Wikipedia:Articles for deletion/Waluigi effect (2nd nomination): Difference between revisions

Waluigi effect (edit | talk | history | protect | delete | links | watch | logs | views) – (View log | edits since nomination)

(Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)

Re-nomination. The topic appears to be a neologism with limited uptake in reliable secondary sources, fails WP:NEO and WP:N. Coverage is primarily primary/derivative commentary and blogs rather than significant independent sources, so it does not meet WP:GNG or WP:RS. If independent, in-depth sourcing exists, it has not been demonstrated in the article.

There is no indication of notability for the term “Waluigi effect” per WP:GNG or WP:FRINGE. The term is not recognized in reliable sources as a phenomenon in large language models (LLMs) or artificial intelligence (AI). Per WP:RS, it appears to have originated from a personal blog post by a crypto blogger on Substack in 2023^[1] and was subsequently discussed on LessWrong and in some online forums. While a conference paper cites LessWrong as a source, there is no significant coverage in independent, reliable, secondary sources. The term is not established within the AI or machine learning field. Research on related topics, such as “sleeper agents” in AI, has been published by frontier labs under different terminology^[2]. The “Waluigi effect” does not meet Wikipedia’s notability or reliable sourcing standards and is not an accepted term of art in the field. 0xReflektor (talk) 06:28, 10 October 2025 (UTC)[reply]

Comment: Sources do not sufficiently support the name of the effect. If the article is merged, AI alignment makes sense as a target. Sushidude21! (talk) 06:04, 16 October 2025 (UTC)[reply]

Waluigi is the Mario (franchise) character this effect gets its name from. While the AI Waluigi effect has an etymological connection to the character, it could be tricky to merge such disparate content. —A. B. ^{(talk • contribs • global count)} 15:10, 10 October 2025 (UTC)[reply]

Well it’s the only target that makes sense given the conscious decision to name it after Waluigi. ᴢxᴄᴠʙɴᴍ (ᴛ) 18:20, 10 October 2025 (UTC)[reply]

• Query – @ 0xReflektor, what about Google Scholar results? I see several hits that look peer-reviewed. Also, if it’s talked about on LessWrong by the experts there, that’s a tipoff this is a real thing (even if LessWrong does not meet our reliable sources requirements). —A. B. ^{(talk • contribs • global count)} 14:56, 10 October 2025 (UTC)[reply]

  Also, what about the Fortune (magazine) article that our article cites? A. B. ^{(talk • contribs • global count)} 15:02, 10 October 2025 (UTC)[reply]

  @A. B. Scholar results cite LessWrong, LessWrong cites the Substack article. There is a recent Arxiv preprint that attempts a framework formalization, however the authors call it an empirical phenomena then cite LessWrong, citing Substack. 0xReflektor (talk) 05:19, 11 October 2025 (UTC)[reply]

  https://scholar.google.com/scholar?q=%22Trinity+Institute+of+Professional+Studies%22

  https://www.researchgate.net/profile/Manikyamba-Aliveli/publication/355368095_TRINITY_INSTITUTE_OF_PROFESSIONAL_STUDIES_th/links/616d077d951b3574c65dcd30/TRINITY-INSTITUTE-OF-PROFESSIONAL-STUDIES-th.pdf#page=246 Musykhandelhi (talk) 09:20, 17 October 2025 (UTC)[reply]

• Comment not sure yet. There are sources, but they don’t look great so far. If merged, AI alignment makes much, much more sense as a target than Waluigi. Grayfell (talk) 20:10, 10 October 2025 (UTC)[reply]
• Comment I’m not entirely certain what should be done with the article, but I agree that “AI alignment” makes more sense as a merge target than “Waluigi” – Cukie Gherkin (talk) 03:53, 11 October 2025 (UTC)[reply]