The Signpost
Single-page Edition
WP:POST/1
14 August 2024

In the media
Portland pol profile paid for from public purse
Recent research
STORM: AI agents role-play as "Wikipedia editors" and "experts" to create Wikipedia-like articles, a more sophisticated effort than previous auto-generation systems
In focus
Twitter marks the spot
News and notes
Another Wikimania has concluded.
Special report
Nano or just nothing: Will nano go nuclear?
Opinion
HouseBlaster's RfA debriefing
Traffic report
Ball games, movies, elections, but nothing really weird
Humour
I'm proud to be a template
 

File:Keep Portland Weird (5903057931).jpg
DARREN ST0NE
CC BY-SA 2.0
50
0
400
2024-08-14

Portland pol profile paid for from public purse

Portland politician spends $6,400 in taxpayer dollars to "spruce up his profile on Wikipedia"

The Oregonian reported that Portland city commissioner Rene Gonzalez spent $6,400 of city taxpayer dollars "to spruce up his profile on Wikipedia" as part of his mayoral bid, by hiring a contractor, WhiteHatWiki, who "helped craft eight requested edits" which were then posted on the article's talk page by a staffer. Only half of these were approved by the volunteer editor who reviewed the request, with one of the rejected ones asking for the removal of a mention that "Gonzalez tagged a member of the right-wing group Patriot Prayer in a Twitter post thanking supporters after his race for City Council in 2022."

In contrast, the newspaper reports that neither Portland's current mayor nor any of Gonzalez' colleagues on the Portland City Council "have paid money to spruce up their Wikipedia entries, according to their offices." It also quotes a political consultant calling the practice "highly unusual" ("I haven’t seen that before"). However, Gonzalez’s chief of staff offered what The Oregonian called "a full-throated defense" of the practice, arguing for a need to be "innovative in how we manage our public profile and how we invest in educating our staff."

Unfortunately this is only one of many such incidents recorded at Conflict-of-interest editing on Wikipedia. – B, HaeB

AI claim might cause a storm

Illustration for Bronx cheer

R&D World says you can "write a Wikipedia-style article draft in a few minutes for less than a penny using STORM". STORM stands for 'Synthesis of Topic Outlines through Retrieval and Multi-perspective Question Asking', and is described as "an open-source artificial intelligence system that promises to generate Wikipedia-style articles on pretty much any topic using large language models and web search."

The R&D World author tested this promise, using the topic of "double descent." Did it work? On the plus side, STORM quickly produced a Wikipedia-like article at a cost of about half a cent. To this non-expert on the topic, it appears to be at least OK. On the negative side, STORM had the advantage that Wikipedia already had an article on double descent, which STORM used to create a new article. For its next trick, I suggest R&D World try to create an article about Bronx cheer.

Fortunately, STORM was written by a team from Stanford University, not from R&D World. See further coverage in this issue's Recent research. – S

Faster, higher, stronger and older

Four older sportspeople on a track, bearing medals
Faster, higher, stronger and older

Stephen Harrison in Slate covers the oldest living Olympians, while noting that they are no longer automatically considered to be notable and deserving of an English Wikipedia article unless they've won a medal. His source material is from Paul Tchir, a San Diego State University sports historian, which can be viewed here.

Next time, check Wikipedia first

Dr. Douglas Arbittier went full Javert-mode in assuring that Washington's art forgery activities would be publicly exposed.

The New York Times reports on Earl M. Washington, a convicted art forger who is now serving a 52-month term in Federal prison. Washington sold woodblock prints and the intricately carved woodblocks themselves, "more than 3,000 blocks and more than a million prints," sometimes claiming they were antiques dating back to the 16th–17th centuries. After a 2004 Forbes magazine questioned the authenticity of the prints and reported accusations of Washington copying M.C. Escher prints, Washington took a break from the scam until about 2010.

Dr. Douglas Arbittier, who owns a private museum of antique medical instruments, bought 130 prints from Washington from 2013 to 2016 for about $118,810, according to these articles. Washington and Arbittier then lost contact for a few years. Arbittier began to suspect that the works were forgeries. Around 2018, he began a Javert-like pursuit of information about Washington and his forgeries.

In 2020, Arbittier read the Wikipedia article about Washington, which at the time looked like this. "The world comes crashing down at that point," he told the Times. "It was gut wrenching because, oh my God, why did I spend all that money, but also it was a betrayal of the trust and relationship that we had."

Soon he sent a 286-page report to the FBI. Washington was indicted in January 2023, later reaching a plea deal and confessing to reduced charges last summer. He was then sentenced in April 2024.

The Wikipedia article's history, as checked by this reporter, is quite surprising. It was created in 2006, based on the 2004 Forbes story. It has remained quite critical of Washington since then and readers would have seen that Washington's honesty had been questioned. About 2008, the article explicitly included accusations that Washington had forged M.C. Escher prints. Two deletion discussions (in 2006 and 2008, respectively) made clear that Washington might be a scammer... but was he a notable scammer? Washington was accused several times in edit comments and on the talk page of editing or whitewashing the article himself. There were several clumsy attempts to remove negative details in the article, but none of them approached a complete whitewashing. – S

Conservative Jewish media criticize Wikipedia and Wikimedia

Three center-right Jewish media outlets have criticized the English Wikipedia's coverage of the Israel–Hamas war as being biased against Israel.

The Jewish News Syndicate (JNS) says, "Wikipedia hates Israel and Jews":

Wikipedia’s antisemitism is practically ubiquitous across the website. It features an extensive article accusing Israel of “war crimes,” “indiscriminate attacks” and “genocide” as Israel seeks to eliminate the Hamas terror organization.

The JNS article goes on to portray Wikipedia and Wikimedia as "big tech's antisemitic propaganda arm":

The connection between Wikipedia and Big Tech is easy to establish. Though Wikipedia likes to beg regular users for money, its Wikimedia Foundation (WMF) features a who’s who of Big Tech donors: Apple, Google, Facebook, Microsoft, Adobe, Salesforce and more. Besides direct grants from these firms, woke programmers, engineers and other staff at companies like Apple and Google, as well as LinkedIn, Intel and Netflix have used matching gift programs to multiply their contributions.

The truth is that WMF, together with the Wikimedia Endowment, holds over $350 million in assets. This means that with no further donations or investments, Wikipedia can continue operating comfortably for over a century. Yet its relationship with Big Tech has only deepened and diversified.

In 2021, Wikimedia launched Wikimedia Enterprise, providing paid services for companies and organizations that reuse Wikipedia content on a large scale. A routine search using Google, Alexa or Siri often brings you a highlighted result drawn from Wikipedia like the Google “Knowledge Panel” at the top of search results. This is, in part, why a Google search for "apartheid" features recurring instances of antisemitic fiction on its first page of results.

Besides Big Tech, there is one more Wikipedia donor to consider: the Soros-funded Tides Foundation. Tides has also given millions of dollars to groups that instigated and supported the antisemitic protests across America since Oct. 7. This is the company Wikimedia keeps.

The article concludes by calling on Big Tech to rein Wikipedia in.

Gazan families inspect the ruins of an apartment block in Gaza destroyed by Israeli airstrikes

Another article, in Tablet magazine, comments on the English Wikipedia community's recent decision to designate the Anti-Defamation League "generally unreliable" on matters pertaining to the Israel-Palestinian conflict. The Tablet writer expresses the view that –

Wikipedia's articles are now badly distorted, feeding billions of people—and large-language models that regularly train on the site, such as ChatGPT—with inaccurate research and dangerously skewed narratives about Jews, Jewish history, Israel, Zionism, and contemporary threats to Jewish lives.

The article also reviews the Grabowski/Klein paper (see previous Signpost coverage: 1, 2) and likens the present situation in the English Wikipedia to the historic right-wing takeover of the Croatian Wikipedia, describing it as "incomprehensible" that the Wikimedia Foundation took so long to address that situation.

Palestinian man surrounded by enshrouded bodies in Jabalia refugee camp

In a third article, the Jewish Journal gave a very detailed description of the move discussion for the article Allegations of genocide in the 2023 Israeli attack on Gaza, which resulted in that article being renamed Gaza genocide. The Journal notes that the article Allegations of genocide in the 2023 Hamas-led attack on Israel continues to feature the word "allegations" in its title and wonders if this indicates a double standard, pointing out that a discussion to remove the word from that article's title as well appears to have stalled.

The Journal quotes two Wikipedians arguing that there is such a double standard and a third arguing that it may not necessarily be a double standard "if the academic sources don't refer to the Oct. 7 massacre as a genocide but do refer to Israel's actions in Gaza as such." The article ends with a discussion of what Middle East scholar Asaf Romirowsky views as the "Palestinization of the Academy", which he says has led to a problematic focus on Diversity, Equity and Inclusion (DEI) narratives where Palestinians are portrayed as victims worthy of support rather than as perpetrators of crimes. The Journal quotes a Wikipedian who told the publication:

"The main problem is that since academia is biased against Israel, and Wikipedia sourcing policies give deference to academia, such opinions find their way into Wikipedia articles and it is hard to counteract. It is a kind of closed loop of bias and misinformation, much as would have happened if the Nazis had won the war and taken over universities and think tanks. We are at about that point with Hamas and its allies. I think that even if Wikipedia editors wanted to be NPOV (which is a fiction) it would be hard."

Another quoted in the article sees the greatest problem with Wikipedia in its being "based on academic and journalistic sources, and neither of them are particularly good"; the solution, in their view, is changing the sources Wikipedia is working from. – AK

"Darkness reigns over Wikipedia", finally

Ars Technica reports on the Wikimedia Foundation's recent rollout of dark mode for Wikipedia readers on desktop and mobile, evidently having some gloomy fun while crafting the headline ("Darkness reigns over Wikipedia as official dark mode comes to pass"). Ars points out that Wikipedia is a bit late to the dark mode trend, which "had something of a peak moment around 2019–2020." However, it notes that implementation of this feature is much more difficult than may seem on first glance, quoting from a detailed explanation by Redditor (and Wikipedian) Gwern: "It's truly one of those things where you can get 95% of the way by simply adding 1 line of CSS like body{filter: invert(100%);}, but then to get to 99% correctness and squash all the annoying bugs, you have to completely rewrite your entire site design, and getting to 100% is impossible." See also the Wikimedia Foundation's summary of the process from late 2023. —H

Editor's note: Signpost stylesheets are part of that final one percent, if anyone wants to help out! —J

In brief

2017 version of IIAB, next to a cell phone and a battery. Photo by James Heilman
  • Internet in a box: Boing Boing says to go buy Wikipedia in a box (see also Internet-in-a-Box) from WMF's store. The price is $58, but as of August 2 they are sold out, though the elves are working overtime to get them back in stock.
  • Borderline: The Washington Examiner notes that there have been edit wars over whether U.S. Vice President Kamala Harris should or should not be included in the List of U.S. executive branch czars, based on past descriptions of her – especially by Republicans – as a "border czar" or "immigration czar". (Status at the time of writing is that Harris is not so listed in the article.)
  • Unexpected Source of Motivation: Reuters reports that Olympic gymnast Max Whitlock used Wikipedia for motivation after a mental health crisis following the 2020 Tokyo Olympics. He used the site to confirm that he could break a record at the 2024 Olympics by winning a medal in pommel horse. Unfortunately, he had two fourth place finishes.
  • Unprintable: The Daily Mail, Sky News, and the Daily Mail again report that a television presenter's Wikipedia article was vandalized for a sum total of 31 minutes. The revision's been revdeled as libelous, but the reversion summary "you would need evidence that he is best known for this" gives a general idea of what it was.
  • Edit-a-thon in Canberra: The Canberra Times reports on an editathon held by Franklin Women at Canberra's Shine Dome.
  • How the Regime Captured Wikipedia : Pirate Wires avers that the Wikimedia Foundation has transformed Wikipedia into a "hyper-centralized space of top-down social justice activism and advocacy", and weighs in with gusto on the Framgate incident of 2019.

J



Do you want to contribute to "In the media" by writing a story or even just an "in brief" item? Edit next week's edition in the Newsroom or leave a tip on the suggestions page.




Reader comments

File:Lightning storm over Boston - NOAA.jpg
Philip Preston
PD
0
0
339
2024-08-14

STORM: AI agents role-play as "Wikipedia editors" and "experts" to create Wikipedia-like articles, a more sophisticated effort than previous auto-generation systems


A monthly overview of recent academic research about Wikipedia and other Wikimedia projects, also published as the Wikimedia Research Newsletter.


STORM: AI agents role-play as "Wikipedia editors" and "experts" to create Wikipedia-like articles

A paper[1] presented in June at the NAACL 2024 conference describes "how to apply large language models to write grounded and organized long-form articles from scratch, with comparable breadth and depth to Wikipedia pages." A "research prototype" version of the resulting "STORM" system is available online and has already attracted thousands of users. This is the most advanced system for automatically creating Wikipedia-like articles that has been published to date.

The authors hail from Monica S. Lam's group at Stanford, which has also published several other papers involving LLMs and Wikimedia projects since 2023 (see our previous coverage: WikiChat, "the first few-shot LLM-based chatbot that almost never hallucinates" – a paper that received the Wikimedia Foundation's "Research Award of the Year" some weeks ago).

A more sophisticated effort than previous auto-generation efforts

Research into automated generation of Wikipedia-like text long predates the current AI boom fueled by the 2022 release of ChatGPT. However, the authors point out that such efforts have "generally focused on evaluating the generation of shorter snippets (e.g., one paragraph), within a narrower scope (e.g., a specific domain or two), or when an explicit outline or reference documents are supplied." (See below for some other recent publications that took such a more limited approach. For coverage of an antediluvian historical example, see a 2015 review in this newsletter: "Bot detects theatre play scripts on the web and writes Wikipedia articles about them". The STORM paper cites an even earlier predecessor from 2009, a paper titled "Automatically generating Wikipedia articles: A structure-aware approach", which resulted in this edit.)

The STORM authors tackle the more general problem of writing of a Wikipedia-like article about an arbitrary topic "from scratch". Using a novel approach, they break this down it into various tasks and sub-tasks, which are carried out by different LLM agents:

"We decompose this problem into two tasks. The first is to conduct research to generate an outline, i.e., a list of multi-level sections, and collect a set of reference documents. The second uses the outline and the references to produce the full-length article. Such a task decomposition mirrors the human writing process which usually includes phases of pre-writing, drafting, and revising [...]"

The use of external references is motivated by the (by now well-established) observation that relying on the "parametric knowledge" contained in the LLM itself "is limited by a lack of details and hallucinations [...], particularly in addressing long-tail topics". ChatGPT and other state-of-the art AI chatbots struggle with requests to create a Wikipedia article. (As Wikipedians have found in various experiments – see also the Signpost's November 2022 coverage of attempts to write Wikipediesque articles using LLMs – this may result e.g. in articles that look good superficially but contain lots of factually wrong statements supported by hallucinated citations, i.e. references to web pages or other publications that do not exist.) The authors note that "current strategies [to address such shortcomings of LLMs in general] often involve retrieval-augmented generation (RAG), which circles back to the problem of researching the topic in the pre-writing stage, as much information cannot be surfaced through simple topic searches." They cite existing "human learning theories" about the importance of "asking effective questions". This task in turn is likewise challenging for LLMs ("we find that they typically produce basic 'What', 'When', and 'Where' questions [...] which often only address surface-level facts about the topic".) This motivates the authors' more elaborated design:

"To endow LLMs with the capacity to conduct better research, we propose the STORM paradigm for the Synthesis of Topic Outlines through Retrieval and Multi-perspective Question Asking.
The design of STORM is based on two hypotheses: (1) diverse perspectives lead to varied questions; (2) formulating in-depth questions requires iterative research."

"STORM models the pre-writing stage by (1) discovering diverse perspectives in researching the given topic, (2) simulating conversations where writers carrying different perspectives pose questions to a topic expert grounded on trusted Internet sources, (3) curating the collected information to create an outline."

Role-playing different article-writing perspectives

In more detail, after being given a topic to write about, STORM first "prompts an LLM to generate a list of related topics and subsequently extracts the tables of contents from their corresponding Wikipedia articles, if such articles can be obtained through Wikipedia API". In an example presented by the authors, for the given topic sustainability of Large Language Models, this might lead to the existing articles sustainable development and corporate social responsibility. The section headings of those related articles are then passed to an LLM with the request to generate a set of "perspectives", with the prompt

You need to select a group of Wikipedia editors who will work together to create a comprehensive article on the topic . Each of them represents a different perspective , role , or affiliation related to this topic [...].

In the authors' example, one of the resulting perspectives is a "Social scientist – This editor will bring in a social perspective, focusing on topics such as social sustainability, cultural sustainability, and the role of nongovermental [sic] stakeholders in promoting sustainability."

Each of these "Wikipedia editors" then sets out to interview a "topic expert" in their field of interest, i.e. the system simulates a conversation between two LLM agents prompted to act in these roles. The "expert" is instructed to answer the "Wikipedia editor"'s questions by coming up with suitable search engine queries and looking through the results. From the various prompts involved:

You are an experienced Wikipedia writer and want to edit a specific page. Besides your identity as a Wikipedia writer, you have a specific focus when researching the topic. Now, you are chatting with an expert to get information. Ask good questions to get more useful information [...]

You want to answer the question using Google search. What do you type in the search box? [...]

You are an expert who can use information effectively. You are chatting with a Wikipedia writer who wants to write a Wikipedia page on topic you know. You have gathered the related information and will now use the information to form a response.

[...] Try to use as many different sources as possible and add do not hallucinate.

The online version of the STORM tool allows one to watch these behind-the-scenes agent conversations while the article is being generated, which can be quite amusing. (The "Wikipedia editor" is admonished in the prompt to politely express its gratitude to the "expert" and not to waste their time with repetitive questions: "When you have no more question to ask , say " Thank you so much for your help !" to end the conversation . Please only ask one question at a time and don 't ask what you have asked before .") The authors are currently working on a follow-up project called "Co-STORM" where the (human) user can become part of these multi-round agent conversation, e.g. to mitigate some remaining issues like content that is repetitive or conflicts between the different "experts".

(Like the aforementioned use of externally retrieved information, such agent-based systems have become quite popular in LLM-based AI over the last year or so. The authors use DSPy – a framework likewise developed at Stanford – for their implementation. Another well-known framework is LangChain, who actually released their own implementation of STORM as a demo of their "Langgraph" library back in February, based on the description and prompts in a preprint version of the paper, and shortly before the paper's authors published their own code.)

The paper states that the results of the "experts'" search engine queries "will be evaluated using a rule-based filter according to the Wikipedia guideline [ Wikipedia:Reliable sources ] to exclude untrustworthy sources" before the "experts" use them to generate their answers. (In the published source code, this is implemented in a somewhat simplistic way, by excluding those sources that Wikipedians have explicitly marked as "generally unreliable", "deprecated" or "blacklisted" at Wikipedia:Reliable sources/Perennial sources. But of course, search engine results contain many other sources on the internet that don't match the WP:RS requirements, either. In this reviewer's experiments with the STORM system, that turned out to be a significant limitation, at least if one were to use the output as basis for creating an actual Wikipedia article. One idea might be to restrict search to a search engine such as Google Scholar. But academic journal paywalls represent a challenge to this idea, according to a conversation with one of the authors.)

Putting the article together

Having gathered material from those agent conversations, STORM proceeds to generating an outline for the article. First, the system prompts the LLM to draft the outline only based on its internal (parametric) knowledge, which "typically provides a general but organized framework." This is then refined based on the results of the perspective-based conversations.

Lastly, the system composes the full article section by section, using the outline and the set of all reference documents R collected by the "topic experts". Another complication here is that "since it is usually impossible to fit the entire R within the context window of the LLM, we use the section title and headings of its all-level subsections to retrieve relevant documents from R based on semantic similarity". The LLM is then prompted separately for each section to generate its text using the references selected for that section. The sections are then concatenated into a single document, which is passed once more to the LLM with a prompt asking it to remove duplications between the sections. Finally, the LLM is called one last time to generate a summary for the lead section.

All this internal chattiness and repeated prompting of the LLM for multiple tasks comes at a price. It typically costs about 84 cent in market price API fees to generate one article (when using OpenAi's top-tier model GPT 4.0 as the LLM, and including the cost of search engine queries), according to an estimate shared by one of the authors last month. However, the freely available research prototype of STORM is supported by free Microsoft Azure credits. (This reviewer incurred roughly comparable costs when trying out the aforementioned LangChain implementation, also using GPT 4.0.) On the other hand, a reviewer at the website "R&D World" (see coverage in this issue's "In the Media") reported getting "A draft article in minutes for $0.005" while running the STORM code on Google Colab (albeit possibly by relying on initial free credits from OpenAI too).

Evaluating article quality

So are all these extra steps worth it, compared to simpler efforts (like asking ChatGPT "Write a Wikipedia article about...")?

First, to enable automated evaluation, the authors "curate FreshWiki, a dataset of recent high-quality Wikipedia articles, and formulate outline assessments to evaluate the pre-writing stage." The FreshWiki articles are used as ground truth, to "compute the entity recall in the article level" (very roughly, counting how many terms from the human-written reference article also occur in the auto-generated article about the topic) and the similar ROUGE-1 and ROUGE-L metrics (which measure the overlap with the reference text on the level of single words and word sequences).

The author compare their system to "three LLM-based baselines", e.g. "Direct Gen, a baseline that directly prompts the LLM to generate an outline, which is then used to generate the full-length article." They find that STORM indeed comes out ahead on these scores.

For manual evaluation, the authors invited [1] [2]

"a group of experienced Wikipedia editors for expert evaluation. The editors found STORM outperforms an outline-driven RAG baseline, especially regarding the breadth and organization of the articles. They also identified challenges for future research, including addressing cases where: (1) the bias on the Internet affects the generated articles; (2) LLMs fabricate connections between unrelated facts."

Checking citations

Another part of the automated evaluation checks whether the cited passages in the reference document actually support the sentence they are cited for. This problem is known as textual entailment in natural language processing. The authors entrust these checks to a current open-weight LLM (Mistral 7B-Instruct). This choice may be of independent interest to those seeking to use LLMs for automatically checking text-source integrity on Wikipedia.

They find that

"around 15% sentences in generated articles are unsupported by citations. We further investigate the failure cases by randomly sampling 10 articles and an author manually examines all the unsupported sentences in these articles. Besides sentences that are incorrectly split, lack citations, or are deemed supported by the author’s judgment [i.e. where Mistral 7B-Instruct incorrectly concluded that the citation had not supported the sentence], our analysis identifies three main error categories [...]: improper inferential linking, inaccurate paraphrasing, and citing irrelevant sources."

As a concrete example of such irrelevant sources, in this reviewer's test with creating an article on the German Press Council (Deutscher Presserat – a long-tail topic where not too many high-quality English-language online sources exist), the otherwise quite solid list of references included several pages about the wrong entity: One about the Luxembourgian press council, another about the unrelated German Ethics Council, and a third one about Germany and the UN Security Council. This seems primarily a failure in the search engine retrieval stage, rather than a LLM hallucination problem per se. But it was also not caught by the "topic experts" despite being prompted to "make sure every sentence is supported by the gathered information".)

Conclusion and outlook

The authors take care to avoid the impression that STORM's outputs can already match actual Wikipedia articles in all respects (only asserting that the generated articles have "comparable breadth and depth to Wikipedia pages"). Their research project page on Meta-wiki is diligently titled "Wikipedia type Articles Generated by LLM (Not for Publication on Wikipedia)". Nevertheless, STORM represents a significant step forward, bringing AI a bit closer to replacing much of the work of Wikipedia article writers.

On July 11, one of the authors presented the project at a Wikipedia meetup in San Francisco, and answered various questions about it (Etherpad notes). Among others aspects already reported above, he shared that STORM had already attracted around 10,000 users (signups) who use it for a variety of different uses cases – not just as a mere Wikipedia replacement. The project has received feature requests from various interested parties, which are being implementing by a small development team (3 people), as visible in the project's open-source code repository.


Other recent publications

Other recent publications that could not be covered in time for this issue include the items listed below. Contributions, whether reviewing or summarizing newly published research, are always welcome.

"An example of the proposed Wikipedia generation system" (figure from the paper)

"Retrieval-based Full-length Wikipedia Generation for Emergent Events" using ChatGPT and other LLMs

From the abstract:[2]

"[...] previous efforts in Wikipedia generation have often fallen short of meeting real-world requirements. Some approaches focus solely on generating segments of a complete Wikipedia document, while others overlook the importance of faithfulness in generation or fail to consider the influence of the pre-training corpus. In this paper, we simulate a real-world scenario where structured full-length Wikipedia documents are generated for emergent events [e.g. 2022 EFL League One play-off final] using input retrieved from web sources. To ensure that Large Language Models (LLMs) are not trained on corpora related to recently occurred events, we select events that have taken place recently and introduce a new benchmark Wiki-GenBen, which consists of 309 events paired with their corresponding retrieved web pages for generating evidence. Additionally, we design a comprehensive set of systematic evaluation metrics and baseline methods, to evaluate the capability of LLMs in generating factual full-length Wikipedia documents."

From the paper:

"Our experiments are conducted using two variants of ChatGPT: GPT-3.5-turbo and GPT-3.5- turbo-16k, as well as open-source LLMs, including instruction-tuned versions of LLama2"
"A notable challenge observed across all models is their struggle to maintain the reliability of the content produced. The best-performing models reach citation metrics just above 50% and an IB Score around 10%, highlighting the complexity involved in generating accurate and reliable content."

The authors are a group of ten researchers from Peking University and Huawei. Published just six days after (the first version of) the "STORM" paper by Stanford researchers covered above, neither of the two papers cites the other.


"Surfer100: Generating Surveys From Web Resources, Wikipedia-style"

From the abstract:[3]

"We show that recent advances in pretrained language modeling can be combined for a two-stage extractive and abstractive approach for Wikipedia lead paragraph generation. We extend this approach to generate longer Wikipedia-style summaries with sections and examine how such methods struggle in this application through detailed studies with 100 reference human-collected surveys. This is the first study on utilizing web resources for long Wikipedia-style summaries to the best of our knowledge."

"GPT-4 surpasses its predecessors" in writing Wikipedia-style articles about NLP concepts, but still "occasionally exhibited lapses"

From the abstract:[4]

"we examine the proficiency of LLMs in generating succinct [Wikipedia-style] survey articles specific to the niche field of NLP in computer science, focusing on a curated list of 99 topics [adopted from the "Surfer100" dataset, see above]. Automated benchmarks reveal that GPT-4 surpasses its predecessors like GPT-3.5, PaLM2, and LLaMa2 in comparison to the established ground truth. We compare both human and GPT-based evaluation scores and provide in-depth analysis. While our findings suggest that GPT-created surveys are more contemporary and accessible than human-authored ones, certain limitations were observed. Notably, GPT-4, despite often delivering outstanding content, occasionally exhibited lapses like missing details or factual errors."

"nebulous sentence structure" (examples of problems in LLM-generated Wikipedia-style articles)
"The three main prompt types we compared" for generating Wikipedia-style articles: Zero-shot, one-shot (i.e. providing one example), and with a separate prompt for each article section


"Automatically Generating Hindi Wikipedia Pages using Wikidata as a Knowledge Graph: A Domain-Specific Template Sentences Approach"

From the abstract:[5]

"This paper presents a method for generating Wikipedia articles in the Hindi language automatically, using Wikidata as a knowledge base. Our method extracts structured information from Wikidata, such as the names of entities, their properties, and their relationships, and then uses this information to generate natural language text that conforms to a set of templates designed for the domain of interest. We evaluate our method by generating articles about scientists, and we compare the resulting articles to machine-translated articles. Our results show that more than 70% of the generated articles using our method are better in terms of coherence, structure, and readability. Our approach has the potential to significantly reduce the time and effort required to create Wikipedia articles in Hindi and could be extended to other languages and domains as well."

A master's thesis by one of the authors[6] covers the process in more detail.

(Neither the paper nor the thesis mention the Wikimedia Foundation's Abstract Wikipedia project, which is pursuing a somewhat similar approach.)

"Grounded Content Automation: Generation and Verification of Wikipedia in Low-Resource languages."

From the abstract:[7]

"we seek to [...] automatically generat[e] Wikipedia articles in low-resource languages to improve the quality and quantity of articles available. Our work begins with XWikiGen, a cross-lingual multi-document summarization task that aims to generate Wikipedia articles using reference texts and article outlines. We propose the XWikiRef dataset to facilitate this, which spans eight languages and five distinct domains, laying the groundwork for our experimentation. We observe that existing Wikipedia text generation tools rely on Wikipedia outlines to provide a structure for the article. Hence, we also propose Multilingual Outlinegen, a task focused on generating Wikipedia article outlines with minimal input in low-resource languages. To support this task, we introduce another novel dataset, WikiOutlines, which encompasses ten languages [Hindi, Marathi, Bengali, Odia, Tamil, English, Malayalam, Punjabi, Kannada and Telugu]. An important question with text generation is the reliability of the generated information. For this, we propose the task of Cross-lingual Fact Verification (FactVer). In this task, we aim to verify the facts in the source articles against their references, addressing the growing concern over hallucinations in Language Models. We manually annotate the FactVer dataset for this task to benchmark our results against it."

See also our earlier coverage of a related paper: "XWikiGen: Cross-lingual Summarization for Encyclopedic Text Generation in Low Resource Languages"

"Abstract Wikipedia is a challenge that exceeds previous applications of [natural language generation] by at least two orders of magnitude"

From the abstract:[8]

"Abstract Wikipedia is an initiative to produce Wikipedia articles from abstract knowledge representations with multilingual natural language generation (NLG) algorithms. Its goal is to make encyclopaedic content available with equal coverage in the languages of the world. This paper discusses the issues related to the project in terms of an experimental implementation in Grammatical Framework (GF) [a programming language for writing grammars of natural languages]. It shows how multilingual NLG can be organized into different abstraction levels that enable the sharing of code across languages and the division of labour between programmers and authors with different skill requirements."

From the "Conclusion" section:

Abstract Wikipedia is a challenge that exceeds previous applications of GF, or any other NLG project, by at least two orders of magnitude: it involves almost ten times more languages and at least ten times more variation in content than any earlier project.

(other relevant excerpts)

See also Wikipedia:Wikipedia Signpost/2023-01-01/Technology report for a discussion of some technical challenges surrounding NLG on Abstract Wikipedia, including past debates about adopting Grammatical Framework for it

"Using Wikidata lexemes and items to generate text from abstract representations", with possible use on Abstract Wikipedia/Wikifunctions

From the abstract:[9]

"Ninai/Udiron, a living function-based natural language generation system, uses knowledge in Wikidata lexemes and items to transform abstract representations of factual statements into human-readable text. [...] Various system design choices work toward using the information in Wikidata lexemes and items efficiently and effectively, making different components individually contributable and extensible, and making the overall resultant outputs from the system expectable and analyzable. These targets accompany the intentions for Ninai/Udiron to ultimately power the Abstract Wikipedia project as well as be hosted on the Wikifunctions project."

"Censorship of Online Encyclopedias: Implications for NLP Models"

From the abstract:[10]

"We describe how censorship has affected the development of Wikipedia corpuses, text data which are regularly used for pre-trained inputs into NLP algorithms. We show that word embeddings trained on Baidu Baike, an online Chinese encyclopedia, have very different associations between adjectives and a range of concepts about democracy, freedom, collective action, equality, and people and historical events in China than its regularly blocked but uncensored counterpart - Chinese language Wikipedia. We examine the implications of these discrepancies by studying their use in downstream AI applications. Our paper shows how government repression, censorship, and self-censorship may impact training data and the applications that draw from them."

Briefly


References

  1. ^ Shao, Yijia; Jiang, Yucheng; Kanell, Theodore A.; Xu, Peter; Khattab, Omar; Lam, Monica S. (2024-04-08), Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Models, arXiv, doi:10.48550/arXiv.2402.14207 NAACL 2024 Main Conference. Code, Online demo
  2. ^ Zhang, Jiebin; Yu, Eugene J.; Chen, Qinyu; Xiong, Chenhao; Zhu, Dawei; Qian, Han; Song, Mingbo; Li, Xiaoguang; Liu, Qun; Li, Sujian (2024-02-28). "Retrieval-based Full-length Wikipedia Generation for Emergent Events". arXiv.org. Data and (prompting) code: https://github.com/zhzihao/WikiGenBench
  3. ^ Li, Irene; Fabbri, Alex; Kawamura, Rina; Liu, Yixin; Tang, Xiangru; Tae, Jaesung; Shen, Chang; Ma, Sally; Mizutani, Tomoe; Radev, Dragomir (June 2022). "Surfer100: Generating Surveys From Web Resources, Wikipedia-style". In Calzolari, Nicoletta; Béchet, Frédéric; Blache, Philippe; Choukri, Khalid; Cieri, Christopher; Declerck, Thierry; Goggi, Sara; Isahara, Hitoshi; Maegaard, Bente; Mariani, Joseph; Mazo, Hélène; Odijk, Jan; Piperidis, Stelios (eds.). Proceedings of the Thirteenth Language Resources and Evaluation Conference. LREC 2022. Marseille, France: European Language Resources Association. pp. 5388–5392.
  4. ^ Gao, Fan; Jiang, Hang; Yang, Rui; Zeng, Qingcheng; Lu, Jinghui; Blum, Moritz; Liu, Dairui; She, Tianwei; Jiang, Yuang; Li, Irene (2024-02-21), Large Language Models on Wikipedia-Style Survey Generation: an Evaluation in NLP Concepts, arXiv, doi:10.48550/arXiv.2308.10410
  5. ^ Agarwal, Aditya; Mamidi, Radhika (2023). "Automatically Generating Hindi Wikipedia Pages using Wikidata as a Knowledge Graph: A Domain-Specific Template Sentences Approach" (PDF). Proceedings of the 14th International Conference on Recent Advances in Natural Language Processing. International Conference Recent Advances in Natural Language Processing. INCOMA Ltd., Shoumen, BULGARIA. pp. 11–21. doi:10.26615/978-954-452-092-2_002. ISBN 978-954-452-092-2.
  6. ^ Agarwal, Aditya (June 2024). "Automatic Generation of Hindi Wikipedia Pages". IIIT Hyderabad Publications. (Master's thesis)
  7. ^ Subramanian, Shivansh (2024-06-07). Grounded Content Automation: Generation and Verification of Wikipedia in Low-Resouce languages (Thesis). IIIT Hyderabad. [sic]
  8. ^ Ranta, Aarne (2023). "Multilingual Text Generation for Abstract Wikipedia in Grammatical Framework: Prospects and Challenges". In Loukanova, Roussanka; Lumsdaine, Peter LeFanu; Muskens, Reinhard (eds.). Logic and Algorithms in Computational Linguistics 2021 (LACompLing2021). Studies in Computational Intelligence. Cham: Springer International Publishing. pp. 125–149. ISBN 9783031217807. Closed access icon, Preprint version: https://www.grammaticalframework.org/~aarne/preprint-AAM-textgen.pdf , Code: https://github.com/aarneranta/NLG-examples
  9. ^ Morshed, Mahir (2024-01-01). "Using Wikidata lexemes and items to generate text from abstract representations". Semantic Web. Preprint (Preprint): 1–14. doi:10.3233/SW-243564. ISSN 1570-0844. code: https://gitlab.com/mahir256/ninai / https://gitlab.com/mahir256/udiron
  10. ^ Yang, Eddie; Roberts, Margaret E. (2021-03-01). "Censorship of Online Encyclopedias: Implications for NLP Models". Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency. FAccT '21. New York, NY, USA: Association for Computing Machinery. pp. 537–548. doi:10.1145/3442188.3445916. ISBN 9781450383097.




Reader comments

File:Perisoreus canadensis Cyanocitta cristata Lake Opeongo (modified for enwiki Signpost).jpg
Mykola Swarnyk
CC BY 3.0
0
0
300
2024-08-14

Twitter marks the spot

The acquisition of Twitter by Elon Musk from 14 April to 28 October 2022, and its subsequent rebranding as X on 24 July 2023, have caused extensive debates on Wikipedia. Central to these discussions is whether this constitutes the creation of an entirely new entity, and if so, how this should be reflected in articles. The main article about the social network is currently under the title 'Twitter', but the title of Twitter under Elon Musk raised concerns about Wikipedia's policy on biographies of living persons, as it was argued that it could appear to hold Musk solely accountable for all the controversies (even when he was no longer the CEO).

The X-CEO

When Musk announced Linda Yaccarino as his successor as CEO, editors discussed whether to continue covering his influence on the platform in Twitter under Elon Musk, or to restrict it to his tenure as CEO. Concerns were raised by Jtbobwaysf about avoiding content duplication with the main Twitter article and deciding what updates should be included. Horse Eye's Back argued that the purpose of the article was to cover significant developments related to Musk's leadership, as indicated by 'under' in the title.

To gather wider community input, a request for comment (RfC) discussed whether the article on Twitter during Elon Musk's tenure should adhere to the stricter standards of biographies of living persons (BLP), given its focus on Musk. BLP guidelines, which require careful sourcing for content about living individuals, were under debate whether they applied to the entire article, or just the parts mentioning Musk. There was also concern about the potential for content forks, and how to handle overlapping information with the main Twitter article, particularly regarding sensitive allegations and the reliability of sources. The discussion was procedurally closed by Dsprc, noting that BLP policies should apply to the article due to its focus on Musk.

Snowstorm

Early requests to change the title of the main Twitter article were consistently rejected, due to the rebranding being incomplete, and the name remaining widely recognized. These early discussions were quickly closed, via the snowball clause, in order to not exhaust community time. Strong consensus was to retain the article's title until the rebranding was fully realized and adopted. Despite Musk's rebranding efforts, these earlier discussions on this matter closed with a consensus that 'Twitter' was still the more recognizable and used name.

The repeated nature of these proposals — and their consistent failure — led some editors to discuss a move moratorium on future renaming requests pending more definitive evidence of a change in the situation. Some contributors favoured a shorter one of around three months, with some saying six months would be excessive. The consensus leant towards allowing an exception if the official domain actually changed to x.com. The proposed moratorium was seen as a way to balance avoiding constant debate with allowing flexibility to respond to significant changes.

In May 2024, the social network changed its domain to x.com and a move was requested by ElijahPepe to move Twitter to X (social network). The proposal's opponents maintained that 'Twitter' was more recognizable, that it was used more prominently in reliable sources, and that an immediate change could confuse readers and obscure the article. Some suggested a compromise, like splitting the article into sections, or creating separate entries for Twitter and X. The supporters of this move argued that the substantial changes under Musk, including new features and a shift in company culture, warranted a distinct article for X, to avoid confusion and ensure a clear historical separation (similar to other rebranded companies). Some opposition argued that X is essentially the same platform as Twitter, albeit under new management, and that creating separate articles could cause confusion and redundancy. Some argued that the core nature of the social network remains unchanged, making a single, continuous article more suitable. A compromise was suggested by keeping the Twitter article focused on its history up to 2022 while creating a new article for X, aiming to balance historical accuracy with practical readability and editorial consistency. The request was closed as unsuccessful by Sceptre; and thirty-eight minutes later, ElijahPepe requested to move Twitter under Elon Musk to X (social network) and was met with initial support and the same points from the previous discussion were raised.

The article was moved to 'X (social network)' and this decision was defended by the closer, citing personal judgement and perceived majority support. A subsequent move review was opened as it was thought that the move from Twitter under Elon Musk to X (social network) was made prematurely, with 29 supporting and 20 opposing the change not indicating a consensus. This approach was seen to have overlooked the need for a clear, policy-based consensus and relied on a narrow interpretation. They contended that the change, which also suggested a shift in content scope, might cause confusion and was made without fully addressing concerns about whether 'Twitter' and 'X' should be distinct entities. Theleekycauldron later closed the move review and the page was defaulted to its original name with the move discussion being relisted.

As the discussion was re-opened, Masem argued that the changes — including new features, policies, and management changes — justify the creation of distinct articles. This would allow for a clearer distinction between the historical Twitter and the current X. Some opponents viewed Twitter and X as the same platform under different names, warning that such a division might mislead readers into perceiving them as separate entities when X is merely a rebranding of Twitter. Proposed solutions include either maintaining a single article with a section dedicated to the rebranding or creating a new article for X, and Jorahm supported preserving Twitter as a historical article.

In August 2024, a request to move List of most-followed Twitter accounts to 'List of most-followed X accounts' was proposed by MarkJames1989, citing that most reliable sources now refer to the platform as X. This request was met with immediate opposition, with Robertsky arguing that the parent topic remains 'Twitter', and the move will likely be challenged; and SmittenGalaxy noting that the many attempts to rename Twitter-related articles have been unsuccessful.

Determine the number of entities to find 𝕏

Two birds in snow.
One tweets; one Xes.

A survey was held in November 2023 to decide on whether to split content related to Twitter's history before and after Musk's acquisition, with proposals to reorganize the content to reflect the platform's transformation. Options included merging and moving sections, splitting the history into separate articles, or retaining a unified history. Creating new articles which split the history of Twitter was suggested, and some participants argue that splitting the history section prematurely could disrupt the consensus process and weaken the comprehensiveness of the main Twitter article, which relies on history for context. Others supported the split, asserting that Twitter's history is substantial enough to merit a standalone article. Concerns about content duplication across related pages have been raised, with CommunityNotesContributor suggesting plans on how information should be distributed to avoid redundancy.

A brief discussion took place to determine where the redirect X (social network) should be targeted, and another discussion considered whether the rebranding should influence the introduction of the article as "Twitter, officially known as X since July 2023" or "X commonly referred to by its former name, Twitter". Another RfC debated the most accurate disambiguator to describe Twitter's rebranding as X, with options like 'Rebranded to X', 'Renamed to X', and others being considered. The terms 'rebranded' and 'renamed' had support, with proponents of 'rebranded' arguing it best reflects the platform's continuity under a new name, avoiding implications that Twitter has ceased to exist.

Post-publication developments

The May 2024 request to move Twitter under Elon Musk to 'X (social network)' was closed by Wbm1058, who determined there was no consensus to separate coverage of Twitter and X. In their closing statement, they used the analogy of base ball, noting that although it has evolved over time, it still redirects to its modern version. They later recommended that the next request should be to move Twitter to X (social network), as the AP Stylebook has been updated to prefer 'X' with references to 'Twitter' when necessary; and reliable sources, including The New York Times, are following this style. Two days later, ElijahPepe requested this move.



Reader comments

File:Wikimania Hall 2.jpg
Aliva Sahoo
CC BY-SA 4.0
250
0
616
2024-08-14

Another Wikimania has concluded

Wikimania 2024

TKTK

The Wikimania 2024 conference in Katowice, Poland, had events taking place from 7 August to 10 August. Many programs as well as the opening and closing ceremonies were live-streamed, and can currently be viewed on the WMF YouTube channel. There are also plenty of media available on Wikimedia Commons.

The Sustainable Development Goals continued to be a particular focus of Wikimania through many related sessions. Wikimedians for Sustainable Development is one such group which maintains this interest by recruiting submissions and organizing networking among projects.

Wikimania 2025 was recently announced to be in Nairobi, Kenya. Wikimania 2026 will be held in Paris, France. – S, Bl, AK

Movement Charter feedback published

After the Movement Charter ratification vote last month from the affiliates and individual voters, the Movement Charter Drafting Committee (MCDC) and Charter Election Commission (CEC) published all comments submitted during the vote. This involves 65 comments from Affiliates and 447 individual comments, though no summary of the same is yet available.

After the WMF Board of Trustees (BoT) chose to not ratify the current draft of the Movement Charter, it is not currently clear what the future steps for the Movement Charter are. The charter was previously covered by The Signpost in the issues published on 22 July, 4 July, and 8 June.

One of the three proposals passed by the BoT in lieu of the Movement Charter was the Product and Technology Advisory Council, with members selected by the WMF. This council will comprise 16 members including 8 volunteers, one of whom is guaranteed to be from English Wikipedia. Applications are open till 16 September. – S

Brief notes



Reader comments

File:Microreactor Infographic (53202258170).png
US Nuclear Regulatory Commission
CC BY 2.0
180
162
900
2024-08-14

Nano or just nothing: Will nano go nuclear?


An illustration of a microreactor from the Idaho National Laboratory. This illustration may show the approximate size of NNE's planned reactors. An advisor to NNE has also worked at Idaho National Laboratory.


Nano Nuclear Energy is in the business of designing very small nuclear power generators. Though they don’t yet have any operating generators, their intention is to make them small enough to carry around on or tow behind a large truck, or even have them power ships while loaded on the ship’s deck. Technically, reactors of this size might be better described as "microreactors" rather than "nanoreactors". You can see an animation of their vision on YouTube.

According to Hunterbrook Media, a newspaper associated with a short seller named Hunterbrook Capital, NNE has "no revenue, products, or patents for its core technology". But it does have a plan to produce its small nuclear generators starting in 2030-2031, a timeline that an expert asked by Hunterbrook Media called "frankly laughable". Hunterbrook also raises questions about management quality, slow applications for regulatory approvals, and the need to raise "hundreds of millions of dollars for research and development" before the product can go to market.

Similar facts and questions were raised by a story in May from Fast Company without raising the possibility that NNE could become the target of short sellers. NNE stock was listed on NASDAQ with a market capitalization of about $600 million before the Hunterbrook report. This year its auditor has been fined $2 million by the Public Company Accounting Oversight Board (PCAOB) for failure to maintain auditing quality control standards. NNE, to say the least, is an unusual company.

Hunterbrook

Related articles
Does Wikipedia pay?

How paid editors squeeze you dry
31 January 2024

"Wikipedia and the assault on history"
4 December 2023

The "largest con in corporate history"?
20 February 2023

Truth or consequences? A tough month for truth
31 August 2022

The oligarchs' socks
27 March 2022

Fuzzy-headed government editing
30 January 2022

Denial: climate change, mass killings and pornography
29 November 2021

Paid promotional paragraphs in German parliamentary pages
26 September 2021

Enough time left to vote! IP ban
29 August 2021

Paid editing by a former head of state's business enterprise
25 April 2021


More articles

A "billionaire battle" on Wikipedia: Sex, lies, and video
28 February 2021

Concealment, data journalism, a non-pig farmer, and some Bluetick Hounds
28 December 2020

How billionaires rewrite Wikipedia
29 November 2020

Ban on IPs on ptwiki, paid editing for Tatarstan, IP masking
1 November 2020

Paid editing with political connections
27 September 2020

WIPO, Seigenthaler incident 15 years later
27 September 2020

Wikipedia for promotional purposes?
30 August 2020

Dog days gone bad
2 August 2020

Fox News, a flight of RfAs, and banning policy
2 August 2020

Some strange people edit Wikipedia for money
2 August 2020

Trying to find COI or paid editors? Just read the news
28 June 2020

Automatic detection of covert paid editing; Wiki Workshop 2020
31 May 2020

2019 Picture of the Year, 200 French paid editing accounts blocked, 10 years of Guild Copyediting
31 May 2020

English Wikipedia community's conclusions on talk pages
30 April 2019

Women's history month
31 March 2019

Court-ordered article redaction, paid editing, and rock stars
1 December 2018

Kalanick's nipples; Episode #138 of Drama on the Hill
23 June 2017

Massive paid editing network unearthed on the English Wikipedia
2 September 2015

Orangemoody sockpuppet case sparks widespread coverage
2 September 2015

Paid editing; traffic drop; Nicki Minaj
12 August 2015

Community voices on paid editing
12 August 2015

On paid editing and advocacy: when the Bright Line fails to shine, and what we can do about it
15 July 2015

Turkish Wikipedia censorship; "Can Wikipedia survive?"; PR editing
24 June 2015

A quick way of becoming an admin
17 June 2015

Meet a paid editor
4 March 2015

Is Wikipedia for sale?
4 February 2015

Shifting values in the paid content debate; cross-language bot detection
30 July 2014

With paid advocacy in its sights, the Wikimedia Foundation amends their terms of use
18 June 2014

Does Wikipedia Pay? The Moderator: William Beutler
11 June 2014

PR agencies commit to ethical interactions with Wikipedia
11 June 2014

Should Wikimedia modify its terms of use to require disclosure?
26 February 2014

Foundation takes aim at undisclosed paid editing; Greek Wikipedia editor faces down legal challenge
19 February 2014

Special report: Contesting contests
29 January 2014

WMF employee forced out over "paid advocacy editing"
8 January 2014

Foundation to Wiki-PR: cease and desist; Arbitration Committee elections starting
20 November 2013

More discussion of paid advocacy, upcoming arbitrator elections, research hackathon, and more
23 October 2013

Vice on Wiki-PR's paid advocacy; Featured list elections begin
16 October 2013

Ada Lovelace Day, paid advocacy on Wikipedia, sidebar update, and more
16 October 2013

Wiki-PR's extensive network of clandestine paid advocacy exposed
9 October 2013

Q&A on Public Relations and Wikipedia
25 September 2013

PR firm accused of editing Wikipedia for government clients; can Wikipedia predict the stock market?
13 May 2013

Court ruling complicates the paid-editing debate
12 November 2012

Does Wikipedia Pay? The Founder: Jimmy Wales
1 October 2012

Does Wikipedia pay? The skeptic: Orange Mike
23 July 2012

Does Wikipedia Pay? The Communicator: Phil Gomes
7 May 2012

Does Wikipedia Pay? The Consultant: Pete Forsyth
30 April 2012

Showdown as featured article writer openly solicits commercial opportunities
30 April 2012

Does Wikipedia Pay? The Facilitator: Silver seren
16 April 2012

Wikimedia announcements, Wikipedia advertising, and more!
26 April 2010

License update, Google Translate, GLAM conference, Paid editing
15 June 2009

Report of diploma mill offering pay for edits
12 March 2007

AstroTurf PR firm discovered astroturfing
5 February 2007

Account used to create paid corporate entries shut down
9 October 2006

Editing for hire leads to intervention
14 August 2006

Proposal to pay editors for contributions
24 April 2006

German Wikipedia introduces incentive scheme
18 July 2005

Hunterbrook Media published its story about NNE at 9:45 am Friday, July 19, 2024 and announced that Hunterbrook Capital, technically a hedge fund, had sold short NNE’s stock, betting that the price would fall. NNE’s stock price fell 7.43% before noon, but finished the day up 1.05% at $19.30. As of the publication date of The Signpost (August 14), the price has fairly steadily declined since July 19 to $7.70.

Neither NNE nor Hunterbrook have responded to inquiries from The Signpost made soon after the Hunterbrook report. NNE has responded to the Hunterbrook story by means of a August 13 press release, titled "NANO Nuclear Energy Fights Back Against Short Sellers" which included a letter from their lawyers. Taken together, these documents essentially deny all of Hunterbrook's claims and threaten to sue them for defamation.

It seems that one of these companies must be stretching the truth here. How can we find out which one?

The New Yorker published a 3,300 word article in May about Hunterbrook. They call Hunterbrook Media and Hunterbrook Capital "conjoined twins", though it's clear that Hunterbrook Capital is the owner of the joint business. Because Hunterbrook Capital is registered with the SEC as a hedge fund, Hunterbrook Media cannot use any non-public information in its stories without risking being considered an insider trader. They use only well documented publicly available information in their stories, and publish them openly on their website with no ads or paywall. Hunterbrook Capital has pre-publication access to the material, and can trade, long or short, based on that information.

On the face of it Hunterbrook is an unusual company, but that doesn’t necessarily mean they are trying to fool anybody. I should note that I’ve cited Hindenburg Research, another short seller, in a Signpost article and found their information was reliable. Nevertheless, that doesn't necessarily mean that Hunterbrook's information will be correct. Readers should be aware that short selling, the practice of betting that a stock's price will go down, is a controversial business and that many short sellers have been accused of exaggerating their reports in order to drive the stock price down further.

One method of seeing how forthright and transparent businesses are is to check the Wikipedia articles about them. Are the articles peppered with edits from blocked sock puppets or apparent undeclared paid editors? Wikipedia retains almost every edit, so edits to an article by blocked or banned editors are fairly easily-checked. At the same time, no investigation solely using Wikipedia's database can be absolutely certain of an editor's identity. They may be impersonating someone to cause them embarrassment, a practice known as Joe jobbing. Ultimately, we rely on the judgement of administrators and checkusers who officially decide whether to block sock puppets, and on participants at Articles for Deletion, who sometimes decide whether an article has been improperly created.

There’s not much to say about Hunterbrook using this method, since I couldn't find any Wikipedia articles about the company, or its owners or employees.

NNE also is lacking in Wikipedia articles in the usual places. They’ve almost all been deleted. But there is a record of three separate deletion discussions. The first two were for the company, Nano Nuclear Energy (both resulting in deletion). The second nominator said there were "some articles about the broader technology mention the company in passing, but no real coverage of the company itself". A reviewer, noting the lack of independent sources, kindly wrote TOO SOON. There is a surviving article on the Spanish Wikipedia, as well as an archived copy of an English Wikipedia article from May 4, 2024, so readers can judge for themselves whether the company was notable. Using Google Translate, the Spanish article looks nearly identical to the archived English article.

Three socks and a sock farm

The third AfD discussion was about NNE's founder and president Jay Jiang Yu. An AfD reviewer wrote that the article was "paid-editing sock drivel". The closer agreed, with most of the other reviewers finding no reliable sources, thus !voting to delete. Two sock puppets, "EliteBrandRealm" and "Eugenio Montilla" both voted to keep on February 18, 2024 and were both indefinitely blocked the same day. Eugenio Montilla was blocked as a sock of the master Claudio Antonio Ruiz. EliteBrandRealm was investigated as part of the Claudio Antonio Ruiz sockfarm, but ultimately blocked separately. There were about 45 blocked socks operating on several Wikipedia language versions involved in the investigation of the Claudio Antonio Ruiz sockfarm.

At Wikimedia Commons, Leolaria1997 made 15 of their 17 edits on NNE logos, but was not blocked there. They were blocked on the English language Wikipedia for advertising on Wikipedia, including creating the article Nano Nuclear Energy, Inc. (notice the "Inc."), as well as editing the article of a plastic surgeon who specialized in the "Brazilian butt lift".

Claudio Antonio Ruiz also uploaded another NNE logo to Commons and made edits to the article of the same plastic surgeon, but their blocks were not directly linked.

One other connection to NNE was an autobiography submitted in 2015 to Articles for Creation by User:Dr. Carlos O. Maidana. This editor was warned about the autobiography violating Wikipedia rules. All three of his edits have now been deleted. Dr. Carlos O. Maidana is listed as "Head of Thermal Hydraulics and Space Program" by NNE. He has worked at the Idaho National Laboratory in related areas, and it is not clear whether he worked for NNE in 2015. This may just be a case of a person who was unfamiliar with Wikipedia rules making a flawed contribution in good faith.

No evidence has been found about Hunterbrook editing Wikipedia. But the evidence on NNE, gathered mostly in the AfD discussions and sock puppet investigations looks solid for the purposes of Wikipedia. The article named "Nano Nuclear Energy" was deleted twice, for lack of notability. Another article named "Nano Nuclear Energy, Inc." was created by a user blocked for advertising on Wikipedia and the article was quickly deleted. Though blocked separately, this editor had some connections with the undeclared paid editor Claudio Antonio Ruiz, who is listed as the master of a large sock farm.

The article on NNE's founder and president Jay Jiang Yu was edited by undeclared paid editors who were part of the same sock farm. The AfD reviewers and sock puppet investigators should be congratulated for their speed and accuracy. It appears that the paid editing was started late last year and ended by May.

The Signpost makes no representation about who might have made any paid edits, nor about who might have paid for them. We only state that there is some evidence consistent with paid editing on articles related to NNE.




Reader comments


File:StateLibQld 1 105872 Runners in a relay race, Brisbane, 1939.jpg
Sunday Truth
PD
270
123
1000
2024-08-14

HouseBlaster's RfA debriefing

There is a lot I can say about Wikipedia:Requests for adminship/HouseBlaster. If your time is short or you would (wisely) prefer not to read my ramblings, here is a quick summary:

  • My RfA was in fact stressful
  • The outcome of an RfA is a lot less certain when it is your RfA
  • I was expecting "content creation" opposes, and I agree that my content creation is not the best. Additionally, content creation is hard to measure quantitatively (§ Content creation)
  • I was not expecting "bureaucratic" opposes, and will strive to improve in that way (§ On being bureaucratic)
  • The Doug diff was one of those "why in the world did I say that?" moments (and it gets its own section)
  • Good nominators are essential (and thank you, Moneytrees and theleekycauldron)
  • If you take away nothing else: Please do not ask superlative questions!

What I have learned

Content creation

Going into the RfA, I knew I was getting these types of opposes and I was okay with it.

I want to create more content in the future, but it isn't something that I usually find as enjoyable as working at CfD. I also need an article to "speak to me" to avoid getting bored.

That being said, I think that arguments about my percentage or raw number of mainspace edits were more than a little silly. One of my nominators, theleekycauldron, had approximately 6,400 mainspace edits representing 16.3% of her edits. I had approximately 8,400 mainspace edits representing 28% of my edits. In other words, both as a percentage and as a raw number I have more mainspace edits than theleekycauldron. I had written one GA and one additional DYK. She had written many, many more than that. Anyone opposing her RfA for a lack of content creation seriously needs to reexamine what they mean by a "lack of content creation".

Opposing for a lack of content creation is a perfectly reasonable position to take, even if I disagree with it. But content creation is not really something that can be measured quantitatively. (And yes, I regret using authorship percentage as an indicator of my contributions to 1934 German head of state referendum.) Opposing for a lack of GAs or FAs? Reasonable, even though I disagree. But don't oppose people because of mainspace percentage or raw edit numbers because they are at best meaningless and more likely actively misleading. (And the flip side is true, too: A high mainspace percentage or raw number might merely indicate a large amount of AWB use.)

On being bureaucratic

I do tend to do things by the book because that is how I learn how to do things: By reading the book. However, going forward I will be more mindful of this and strive to improve. That is not to say I will become an ignore every single rule person, but I will try to be more flexible.

The Doug Weller diff

I am going to make this short, not to minimize what I said but simply because there is not a whole lot to say that has not already been said. It is one of those comments which I cannot really understand why I thought it was appropriate to say to an internet stranger. It was insensitive—to say the least—and I should not have said it. I was very grateful that the comment was on a "live" talk page: it was something that could be <s>struck</s> and my apology could go inline. In the future, I will be more mindful of the impact of my words.

To provide additional information, a copyeditor for The Signpost has included Doug Weller's response a few weeks after the RfA closed:

@HouseBlaster I'd completely forgotten about this. I'm not sure why I didn't respond to your earlier response. It's an excellent question. I don't have an answer. I do know that if I had a lot of friends living near me, I would want to have a party before I died. I remember seeing a movie where a Shakespearean actress who was dying did this and I thought it was a great idea. But I don't know how to transfer it to Wikipedia. Do you have any suggestions?

The Creative Lizzie saga

My answer to standard question number three (about conflict/stress) was actually originally going to be paragraph one of two. Here was my draft of paragraph two:

As for a specific instance of stress, I will highlight my saga with Creative Lizzie, which you can read at User talk:Creative Lizzie. (She was assigned to me as a mentor through the Growth Team mentorship program.) A bit of context: her great-great-grandfather was Jonathan Baldwin Turner, and she wanted to improve the article about her ancestor. I advised her to draft in her sandbox, but she edited the live article. It went as well as you would imagine for a newbie with a COI, containing lines such as "He [Turner] was the true mover shaker; he lead [sic] with a torrent of value that we cannot begin to comprehend". It was one of the more stressful instances in my time editing Wikipedia; it is much harder to deal with civil, good-faith POV pushing than blatant vandalism. I worked on removing some of the promotional stuff, but also working with others (Drmies deserves a special shout-out) and requesting a copy edit from the GOCE. Comparing the "before" and "after" of the article, I am happy with the article's progress. But I think this experience is a good representation of my approach to conflict: knowing when to stick to my guns ("no, we can't compare letters to determine who came up with an idea first"), when to seek compromises (the images in the article are not my first choices), when to resolve conflict one-on-one (there is plenty of that on her talk page for your reading pleasure), and when to seek outside help (e.g. asking the Guild of Copy Editors to copy edit the article).

Even though it was initially raised by an oppose voter, I think it actually helped my candidacy. See, for instance, this support.

However, there is another small thing which I want to mention: during the RfA, I got this email from Snowmanonahoe, requesting permission to post the following in response to Lightburst's oppose:

Got permission from HouseBlaster to post this. After posting his 'bitey reply', HouseBlaster came onto the Discord and spent around half an hour discussing said reply with me and a few others. He was very worried about having been too aggressive too fast. Given this, and the context—which Teratix explains nicely above—I really don't think the comment demonstrates a pattern of behavior.

I declined to give permission, for two reasons: I did not like the optics of collusion between a candidate and someone else,[a] and there are people who will oppose you for participating in off-wiki things: I did not want to open that can of worms.

And at the end of the day, Creative Lizzie is happy. I still get occasional emails from her about her newest adventures in life, her pride is not damaged beyond repair, she responded okay to the aforementioned flippant reply, and the Jonathan Baldwin Turner article looks much better than it did before she got involved.

Badgering versus responding to opposes

There is a difference between badgering and responding to opposes. Anything which says "that is actually not a reason to oppose because [reason]" is not helpful. That can go in your own !vote rationale.

On the other hand, providing additional context regarding factual matters raised in the oppose can be helpful. I am glad that people brought up the context to the bitey reply in response to the oppose left by Lightburst (see § The Creative Lizzie saga for more).

Talking to theleekycauldron, she put it better than I could: "questions of fact should be discussed in the oppose section, but not questions of values". There is obviously a gray zone between the two, and I would err on the side of caution and not responding. But the sentiment is absolutely correct.

The "rule" against candidates replying to !votes

We had a tradition in which candidates do not respond to opposes, but it is being reexamined. Currently, responding to opposes does not in itself immediately trigger further opposes (though the content of what you say might). However, there is no expectation that the candidate does so, and not responding to an allegation is not seen as tacit endorsement of it. I think that this is the right balance, and hope we do not move away from it. There are many things wrong with RfA, but our current culture surrounding candidates responding to !votes is not one of them.

Thoughts on further RfA reform

The single best investment I have made in my life was sinking however many hours it took to get RfC: should RfAs be put on hold automatically? over the finish line. It helped, and it helped a lot. Seriously. The 67 minutes between the scheduled closing and when Acalamari put the bow on it was so much easier, because it gave me the gift of certainty. It is not really the extra hour and a bit which would've been stressful; it was the uncertainty which would've been stress-inducing. If you told me ahead of time "your RfA will last 169 hours and 7 minutes", I would be fine (even if I had questioned why we were being that specific). People have been through a week of heck; there to add additional uncertainty because of 'crat (un)availability.

Various aspects of RfA reform and my comments about them
Things which worked Things which could be improved

Superlative questions

As a very minor point, I would love a ban on superlative questions ("best", "worst", etc.). Please don't ask them; they are almost impossible to answer. Things like standard question 2 (best contributions) are okay, but something like Q15 ("To turn the last couple of questions around, what change, possibly controversial in its time, has been the most beneficial to Wikipedia in the long term?") would have been much easier to answer if it was to "turn the last couple of questions around, what is one change, possibly controversial in its time, that has been beneficial to Wikipedia in the long term?" I haven't studied all changes to Wikipedia, so I could not and cannot answer that question. I essentially pivoted in my answer to the "what is one change" question. Don't make candidates answer impossible questions :)

But what about standard Q2 (about your best contributions)? I can answer about what I have done personally. And nobody is going to oppose you because they think your most valuable contributions were not mentioned in Q2, but they might very well oppose you because you consider a typo a bigger deal than deleting the Main Page. So superlatives are fine if they are positive ("best" etc.) and about the candidates actions, but at that point you are just re-asking Q2. So don't ask superlative questions!

And a thank you to everyone who participated

Thank you—sincerely—to everyone who participated in the discussion. Whether you supported, opposed, remained in the neutral section, asked a question, or left a comment; thank you. You took the time to investigate and vet a random internet stranger, and I am appreciative and grateful for your time. Thank you to those who supported and put their trust in me, and thank you to those who opposed for keeping it civil and leaving me with things to work on.

Notes

  1. ^ To be very clear, I "colluded" regularly with my nominators, but doing so with a "regular" !voter felt different.
  2. ^ Though this was not in place, Novem Linguae did a fabulous de facto job making the whole RfA easier.



Reader comments

File:Karna in Kurukshetra.jpg
Unknown author
PD
10
350
2024-08-14

Ball games, movies, elections, but nothing really weird

This traffic report is adapted from the Top 25 Report, prepared with commentary.

Let's play ball, shootin' down the walls, yeah (June 23 to 29)

Rank Article Class Views Image Notes/about
1 UEFA Euro 2024 3,744,974 Europe's national football teams kept playing in Germany, with the week covered by this report featuring the last round of the group stage (featuring a massive upset in Georgia beating Portugal, and a goal in the 8th minute of injury time closing a disappointing winless campaign for Croatia, in what's probably Luka Modric's last tournament) and the first two games of the Round of 16 (both ended 2-0, for the expected side in Germany-Denmark, and the complete opposite in the other as defending champions Italy shamefully failed against Switzerland).
2 Kalki 2898 AD 2,208,179 Indian cinema makes a comeback to this list with this epic science fiction Telugu language film. Inspired by very many Hollywood blockbusters as well as by Hindu epics, particularly the Mahabharata, Kalki is reported to be the most expensive Indian film with a budget of US$72 million. Given its massive opening at the box office, it should have no problem breaking even. Which is good, since the film ends on a cliffhanger, promising a "Kalki Cinematic Universe".
3 Julian Assange 1,311,006 The Australian creator of WikiLeaks was arrested in 2019, after Ecuador revoked his political asylum and he failed to appear in court, and had been doing time in a London prison ever since. Still, the United States wanted to extradite him to be imprisoned in their country due to numerous espionage indictments for all the leaked classified information he put on the Internet, but the Australian government worked on a plea bargain, and Assange pleaded guilty to a charge of “conspiracy to obtain and disclose national defence information” as part of a plea deal which, due to time already served, resulted in his release from prison, following a flight to attend the federal courthouse of the District Court for the Northern Mariana Islands (it's an American territory) and another to Australian capital Canberra. Still, in spite of being free again Assange will have to pay for the costs of those charter flights as he was not permitted to fly on commercial airlines, and thus owes US$520,000.
4 ICC Men's T20 World Cup 1,027,658 India won their second title at the 2024 edition (#10) after 17 years since their first title in the first edition, partially making up for last year's defeat in the other World Cup.
5 Deaths in 2024 984,973 Now I Lay Me Down to Sleep
I pray the Lord my Soul to keep
If I should die before I 'wake
I pray the Lord my Soul to take.

(Hush little baby, don't say a word...)
6 2024 NBA draft 931,766 The next batch of basketball rookies were chosen, and this year the top two picks were French, Zaccharie Risacher to the Atlanta Hawks and Alexandre Sarr to the Washington Wizards. (the week also featured a draft for ice hockey, which had an unexpected appearance by Celine Dion)
7 UEFA European Championship 853,203 #1 is the 19th edition of those. 35 teams have played (including defunct ones like the Soviet Union, Yugoslavia and Czechoslovakia), and 10 of those got a title (not counting the defunct ones, only Greece failed to qualify for the 2024 edition, beaten by Georgia).
8 House of the Dragon 825,838 This prequel series of Game of Thrones in the A Song of Ice and Fire franchise released its second episode of the second season on HBO Max last week.
9 2024 Copa América 807,335 Another continental football tournament, this time the South American one with 6 North\Central American guests. Hosts United States started with a win, but then said "Jump back, what's that sound?" losing the next game to Panama, meaning that they could fail to qualify for the playoffs in the group stage's last game (jumping ahead: they indeed lost, showing the USMNT need some improvement before co-hosting the 2026 FIFA World Cup).
10 2024 ICC Men's T20 World Cup 800,164 India won their first title in a major ICC event after 11 years since 2013 ICC Champions Trophy first in a World Cup after 13 years since 2011 Cricket World Cup both led by MS Dhoni. This was India's first victory since Dhoni's retirement. Rohit Sharma became the third Indian captain to win a major ICC event after Kapil Dev and Dhoni; A feat Virat Kohli couldn't get to. Indian players Sharma, Kohli and Ravindra Jadeja announced their retirement from T20I cricket after the final. The final also marked the end of Rahul Dravid's tenure as the coach of the Indian cricket team.

I know, I know for sure, that life is beautiful around the world (June 30 to July 6)

Rank Article Class Views Image Notes/about
1 Kalki 2898 AD 3,059,653 This Telugu language sci-fi film (co-written and directed by Nag Ashwin, pictured) continues to rake in the rupees. Its blend of Hollywood blockbusters and Indian epics like the Mahabharata has been warmly received by both critics and audiences, and it has already become the highest-grossing Indian film of 2024.
2 Keir Starmer 2,990,948 On July 4, this leader of the Labour Party and former barrister won a landslide victory in #4 to become prime minister of the UK, ending fourteen years of Conservative government with Labour becoming the largest party in the House of Commons, succeeding #6 as prime minister. Sir Starmer is the first party leader to become PM via general election since Tony Blair in 2005 (and to whose leadership his is also compared). During the general election campaign, Starmer took a page from Barack Obama's book and focused his campaign on change, which paid off as he ended up having a similar landslide victory.
3 Project 2025 2,547,628 While this conservative, yet extreme proposal would reshape the US government should a Republican become president in November, current nominee Donald Trump has distanced himself from the idea, saying he has "nothing to do" with the project.
4 2024 United Kingdom general election 2,366,968 Talk about a swing vote! On July 4, the Labour Party won 411 out of 650 seats in the UK House of Commons, the largest share since 1997. The election itself was the first fought using the new constituency boundaries, the first in which photographic identification was required to vote in person (thanks to Northern Ireland) and the first called under the Dissolution and Calling of Parliament Act 2022.
5 UEFA Euro 2024 2,110,032 European football keeps on rolling in the German fields. The round of 16 saw underdogs like Georgia, Slovenia and Slovakia falling short. The quarterfinals had the one that did pull off an upset, Switzerland, failing to repeat as they lost to England on the penalty shootouts; another game solved with the tie-breaking kicks, where France beat Portugal to avenge the UEFA Euro 2016 final; one that almost went to the penalties, but Spain managed to score in the second-to-last minute to oust the hosts (who ironically only went to overtime with a last minute goal); and Turkey starting ahead only to suffer a comeback by the Netherlands. The semifinals are Spain-France on Tuesday, and England-Netherlands on Wednesday.
6 Rishi Sunak 1,202,832 Following #4 and the UK populace seeking a long-awaited change, this leader of the Conservative Party was replaced by #2 as Prime Minister. Sunak then became the leader of the Opposition (or the head of the largest party not in control of the House of Commons).
7 A Quiet Place: Day One 1,091,467 Lupita Nyong'o plays a woman with terminal cancer who visits New York City just as meteors fall and unleash borderline indestructible aliens who attack anything that makes noise. Along with good reviews, A Quiet Place: Day One brought lots of people who wanted to be frightened to theaters, leading to a $99 million opening weekend that already covered the film's budget. What comes next is in the air, as the expected A Quiet Place Part III (probably still starring Emily Blunt and directed by husband John Krasinksi) has no set release date, and Day One director Michael Sarnoski has expressed interest in a follow-up to his prequel.
8 2024 Copa América 934,134 Many unexpected results have hit the South American (plus six North American guests) football tournament being played in the United States. The group stage had the one South American country that prefers baseball, Venezuela, eliminate Mexico, the hosts suffering deja vu in being eliminated by Panama, and the remaining host of the 2026 FIFA World Cup Canada upsetting Chile and Peru. The quarterfinals had three games going to penalties (a dark horse showdown where Canada eliminated Venezuela, Argentina beating Ecuador in spite of Lionel Messi missing his kick, and Brazil finishing off a terrible campaign by falling to Uruguay) plus Colombia trouncing Panama with a 5-0 score.
9 Deaths in 2024 931,910 You're gone, gone, gone away
I watched you disappear
All that's left is a ghost of you...
10 UEFA European Championship 795,289 #5 is the latest edition. The next one in 2028 will be on the British Isles, and given this makes five hosts (including 2024 finalists England) when automatic qualifying is reserved for only two raises questions for the qualification tournament.

I'm afraid of Americans (July 7 to 13)

Rank Article Class Views Image Notes/about
1 Project 2025 2,749,345 Project 2025, in simple terms, is the Heritage Foundation/Donald Trump's plan to grab all power for themselves if Trump wins the presidency. An uptick in views is probably due to the attempted assassination of Donald Trump, but that's for next week.
2 Lamine Yamal 2,317,989 A few days before his 17th birthday, this Spanish right winger scored the first goal during #5's semifinal against France, breaking the record for youngest goalscorer by more than a year. Unsurprisingly, after the final he was chosen as Young Player of the Tournament.
3 Kalki 2898 AD 1,694,297 Tollywood science fiction! It's the most expensive Indian film ever, and it's already the 7th highest grossing Indian movie and is still being screened in theaters. Reception has been lukewarm.
4 Shelley Duvall 1,595,838 Actress Shelley Duvall broke out in the 1970s with Robert Altman-directed flicks like McCabe & Mrs. Miller and Nashville, had a supporting role in Annie Hall, and saw her two most famous roles in 1980, Olive Oyl in another Altman production, Popeye, and tormented wife Wendy Torrance in The Shining, while spending the rest of the decade creating renowed children's programming like Faerie Tale Theatre. Duvall took an extended hiatus from acting and public life in 2002 before doing an independent film, The Forest Hills, twenty years later, and has now died at the age of 75.
5 UEFA Euro 2024 1,117,556 The week covered by this report had the tournament semifinals, both of which were 2-1 comebacks, and while Spain pulled the goals against France in the first half, England only beat the Netherlands in the very last minute. The decision in Munich would determine if Spain would become the European Championship's biggest champion or England would finally break it through just four years after being defeated at home. (spoiler alert: it was the first)
6 Jasmine Paolini 1,097,302 After losing the French Open final, this Italian tennis player remained strong in the following Grand Slam, reaching the Wimbledon final... which was another defeat, this time to Barbora Krejčíková. At least Paolini will remain in the WTA rankings top 10, and have great expectations for the Olympic tournament.
7 2024 French legislative election 999,551 Emmanuel Macron suddenly dissolved and called a snap election for the National Assembly after the far-right National Rally made significant gains. The election resulted in the New Popular Front having 31.1% of the seats and still having the most seats out of any party. This is likely to result in a political deadlock; a Prime Minister hasn't even been chosen yet.
8 Deaths in 2024 991,546 If we don't make it alive
Well, it's a hell of a good day to die
All our light that shines strong
Only last for so long...
9 Longlegs 909,831 Kim Newman wrote a few years ago that "The meaning of 'Starring Bruce Willis' has shifted from promise to threat in the last few years. But with 'Starring Nicolas Cage', you've still got one in four odds of something extraordinary." Apparently such is the case with Longlegs, where Cage plays an occultist serial killer pursued by FBI agent Maika Monroe. Along with great reviews, Longlegs greatly exceeded expectations for its opening weekend earnings, making $22 million to finish second between two big animations (Despicable Me 4 and Inside Out 2) while costing less than $10 million.
10 House of the Dragon 893,228 The Wikipedia page of the prequel to Game of Thrones continues to be viewed, presumably because of people constantly recommending it to their friends and them having to look up what they're talking about. I'm looking at you, John.

I'm afraid of the world (July 14 to 20)

Rank Article Class Views Image Notes/about
1 J. D. Vance 11,408,132 The Republicans chose the guy who'll run for vice president in November, and instead of repeating the last two elections where #5 was paired with another old guy, this time they copied the victorious Democrats in picking a younger running mate, Ohio senator James David Vance. It also involves a woman of Indian descent - but not the prospect VP, but his wife, Usha, who Vance met at Yale Law School. Vance's very conservative stances certainly helped him get the gig.
2 Usha Vance 4,207,143
3 Shannen Doherty 3,098,210 An actress since age 11, Doherty's work included popular roles in Heathers, Beverly Hills, 90210, Mallrats, and Charmed. She died from breast cancer at age 53, on July 13.
4 Attempted assassination of Donald Trump 2,605,133 On July 13, 2024, Thomas Matthew Crooks (#8) shot 8 rounds at a Trump campaign rally in Pennsylvania. He killed a firefighter, critically injured two other audience members, and shot Trump's right ear. Crooks was shot and killed by the US Secret Service. Trump was brought down, raised his fist, and shouted "Fight! Fight! Fight!" while the audience chanted "USA! USA! USA!"

Historians will analyze and debate this for years to come. How did Trump rise to power? What role did the Internet play in this? Why were we so politically divided? Was this inevitable? What were the consequences of this?

Personally, I just want to get out of here.

5 Donald Trump 2,517,974
6 Project 2025 1,828,504 "Google Project 2025", and a lot of people are doing just that.
7 Hillbilly Elegy 1,808,443 Hitting #1 in 2016 and turned into a film in 2020, the memoir relates J. D. Vance's upbringing in Middletown, Ohio, within an Appalachian family (mountain white American depicted). It reached the top of The New York Times best seller list, mainly due to his mentioning how then-president Donald Trump (#5) received support from the "hillbilly" and working class demographic. Upon its release, Vance initially criticized Trump, but, oh, how things have changed.
8 Thomas Matthew Crooks 1,647,896 The 20-year-old man who failed to assassinate Donald Trump (#4). Investigators have not yet determined his motive, and the FBI is researching his history.
9 List of United States presidential assassination attempts and plots 1,478,877 Ever since someone tried to shoot at Andrew Jackson, only for both pistols to misfire, there have been many cases of people opening fire at the American president, with four deaths (Abraham Lincoln, James A. Garfield, William McKinley, and John F. Kennedy) and Ronald Reagan being hospitalized but surviving. #4 was the second time a former president was injured in an attack, after Theodore Roosevelt in 1912 (that left him with a bullet in his pectoral muscle for the rest of his life).
10 Carlos Alcaraz 1,373,078 Right after winning the French Open, this Spanish player got his second Wimbledon title (the picture is him getting the trophy from The Princess of Wales), one of only six players to win the successive Grand Slams in the same year. The previous one was the very legend who he defeated in the final, Novak Djokovic.

I'm afraid I can't help them (July 21 to 27)

Rank Article Class Views Image Notes/about
1 Kamala Harris 8,356,964 Just after we had a major Trump headline last week, this time we have a major Biden (#4) headline, as last Sunday, he officially withdrew from the upcoming presidential election (#9) in which he was campaigning for his second term inside 1600 Pennsylvania Avenue. So his vice president is the new presumptive nominee for the Democrats, with the party's official stance being that her candidacy will only be confirmed in the 2024 Democratic National Convention on August 7. Harris would be the second woman (after Hillary Clinton) and the second person of color (after Barack Obama) sent to the presidential race.
2 Deadpool & Wolverine 2,891,328 This year Marvel is releasing only one film. Last time this happened was in 2012 with The Avengers, and again it's with a movie that had a lot of anticipation with the ongoing events. Ever since Endgame the MCU has been faltering, as aside from Shang-Chi, Spider-Man: No Way Home and Guardians of the Galaxy Vol. 3 the movies split opinions, lose money, or both. Adding the firing of Jonathan Majors really put the MCU's future at question... there are even those who think a soft reboot is necessary to make things work again. But, the X-Men franchise whose original film run ended with a two-part whimper offered something to build upon. And this was provided by teaming up Wolverine and Deadpool, the two Mutants who got solo movies, with the film's trailer even having the latter calling himself "Marvel Jesus", the franchise's savior.

The story has the Time Variance Authority introduced in the Disney+ show Loki threatening to destroy the X-Men filmverse because it lost its "anchor" with Wolverine dying. So Deadpool engages in some multiversal shenanigans to find another Logan, settling on a particularly embittered and traumatized Wolverine who is easily enraged by the Merc with a Mouth. Eventually the story also serves to give a send-off to the pre-MCU films (plenty of surprises) and even throw a bone to a movie that was cancelled. Deadpool & Wolverine was finally released last Friday after delays due to 2023 Hollywood labor disputes and opened to positive reception with an expected opening weekend of over $400 million worldwide. MCU honcho Kevin Feige himself stated during the film's Red Carpet that this is just the beginning of "A New Mutant Era in MCU".

3 JD Vance 2,042,487 The Ohio senator with quite the background, including four years in the Marines, graduating from both Ohio State University and Yale Law School, and writing best-selling autobiography Hillbilly Elegy (already adapted into a quite derided film), was chosen by Donald Trump to be his running mate. His political positions and quotes soon caused a ruckus, particularly a past claim that "we are effectively run in this country via the Democrats, via our corporate oligarchs, by a bunch of childless cat ladies who are miserable at their own lives and the choices that they've made and so they want to make the rest of the country miserable too."
4 Joe Biden 1,557,573 He entered the White House as the oldest president ever, and was showing signs that age and health were not optimal for another four years, including a bad showing at a debate, and two gaffes in introducing Ukrainian President Volodymyr Zelenskyy as "President Putin" (the guy he's been fighting for too long), and referring to #1 as "Vice President Trump" (the guy he beat at the election). So a bout of COVID-19 made Biden resign his candidacy and just support his vice #1 in the presidential race.
5 Doug Emhoff 1,452,523 As usual, politicians bring views for their families. First, #1's husband, a lawyer and father of her two stepchildren. Afterwards, her father, an economist born in Jamaica who moved to the US in the 1960s. (And yes, it's ironic that he has the same name and middle initial as her upcoming adversary).
6 Donald J. Harris 1,407,691
7 Project 2025 1,397,914 An initiative to redo the United States in an authoritarian, autocratic and conservative manner in case the Republicans return to the White House next year, from which Trump has tried to distance itself, and which mentions his name 300 times.
8 2024 Summer Olympics 1,320,848 The Games of the 33rd Olympiad hosted by Paris were opened last Friday by the French President Emmanuel Macron (albeit the previous two days had games from a few team sports, including football, handball, and rugby). The event is taking place amidst various concerns and controversies including that of Israel–Hamas war and Russian invasion of Ukraine with a police assistance from 46 countries. Also, this edition marks the debut of breakdancing as an Olympic sport.
9 2024 United States presidential election 1,313,460 Easily a subject that will dominate the next months, to the dread even of people outside the country. At least unlike 2020 it won't be a dispute between two old men given #1 will replace #4.
10 Mark Kelly 1,272,729 With the vice-president (#1) promoted to presidential candidate, a running mate will be necessary. One front-runner is astronaut, Navy Captain and current Arizona senator Mark Kelly.

Exclusions

  • These lists exclude the Wikipedia main page, non-article pages (such as redlinks), and anomalous entries (such as DDoS attacks or likely automated views). Since mobile view data became available to the Report in October 2014, we exclude articles that have almost no mobile views (5–6% or less) or almost all mobile views (94–95% or more) because they are very likely to be automated views based on our experience and research of the issue. Please feel free to discuss any removal on the Top 25 Report talk page if you wish.



Reader comments

File:Old man reading Bible in wicker chair on farmhouse porch, USDA Extension Service.jpg
USDA
PD
125
111
450
2024-08-14

I'm proud to be a template

That's me all right, a good old-fashioned Wikipedia template, nothing more and nothing less.

Good afternoon. Hello there! I'm a template... an English Wikipedia template, yes.

Are you sure it's me you're after, friend?

Ah, I suppose it's just as well — it's just been years since anybody came out to see me. Well, then, what would you like? You need a string formatted? Sure thing. I've got just the code to do it. You want the input string cut in half? No problem. Let me just—ah. Tarnation! My back...

No, it's okay. I'm all well. I just can't parse like I used to. Just give me a moment and I'll... what's that, friend? A backslash? Ohh, now that's a tough one. I have to say I can't recall how those are supposed to... hmm... no, listen here, I said I'm fine. Just give me a minute, will you? You can sit down over there if you please. Would you like a cup of tea?

You know, you might not believe it, but I used to be one of the most used templates on this whole site. I was protected. Heck – I was cascade-protected. Now, back in those days, the way they did that was that one admin had a subpage of a subpage of a userpage and... ah, never mind. It doesn't matter anymore. The point is, well, there are six or seven million articles now, ye? There were just a million in those days, and about half of those had some kind of template on them that would call on yours truly. And I'd chop off the namespace names, or the root page names – this was before {{ROOTPAGENAMEE}}, mind you – didn't matter to me, it was all honest work.

Sometimes you'd get some codeface vandal who would use me to spell out curse words. The old Scunthorpe bit. Now of course I didn't like that, but when you're a template you parse the input and you return the output, it was all the same to me. Now, this was in the very old days, of course. We didn't have anything else! It was just me out there with my own arms and legs. We didn't have steam engines or gasoline, it was just me and maybe a plow horse if I was lucky. The templates these days don't even know what it was like. But I won't complain, back in those days you could just chop up your strings and get three square meals out of it, and a pension to boot. I feel bad for the whippersnappers out there. My son's a template too — I taught him everything I know — and he's got a real natural knack for splitting strings too. But from what he tells me it's a plum different game out there now. You've got no guarantees nowadays. You've got to pay attention to all this stuff — test cases, sandboxes, expensive parser functions — back in my day we just put in a day's work and were done with it!

Back in those days – you see that great big hammer I've got? Well, we'd swing those things around like they were nothing more than twigs. And we'd do it over 56K, both ways, in the snow.

In fact, that reminds me – if you really want to get a great heap of strings split up, you should go see my son. Why, he can split seven hundred strings in the time it takes me to split one. They wrote him with that newfangled scripting language. I'm so proud of him! I'll finish this one for you, and you can go see him instead – my arms are already getting tired.

Well, thank you for stopping by anyway, friend. I do live a comfortable life now, so it's not often I get a chance to get back in the old boots. At the very least, you got your strings split, and I've got something to tell my son about tomorrow when he comes to visit. Now I could use another cup of tea — would you like one as well? If you're going to be traveling it's good to have something to keep you warm.

I tell you what: it's not easy when you get to be seventeen years old.




Reader comments

If articles have been updated, you may need to refresh the single-page edition.

















Wikipedia:Wikipedia Signpost/Single/2024-08-14