User talk:Tkerby

Here's wishing you a belated welcome to Wikipedia, Tkerby! I see that you've already been around a while and wanted to thank you for your contributions. Though you seem to have been successful in finding your way around, you may still benefit from following some of the links below, which help editors get the most out of Wikipedia:

If you have questions, just use this link to ask for help; a volunteer will visit you here shortly!

Need some ideas of what kind of things need doing? Try the Task Center.

I hope you enjoy editing here and being a Wikipedian! Again, welcome! AntiDionysius (talk) 20:29, 2 February 2026 (UTC)

Hello, Tkerby! Having an article draft declined at Articles for Creation can be disappointing. If you are wondering why your article submission was declined, please post a question at the Articles for creation help desk. If you have any other questions about your editing experience, we'd love to help you at the Teahouse, a friendly space on Wikipedia where experienced editors lend a hand to help new editors like yourself! See you there! ScalarFactor (talk) 06:22, 3 February 2026 (UTC)

Hello, I'm Newslinger. A comment that you recently posted to Talk:List of engineering societies seemed to be generated using a large language model (an "AI chatbot" or other application using such technology). Editors should not use LLMs to write comments generatively. Communication is at the root of Wikipedia's decision-making process, and it is presumed that editors possess the ability to come up with their own ideas. Your comment may have been collapsed per the relevant guideline. If you want to practice editing, please use your sandbox. If you think a mistake was made, or if you have any questions, you can leave me a message on my talk page. Thanks. — Newslinger talk 09:22, 9 March 2026 (UTC)

Hi

This wasnt AI generated. I may not have seen every rule on how to write a good RFC as I'm new to the process but this has come about following some disagreement on the page and I wanted to capture the history of that if any other editors wanted to see the status. If it was AI then it wouldn't have my typos in it!

Can you please remove that tag and reopen for me. If you have any editing suggestions as to how to write a better RFC then please let me know constructively rather than incorrectly flagging text as AI generated.

Thanks Tkerby (talk) 19:00, 9 March 2026 (UTC)

Hi Tkerby, your comment Special:Diff/1342035482 was clearly copied and pasted into Wikipedia, based on its formatting, and the latter half was structured the same way an AI chatbot would attempt to write an Wikipedia RfC statement when prompted. Additionally, you have already attempted to submit an LLM-generated draft (first revision), which was recognized and declined as LLM-generated. On the English Wikipedia, any LLM use in your edits should be disclosed, and LLM use is restricted by the WP:NEWLLM guideline (which prohibits the creation of LLM-generated articles) and the WP:AITALK guideline (which permits the collapsing of LLM-generated comments). The Wikipedia:Requests for comment page provides an overview of the RfC process on Wikipedia, with WP:RFCBRIEF containing examples of "neutral and brief" RfC statements. — Newslinger talk 19:38, 9 March 2026 (UTC)

Per WP:AGF it would be better to ask another editor or make genuine suggestions for improvements rather than incorrectly tagging content as AI generated. I have been using an external text editor and you're correct that I've pasted in from there buut entirely incorrect about AI use. As to the new article you cite, I'm a new editor and had incorrectly used some markdown rather than wikipedia markup if you read the discussion around that article being flagged and hadn't noticed as I'd previewed in an offline editor (which I'm now doing in the WP preview tools instead). Repeated accusations of AI risk violating WP:ASPERSIONS and damaging the reputation of new editors like myself who are trying to get to grips with editing and keeping up with all the policies. It's quite a lazy approach in my opinion when it would be far better to describe the issues and help people improve their content rather than otherwise flagging it. It also discriminates against people who do a lot of technical writing and generally present things in a way that AI tries to replicate! Tkerby (talk) 19:54, 9 March 2026 (UTC)

I've replaced the AI-specific collapse template with a generic one in Special:Diff/1342591506, as you have provided a shortened RfC statement. I apologize for the false positive in my identification here. The English Wikipedia's sentiment toward LLM use is generally negative, and inappropriate LLM use is the reason for many conduct disputes on Wikipedia today. If you do choose to use an LLM to assist with your editing, as you seem to indicate a willingness to do at Special:Diff/1342587610, please note that disclosed LLM use is consistently less controversial on the English Wikipedia than undisclosed LLM use that is later discovered. — Newslinger talk 20:31, 9 March 2026 (UTC)

Thanks for that. As a part time academic I do play by the rules for any kind of publications including Wikipedia as I'd expect of my students. But I also know how difficult it is to detect reliably and most university policy in the UK now favours responsible use and disclosure where used so I tend to support its use if done correctly. I've also seen the impact false accusations of misconduct can have of students and authors so policy generally errs very much on the side of assuming good faith (to the extent that I couldn't do anything beyond a pointed comment when a student had clearly fed my course notes into AI, generating an essay that made up unpublished papers attributed to me). Never easy to find a balance. Personally I don't care if AI has been used as long as the content is good and I'd rather feed back on any specific content points than risk falsely accusing someone.

Would like to avoid this happening again as a newbie editor and I suspect I had some bad habits from using markdown a lot in the past and how I tend to use white space. Was it extra unneeded line returns there? If so, complete newbie error on my part and not noticed when I've used an external text editor as the formatting looks fine in preview. Tkerby (talk) 20:55, 9 March 2026 (UTC)

In Special:Diff/1342035482, one part that stood out to me was that several lines of text were missing spaces between words ("andtherefore", "orimprove", and "forlist") at roughly the same point in the sentence. Some AI chatbot web interfaces render line breaks in this way when text is manually copied and pasted from them. In this case, I should have paid more attention to the fact that the missing spaces showed up just before the 120th character in the line of text, which is more likely to be a sign of text editor use than AI chatbot use. My apologies for this.

The challenge with Wikipedia is that our system of decision-making (consensus) is highly dependent on editors being required to spend a certain amount of time and effort into edits before submitting them. Without this cost, bad actors would be able to influence Wikipedia content much more easily by recruiting individuals to post LLM-generated arguments in favor of certain content or policy decisions, as using an LLM is cheaper than using a person. Additionally, some of our anti-abuse measures (e.g. against abusing multiple accounts) depend on editors having unique and distinguishable human qualities.

I'm sorry to hear about your experiences with AI-generated student papers, although I'm not sure about the extent the university's policy would be appropriate for Wikipedia, because Wikipedia's status as a human-written encyclopedia seems more existential as AI-generated web sources become more common. From my point of view, I want to ensure that every editor on Wikipedia is able to contribute manually without having to argue with bots, which many editors in past conduct disputes have described as intensely frustrating. But, I acknowledge that being falsely accused of LLM use is also frustrating, and the Wikipedia community is still trying to find an appropriate solution that works for as many people as possible. — Newslinger talk 21:42, 9 March 2026 (UTC)

Thanks for the reasoned discussion on this one. I agree with the frustrations especially getting AI to argue a point for you! I suspect we will see more of that with some of the agentic AIs as people will get them to edit for them. We are already seeing complaints on GitHub from agents where AI changes to code haven’t been accepted.

As to those missing spaces, I suspected a faulty spacebar as I’ve just replaced a sticking keyboard but think you might be right with my text editor line length settings. I tend to proofread read talk a lot less than articles so missed it completely as I was too focussed on figuring out how to use the rfc template for the first time! Tkerby (talk) 22:05, 9 March 2026 (UTC)