TheWikimedia Forum is a central place for questions, announcements and other discussions about theWikimedia Foundation and its projects. (For discussion about the Meta wiki, seeMeta:Babel.) This is not the place to make technical queries regarding theMediaWiki software; please ask such questions at theMediaWiki support desk; technical questions about Wikimedia wikis, however, can be placed onTech page.
You can reply to a topic by clicking the "[edit]" link beside that section, or you canstart a new discussion.
Latest comment:1 day ago8 comments6 people in discussion
On a conservative estimate, there's probably been more than 1000+ articles related to airlines routes, train-lines, roads and so forth as well as their associated stations/airports deleted from EN Wikipedia for essentially being unencyclopaedic. In almost all cases these were at least verifiable, they just lacked notability.
I've been told multiple times that people actually find these articles useful. Having looked at WikiTravel, this is not travel information per se so it can't be interwikied there (or at least they don't want it). It seems like there's a corpus of useful content for which there are readers out there that could usefully be undeleted and hosted somewhere, just not on EN Wikipedia. For this reason a new Wikimedia wiki covering this kind of air/road/rail/sea transport link-and-hub information seems in order.
I think this material is encyclopedic, and I think Wikipedia should include it. But I'd also be happy to see it on Wikidata in addition.
At the English Wikipedia, the usual argument against it is that we should include a complete list of connections for certain kinds of transportation infrastructure, but not for other kinds, based on their personal perception of whether that type is the sort thatcould change without building expensive, permanent infrastructure. (NBcould, notactually does in all cases.) So, they say, train routes and highways should get reported in full, but air and bus routes should not.
Next, they say that you can report a "temporary" service (e.g., one that might last for decades) if it's covered in independent sources, but (a) they acknowledge that almost all such changes are reported in the newspapers covering the affected airport, and (b) they believe it would be better to have the Wikipedia article be incorrect and out of date than to cite any 'non-independent' source, even if it's authoritative. Double standards apply: Announcements from airlines about their routes are bad; announcements from train operators about their routes are acceptable.WhatamIdoing (talk)14:53, 8 July 2025 (UTC)Reply
I feel like this is way more almanac/gazetteer-ish info than encyclopedic. I would not expect to find a list of routes in an encyclopedia. Wikidata probably, and maybe even a loose-fit for Wikibooks.Going off a tangent here, but why would you document air connections? enwiki doesn't include articles on flight routes last I checked, and the possibilities are pretty infinite.Aaron Liu (talk)23:58, 8 July 2025 (UTC)Reply
It might, butw:en:WP:5P1 says "Wikipedia combines many features of general and specialized encyclopedias,almanacs, and gazetteers", so that's not an inherent problem.
I wouldn't expect an article on "Flights from London to New York". But I would expect an article about an airport to indicate which other cities/airports (if any) are regularly flown to. There's a big difference between "Tinyville Airport has scheduled passenger flights twice a week, to Big City only" and "Heathrow Airport is a major hub, with direct flights tox domestic destinations andy international destinations, including Amsterdam, Berlin, Chicago, Dublin..."WhatamIdoing (talk)01:13, 9 July 2025 (UTC)Reply
Latest comment:23 days ago3 comments2 people in discussion
Hi,
I was asked to take a survey from a box in an English Wikipedia article sayingTake a short survey and help us improve Wikipedia. I took this survey, which turned out to concern AI-generated “simple summaries” of articles on Wikimedia wikis.
I interpret this as plans to include or consider including, and build technology for such inclusion, AI-generated summaries of articles on Wikimedia wikis.
I have not found another place more appropriate for this discussion, so I am writing it here. If you know of a better place to write this, whether that is because it is specifically about these plans or this survey, or you believe it would be more helpful to the community at large to write this there, please tell me.
I want to voice the strongest opposition to these plans.
Note that I will be using “AI” here to refer to large language models and generative AI, that is, AI that works on natural language and complex human cultural artifacts to do knowledge tasks, such as OpenAI’s GPT models and LLMs in general.
The plan to include AI content in Wikimedia wikis, as it appeared in the survey, is harmful and useless. This can be said despite the fact that the survey was vague in its description of the AI content. I assume this is because the concrete form and finer details of it are still being discussed.
I can say this because it is clear from the survey and the included demonstration video"Simple summaries EN demo" on YouTube (external site) that Wikimedia Foundation or the developers of this feature are concerned or appear to be concerned about the accuracy of AI.They recognise that AI can make mistakes, and that this is a problem. I can infer this because many questions were concerned with ways to manage AI content and make it more useful and accurate.
However, due to the lacklustre factuality, neutrality and completeness of current AI models in answering knowledge questions,the amount of safeguards necessary to restrain AI to achieve an acceptable level of accuracy is so great that AI ceases to be useful. The benefit of AI is in its ability to automate these knowledge tasks. But the manual patrolling and editing of these summaries will be a similar amount of work as just writing these summaries manually.
Among the measures considered in the survey were limiting AI summaries to a smaller number of pages than all pages of the wiki, and enhanced tools for administrators and editors to review AI summaries and spot bad AI summaries. These are tasks that are already being done with human editors, and they are tasks that will not get easier using AI. Specifically the restriction of AI to a subset of pages makes it easier to instead use humans.
Wikipedia articles already have short summaries of articles for search engines and link tooltips. Any summary of this sort could be implemented in a similar way.
It is also clear that the plans intend to focus on simple language in these summaries. They are calledsimple summaries, after all. Such simplification is admittedly easy to mess up in many cases, promoting false analogies, and it is difficult to find the best simple phrasing for a complicated topic. But writing simple summaries of an acceptable quality, similar to that of AI, is easy for humans.
It is important to emphasise thatany implementation of these plans has a high chance of producing significant harm. It is known thatAI reproduces systemic biases and that thehumans trust AI, even if AI is wrong. SinceAI makes a shocking amount of mistakes, both simple and complex, both subtle and obvious, the potential for harm is immense.Particularly in areas like medicine, law, finance, health, and home and workplace safety, AI summaries will have enormous direct impact. Additionally, in areas likehistory, social studies, politics, philosophy, AI will contribute to existing patterns of abuse and misinformation, and help spread false ideas that will influence people’s fundamental world views.
But if it is assumed outright that AI is needed,it is very likely that the implementation will not be adequate. Given that AI is being considered at all, despiteoverwhelming signs that AI is harmful to the information landscape, I find it likely thatAI is being considered not because of its virtues, but in spite of its failures, and that such immense problems with AI will be overlooked in part because of a desire to use AI, divorced from its usefulness.
Therefore, I call on the Wikimedia Foundation toimmediately distance themselves from this plan and consider other uses of AI, and strongly consider abandoning any use of AI, unless significant developments in AI model accuracy are made,and explain how they will avoid the temptation to use AI due to trends in technology divorced from its actual usefulness.
—
Additionally, I want to note that, on the last slide of the AI survey, I was taken aback at how the labels for the “positive” and “negative” options on the sliders “switched places”, so to say. Before, I was asked to rate on a scale of “very good” on the left side of the screen to “very bad” on the right side of the screen. Now, I was asked to rate on a scale of “strongly disagree” on the left and “strongly agree” on the right.
I believe this unintuitive swap will contribute to respondents mistakenly selecting e.g. “strongly agree” when they meant “strongly disagree”. Because I believe that my sentiment about AI is widely shared at least among Wikipedia editors, I expect that this will lead to a skew in the data towards a positive sentiment towards AI summaries.I think it is important that this is considered when evaluating the survey data.Theanswertolifetheuniverseandeverything (talk)12:00, 4 June 2025 (UTC)Reply
This year, the Wikimedia community will vote in late August through September 2025 to fill two (2) seats on the Foundation Board. Could you – or someone you know – be a good fit to join the Wikimedia Foundation's Board of Trustees? [3]
Learn more about what it takes to stand for these leadership positions and how to submit your candidacy onthis Meta-wiki page or encourage someone else to run in this year's election.
Best regards,
Abhishek Suryawanshi Chair of the Elections Committee
On behalf of the Elections Committee and Governance Committee
Latest comment:13 days ago3 comments2 people in discussion
Hello. OnSpecial:WikiSets/14, there are currently 111 wikibooks listed that allow global bots. I went through first 15 sites, and most of them have been shut down. Is there a list of currently active projects? Or am I using outdated information? This is about deployingKiranBOT where global bots are allowed. Regards, —usernamekiran(talk)08:38, 22 June 2025 (UTC)Reply
Latest comment:14 days ago1 comment1 person in discussion
Hello everyone,
My name is Phil - I work in the Wikimedia Foundation’s Legal department, and I’m here to provide two updates on our legal challenge to the UK Online Safety Act’s“categorisation rules”. Those rules are written so broadly that Wikipedia could be lumped in as a “Category 1 service”. This would subject it to extra duties under the Act that could seriously impact the privacy, safety and empowerment of the Wikipedia community, and our collective ability to sustain the Wikimedia projects. For background on the OSA and our legal challenge, seehere (Diff), or a more detailed posthere (Medium).
First, an administrative note: the High Court has agreed to expedite our case, and set a two-day trial next month:July 22-23. We expect the hearings to be public, and can be observed in person at the beautifulRoyal Courts of Justice in London.
Second: the Foundation will be joined in this case by a Wikipedian, as joint claimant.User:Zzuuzz, a longterm UK-based user, will play a pivotal role in articulating the human rights implications of this case, including for your rights to privacy, safety, free speech, and association.
I hope you’ll join us in expressing deep appreciation to User:Zzuuzz for volunteering to take this extraordinary step, and standing up for the Wikimedia movement worldwide. This might be legal history in the making: our early searches haven’t turned up any legal precedent of a website’s host and its users proactively joining forces to bring a legal challenge.
We’ll aim to provide further updates on Meta, and we’ll watch discussions for a few days in case there are questions we can usefully answer. As this is a critical moment in active litigation, we apologise for not commenting as freely as we’d like. Best regards,
A vision of relevant, accessible, and impactful free knowledge has always guided the Wikimedia Movement. As the ecosystem of Wikimedia projects continues to evolve, it is crucial that we periodically review existing projects to ensure they still align with our goals and community capacity.
Despite their noble intent, some projects may no longer effectively serve their original purpose.Reviewing such projects is not about giving up – it's about responsible stewardship of shared resources. Volunteer time, staff support, infrastructure, and community attention are finite, and the non-technical costs tend to grow significantly as our ecosystem has entered a different age of the internet than the one we were founded in. Supporting inactive projects or projects that didn't meet our ambitions can unintentionally divert these resources from areas with more potential impact.
Moreover, maintaining projects that no longer reflect the quality and reliability of the Wikimedia name stands for, involves a reputational risk. An abandoned or less reliable project affects trust in the Wikimedia movement.
Lastly,failing to sunset or reimagine projects that are no longer working can make it much harder to start new ones. When the community feels bound to every past decision – no matter how outdated – we risk stagnation. A healthy ecosystem must allow for evolution, adaptation, and, when necessary, letting go. If we create the expectation that every project must exist indefinitely, we limit our ability to experiment and innovate.
Because of this, SPTF reviewed two requests concerning the lifecycle of the Sister Projects to work through and demonstrate the review process. We chose Wikispore as a case study for a possible new Sister Project opening and Wikinews as a case study for a review of an existing project. Preliminary findings were discussed with the CAC, and a community consultation on both proposals was recommended.
Theapplication to consider Wikispore was submitted in 2019. SPTF decided to review this request in more depth because rather than being concentrated on a specific topic, as most of the proposals for the new Sister Projects are, Wikispore has the potential to nurture multiple start-up Sister Projects.
After careful consideration, the SPTF has decidednot to recommend Wikispore as a Wikimedia Sister Project. Considering the current activity level, the current arrangement allowsbetter flexibility and experimentation while WMF provides core infrastructural support.
We acknowledge the initiative's potential and seek community input on what would constitute a sufficient level of activity and engagement to reconsider its status in the future.
As part of the process, we shared the decision with the Wikispore community and invited one of its leaders, Pharos, to an SPTF meeting.
Currently, we especially invite feedback on measurable criteria indicating the project's readiness, such as contributor numbers, content volume, and sustained community support. This would clarify the criteria sufficient for opening a new Sister Project, including possible future Wikispore re-application. However, the numbers will always be a guide because any number can be gamed.
We chose to review Wikinews among existing Sister Projects because it is the one for which we have observed the highest level of concern in multiple ways.
Since the SPTF was convened in 2023, its members have asked for the community's opinions during conferences and community calls about Sister Projects that did not fulfil their promise in the Wikimedia movement.[1][2][3] Wikinews was the leading candidate for an evaluation because people from multiple language communities proposed it. Additionally, by most measures, it is the least active Sister Project, with the greatest drop in activity over the years.
While the Language Committee routinely opens and closes language versions of the Sister Projects in small languages, there has never been a valid proposal to close Wikipedia in major languages or any project in English. This is not true for Wikinews, where there was a proposal to close English Wikinews, which gained some traction but did not result in any action[4][5], see section 5 as well as a draft proposal to close all languages of Wikinews[6].
Initial metrics compiled by WMF staff also support the community's concerns about Wikinews.
Based on this report, SPTF recommends a community reevaluation of Wikinews. We conclude that its current structure and activity levels are the lowest among the existing sister projects. SPTF also recommends pausing the opening of new language editions while the consultation runs.
SPTF brings this analysis to a discussion and welcomes discussions of alternative outcomes, including potential restructuring efforts or integration with other Wikimedia initiatives.
Options mentioned so far (which might be applied to just low-activity languages or all languages) include but are not limited to:
Restructure how Wikinews works and is linked to other current events efforts on the projects,
Merge the content of Wikinews into the relevant language Wikipedias, possibly in a new namespace,
Merge content into compatibly licensed external projects,
Archive Wikinews projects.
Your insights and perspectives are invaluable in shaping the future of these projects. We encourage all interested community members to share their thoughts on the relevant discussion pages or through other designated feedback channels.
We'd be grateful if you want to take part in a conversation on the future of these projects and the review process. We are setting up two different project pages:Public consultation about Wikispore andPublic consultation about Wikinews. Please participate between 27 June 2025 and 27 July 2025, after which we will summarize the discussion to move forward. You can write in your own language.
I will also host a community conversation 16th July Wednesday 11.00 UTC and 17th July Thursday 17.00 UTC (call links to follow shortly) and will be around at Wikimania for more discussions.
Latest comment:6 days ago6 comments2 people in discussion
The Meta Wiki is always expanding, and for its better administration, there is no longer a concrete decision to provide opportunities for volunteers in the area of information technology.
Methodologies are created for the insertion of goodwill conducts in the administration of information technology, with a focus on high school students, with a focus on institutions such as ETEC and FATEC, an example in Brazil.Chadecesio (talk)11:06, 3 July 2025 (UTC)Reply
A Meta Wiki esta sempre em expansão, e para sua melhor administração, nao ha mais concreta decisao de dar oportunidades para voluntários na area de tecnologia da informação. Se cria metodologias de inserção de condutas de boa vontade da administração da tecnologia da informação, com foco em estudantes do ensino médio, com foco nas instituições comp ETEC e FATEC.
