicate that the page could only plausibly have been generated ... and would have been removed by any reasonable human review part. --Gurkubondinn (talk) 19:25, 2 April 2026 (UTC)

Most reasonable humans looking at a preview of the page or working in Visual Editor would never see the URL, and most of those that do see it wouldn't actually read it or realize anything is wrong with it, as it's a perfectly functional URL. I'm not disputing that this was AI-generated, just that G15 doesn't apply, and we currently have a significant gap between what is eligible for G15 and what is prohibited by NEWLLM. Toadspike [Talk] 20:17, 2 April 2026 (UTC)

Oh, I definitely agree on the gap between G15 and NEWLLM. But I slightly disagree with what reasonable humans should (not would) notice. But that's fine. Like I said above, our job here is done as far as I'm concerned. I don't care about if the draft gets deleted or not, I just didn't want this draft (not subject) to make it into article space. --Gurkubondinn (talk) 20:20, 2 April 2026 (UTC)

Upper Arlington High School

Thank you for your attempt to fix Upper Arlington High School and the no wiki text issue. I went in and fixed it after reverting your changes, what I had done was rewrite the section, so it doesn't have to rely on the star symbol period, that way it fixes the issue.

Thanks again! JLN2026 (talk) 00:13, 3 April 2026 (UTC)

Yeah, sorry about that. It looked like a wikitext markup mistake with nested lists at first, but when you reverted my non-fix then I saw how you had meant it. I was going to suggest using {{efn}} footnotes instead, but rewriting the prose like you did is a much more elegant solution. Thanks for taking care of that! --Gurkubondinn (talk) 00:18, 3 April 2026 (UTC)