The Signpost


Wikipedia in the age of personality-driven knowledge

Contribute  —  
Share this
By Maryana Pinchuk
Maryana Pinchuk is the Wikimedia Foundation's Principal Product Manager, and has been a Wikipedian since 2011.

In March 2023, I helped organize a Zoom call to discuss external trends that may impact the Wikimedia movement, and over a hundred Wikimedians attended to talk about the opportunities and risks of generative AI. But a different concern also surfaced that really surprised and stuck with me. Longtime English Wikipedia editor Ragesoss noted that when observing how his children like to learn things online, he saw that instead of going to websites, they sought out personalities who had built large followings on platforms like YouTube, and could provide information from an authentically human point of view. In the year since that AI call, the English Wikipedia community has heatedly debated how to respond to the opportunities and threats of generative AI. But this other trend – of people (especially younger people) seeking out personalities to aggregate knowledge for them, on social platforms that make it easy and rewarding for contributors to participate in new formats like short video – hasn’t been discussed as much. And while some Wikipedians may wonder what any of this has to do with encyclopedic content, I believe that this trend deserves as much attention and conversation in our community.

Despite the excitement (and hype) around generative AI in the press over the last year, Wikipedia is still used by more people globally than ChatGPT.[1] But, when looking at Wikipedia usage compared to usage of popular social platforms frequented by young people, the comparison is far less favorable for us. TikTok has close to two billion monthly global users – twice as many as Wikipedia – and is still growing.[2] And while it's tempting to dismiss TikTok and apps like it as entertainment platforms where young people only share and watch lip-syncing and viral dances, consider that a third of US people under age 30 recently reported regularly getting their news from TikTok, and one in four US TikTok users say they come to the platform for educational purposes like learning history.

Meanwhile, recent surveys by the Wikimedia Foundation Brand Health team have shown that worldwide, people aged 18–24 report being less aware of, less likely to use, and less likely to recommend Wikipedia to a friend than older age groups. Many of these younger audiences report preferring to get information via video platforms like YouTube or social knowledge sites like Quora. Traditionally, young people have always been a huge staple of Wikipedia readership. What could account for fewer young people than retirees today saying that they use or are even aware of Wikipedia's existence, and more going to social apps for educational content?

I think it's important to look outward at what's happened in the world in the last decade to better understand these data:

Given all these changes to how people expect to get information online, what opportunities does Wikipedia have to continue to meet the needs of this younger generation and generations to come?

I think there are a few strategies we could pursue.

One strategy, which I’ve heard expressed by some Wikipedians, is evolving what our content or projects looks like, to better fit some of these new expectations and preferences. There are different proposals for how to achieve this that have recently been proposed and discussed on the Wikimedia-l mailing list: for example, should we provide the ability to create and include more interactive elements (like graphs, timelines, or quizzes) on Wikipedia articles? Should we produce and include video explainers of important topics on Wikipedia? Though the specifics differ, fundamentally these ideas are about making Wikipedia more of a “knowledge destination” that younger audiences will want to visit – to read and to contribute to – instead of going to places like TikTok or YouTube. But when this conversation came up at WikiConference North America last fall, I heard many Wikipedians express concern about where pursuing this strategy could lead us. There was fear of making Wikipedia into something it isn’t. There was also fear about the cost and risks of building big new software features and trying to compete with massive for-profit technology companies for users. I think all of these concerns are very valid.

Some alternative strategies might instead focus on:

  1. Making sure people continue to get reliable information from Wikipedia, and that they understand what Wikipedia is and that they can join our movement, no matter where they like to share or receive information online. This “free knowledge everywhere” strategy seems promising to me because creators on YouTube and TikTok already routinely harvest facts, ideas, and images from Wikipedia and include them in their videos (and it's possible that in this way, information from Wikipedia is already being viewed by more young people outside of Wikipedia than on it, in the formats that they prefer). However, Wikipedia is rarely credited or acknowledged in these videos, so audiences have no way of knowing about the amazing volunteer project that brings them this information for free, let alone that they can join in and contribute. One way to change this might be to make using our new sound logo a more consistent practice whenever Wikipedia knowledge is shared externally.
  2. Evolving the habit that has been so standard among people of my generation when reading or hearing some claim that sounds dubious: opening up a new tab and checking what Wikipedia has to say. This strategy is about seeing if we can continue to make Wikipedia the "Internet's conscience" – i.e., the knowledge commons that provides additional context; organizes secondary sources; and gives ordinary people the ability to see, participate in, and influence the consensus on what is known – while adapting to the fact that "opening up a new tab and reading longform text" might not be the way younger audiences use the Internet anymore. One way to do this might be to provide instant verification with Wikipedia wherever people are getting information.

My team at the Wikimedia Foundation, Future Audiences, is exploring these different strategies, and our mandate is to evaluate them with quick experiments (including an experiment to build and test that last example, just-in-time third-party information verification with Wikipedia). But we are not the only ones within the Wikimedia movement who are interested in better understanding and showing up in this space: Basque Wikipedians have been turning Wikipedia articles into short videos for YouTube and TikTok as well as adding them back to Wikipedia, and the Wikimedia UK chapter has recently started their own TikTok channel.

I would love to hear from more Wikipedians: what role could Wikipedia play in this new personality-driven knowledge era? What new strategies seem most promising (or not) to you? What aren't we as a movement talking about or doing that we should be?


In this issue
+ Add a comment

Discuss this story

That Signpost feature also points out how few articles actually have interactive or other fancy graphs - a clear indication that "build it and they will come" isn't always true, particularly for an all-volunteer content creation process.
As for the current situation: as the article pointed out, creators on YouTube and TikTok already routinely harvest facts, ideas, and images from Wikipedia. If our goal is to push out accurate information, as opposed to getting Wikipedia as many page views as possible, then YouTube and TikTok use of Wikipedia may be worth giving more encouragement, not reacting with despair.
Finally, I think the most important metric is how many active editors we have, and how much editing they do. If that turns negative, then we have a very serious problem. -- John Broughton (♫♫) 23:29, 29 March 2024 (UTC)[reply]
Indeed, those Youtubers I watch already cite Wikipedia or even show it on screen occasionally. That is a practice we should encourage and make easy: Always cite your sources!
Proper attribution should be standard behaviour and easy, when reusing Wikimedia content. As of today, there is still much UX design and metadata cleanup work to do in this regard, cf. e. g. this recent blogpost. HHill (talk) 11:53, 30 March 2024 (UTC)[reply]
@HHill and @John Broughton: Thank you for reading and sharing your thoughts! @John Broughton: I think you're spot on in your summary here:

If our goal is to push out accurate information, as opposed to getting Wikipedia as many page views as possible, then YouTube and TikTok use of Wikipedia may be worth giving more encouragement, not reacting with despair.

Our mission has never been to drive clicks to a website. It is to share the sum of human knowledge, and if this knowledge finds its way to people in new ways, I think that should be celebrated. But as @HHill points out, attribution is an important ingredient in ensuring that Wikipedia knowledge is shared sustainably – without any paper trail back to the concept of Wikipedia as the source, a) it's hard for viewers to know if any given YouTube or TikTok video is full of carefully checked facts harvested from Wikipedia or... not, and b) over time, the awareness that Wikipedia is a place where reliable knowledge is collected may fade from people's consciousness. And with fading awareness that Wikipedia is a website you can read, we may indeed start to see declines in people who experience the magic of discovering that Wikipedia is a website they can edit...
A few years ago, some colleagues and I worked on this attribution guide intended for 3rd party content reusers, but I'm interested in hearing any thoughts or ideas you have about how to approach attribution in a way that gets it to stick – because asking people to read and follow a guide only goes so far. I'm also very curious to see if we could create that magic "I can edit!" moment wherever Wikipedia facts/information get reused (but also very mindful of the careful balance between being more open/inviting to new users and not overburdening the existing community with cleanup work). I'm inspired by the work of my colleagues in the WMF Growth team (e.g., structured safe-to-try tasks) and am thinking about how some things like that might show up in places like TikTok or YouTube. Your thoughts/ideas on either of these topics (attribution, off-platform contributions) would be very appreciated! Maryana Pinchuk (WMF) (talk) 21:32, 1 April 2024 (UTC)[reply]
While it's great that there's a guide for this, I honestly doubt the vast majority of content reusers know it exists. From where I'm standing, I think the problem with non-attribution comes down to a few factors:
  1. Most people aren't even aware of our license, let alone its requirement to attribute. Even new (or sometimes seasoned) editors copying content within Wikipedia need consistent reminders that they need to provide proper attribution. Without displaying our license more prominently, people aren't even going to know about it, let alone care.
  2. Almost as a rule, Wikipedia is seen as a monolithic entity where content just exists, not an easily-editable platform that thousands of thousands of real people put time and labour into. You would be surprised (or maybe you wouldn't) at the amount of times I've told people I'm a Wikipedia editor, and they didn't even understand that was a thing that existed. If it's seen as just a thing and not a collection of people, then many are going to think it's fine to lift (or even plagiarise) from.
  3. Maybe I'm not informed enough on this, so feel free to correct me if I'm wrong, but I get the feeling that the Wikimedia Foundation doesn't see enforcing its license as an issue worth pursuing. If our license requires attribution, but there are no consequences (no matter how minor) for not providing attribution, then many unfortunately aren't going to see it as necessary.
  4. This is something I actually saw the YouTuber hbomberguy bring up in his recent video about plagiarism (which highlights a couple cases of YouTubers plagiarising Wikipedia): reusers that lift content without attribution almost invariably see the people they're exploiting as beneath respect, they're not going to attribute our work to us because they don't see us as worthy of attribution. It is sometimes shocking to me how much contempt is levelled at Wikipedia editors and our work by different parts of the internet; many people see our work as worthless, so why would they bother pointing out that they got their information from a platform they consider so contemptible?
I don't think any of these are unfixable problems, although the last one may be a lot harder to fix as it requires broader cultural changes. In fact, I think the first three have solutions that we can do in house:
  1. Display the CC license more clearly on more pages (not in small print at the very bottom of the page), make it even harder to miss that there are clear terms outlined for reuse of our content. The only place I'm regularly confronted with the terms of licensing is on pictures on Wikicommons, and even then it is very easy to miss.
  2. We need to do better ourselves at properly attributing the work of our users. Right now, if I want to find out the authorship of an article, I need to click on "View history" (which many don't know exists), then click an external link to a separate website, then scroll down to see the authorship in terms of percentage and characters. In contrast, Britannica prominently lists the primary authors of its articles underneath the article title, with secondary contributors also being displayed prominently in the history. Make it clearer that there are real people behind these articles and who they are, and this'll contribute to more people understanding that there are real people doing this work and they deserve attribution.
  3. We should probably have a form set up where users can submit information about their work being reused without attribution, so that the Wikimedia Foundation can pursue the issue. I've tried being direct with content reusers about attributing the work of Wikipedians/Wikimedians, whether through emails or comments, but it doesn't work, they usually just ignore such messages. So we really do need the force and resources of the Foundation to help us with this. We should be active in collecting data on content reuse without attribution, leaning on reusers to provide attribution and pursuing them further if they continue unapologetically breaking the terms of our license. It should be made clear to reusers that reuse without attribution isn't just wrong, but has potential consequences.
These are just my thoughts on the matter, which I've been mulling over for quite some time. (I think I started thinking about this sometime last year, after I saw my work getting plagiarised twice in one week) I understand they may not all be workable, to others' liking or if this is even the right venue for bringing this up. I just really wanted to get all this down in writing. Apologies if I've gone on for too long about it. In any case, I think off-wiki reuse without attribution is a major issue that needs addressing, and I'd be happy to see more done on that front. --Grnrchst (talk) 09:50, 2 April 2024 (UTC)[reply]
@Grnrchst thank you for sharing these thoughts, and it's so painful and sad to hear how uncaring some reusers can be about taking the hard work of Wikipedians without given them proper (or any...) credit. This is definitely something I and my colleagues care a lot about. IANAL, but my understanding is that we're in a complex and tricky spot to approach this topic from a legal POV, for the following reasons: the Wikimedia Foundation doesn't "own" the licenses used on our projects (the license terms are specified and maintained by Creative Commons) and we certainly don't own the underlying content (which belongs to the original authors). Furthermore, the CC license terms on attribution are flexible by design, to accommodate unforeseen types of reuse (i.e., via new technologies or new creative applications) – but that flexibility makes it harder to argue why something is/isn't reasonably attributed. So, my (non-lawyer) understanding is that pursuing legal action against every reuser who doesn't attribute correctly would be quite resource-intensive, and we would have to balance an investment in WMF Legal staff time spent on this against other important matters (like trademark infringement, Terms of Use violations, etc.).
For all the above reasons, I wonder if taking a different (i.e., more of an awareness/marketing) approach could be both more effective and more cost-effective at getting reusers to attribute – i.e., showcasing the existence of our editing community and highlighting how to do attribution right via popular social channels, doing more to promote the importance of reuse among the tech sector, etc...
Would you mind if I put you on my mental list of Wikipedians to reach out to as I and others at WMF think about how to do some interventions in this area? Maryana Pinchuk (WMF) (talk) 17:23, 5 April 2024 (UTC)[reply]
@MPinchuk (WMF): Thanks for the response, I understand my suggestions may not be workable, was just spitballing because I think something needs doing. And aye feel free to reach out at any point. :) --Grnrchst (talk) 11:12, 6 April 2024 (UTC)[reply]
The "BY" requirement of CC BY-SA doesn't kick in if you're just citing or paraphrasing Wikipedia. And featuring the usernames of contributors more prominently would be incompatible with WP:OWN. Anybody who contributes to Wikipedia should expect they will get no credit for their work.
I see no problem with content creators not crediting Wikipedia. If they referenced Wikipedia for their research, what I want from them is to verify the claim with the sources cited in the article they read and then credit those sources, not to credit Wikipedia, which we all know is not itself a reliable source. Nardog (talk) 08:33, 6 April 2024 (UTC)[reply]
I strongly disagree that crediting contributors has anything to do with ownership or rewards (in particular I take issue with the idea that proper credit is a reward and not basic decency). In fact, neither of those articles you link to mention anything about authorship credit. --Grnrchst (talk) 11:20, 6 April 2024 (UTC)[reply]
@Grnrchst: I agree with you on the question of rewards; WP:DNER was the wrong essay to cite, especially since it's primarily an essay about on-wiki rewards. (IOW, don't expect Wikipedia to reward your editing with special privilege.) That doesn't apply here.
However, I largely agree with Nardog on the question of crediting editors more prominently — how does that even work, when a typical article is the collective product of hundreds if not thousands of individual contributors? What makes someone a "primary author" of any given Wikipedia article? How would we even decide?
"source: Wikipedia" IMHO satisfies both the ethical and moral obligations of the project license. (I won't comment on legal obligations because IANAL.) Asking people who reuse Wikipedia content to name-check everyone with a Wikipedia account who's ever tweaked the article's punctuation feels not only unworkable, but counterproductive. And personally, I don't see it as in any way necessary, when it comes to my own contributions. Others' opinions may, of course, vary. FeRDNYC (talk) 17:44, 6 April 2024 (UTC)[reply]
I don't see how DNER is inapplicable here (on-wiki rewards are just some of the rewards it discusses, not the only kinds), but if you say so, there are also WP:YANI, WP:NOTYOU, and WP:DGAF. It's not a novel or obscure concept. Nardog (talk) 10:39, 7 April 2024 (UTC)[reply]
If you're a content creator and your only source is a Wikipedia article, then you have a moral obligation with your audience (and perhaps with society at large) to disclose where you got that information. But unless you're redistributing the whole or a part of the article verbatim or adapting it beat by beat, you have neither legal nor moral obligation to credit individual contributors. We document facts, and nobody owns them. Nardog (talk) 04:41, 7 April 2024 (UTC)[reply]
@Grnrchst: I actually made a quick video about some of the stuff you are talking about with regards to Wikipedia and Hbomberguy. :PMJLTalk 00:39, 14 April 2024 (UTC)[reply]
@MPinchuk (WMF): Videos are indeed a lot of work, which I can confirm from my own Youtube work. At this point I've spent months on single videos, but those videos do tend to be fairly long (ten minutes +). I don't have any incentives to make those videos, as I am not making any money whatsoever on Youtube with my skills. One or two-minute explainers on Youtube would be a lot simpler to make, perhaps even less work than doing a full-article audio recording (and we have surprisingly many of them). Technologically, I don't think we need to implement anything new; Commons video would already work sufficiently. I am realizing this concept sounds more like a Wikiversity thing than a Wikipedia thing, but I do think this kind of thing falls within our existing policies, similar to making explainer graphics. If someone takes the lead, who knows. ~Maplestrip/Mable (chat) 06:24, 6 April 2024 (UTC)[reply]
One of areas of Wikipedia that I worked on from 2006 through 2008 was U.S. Congressional elections. Since then, I've seen a very significant drop-off in this area - lots of Congressional elections have no articles, or minimal articles, and biographical articles are often not updated with more recent election results. And this is essentially text-only. Why do we think that editors would flock to Wikipedia if it offered (more) rich-media editing, particularly since we essentially offer no (visible) credit to authors. -- John Broughton (♫♫) 20:25, 5 April 2024 (UTC)[reply]
We almost (edit conflict)'ed saying basically the same thing
See my response to @Maplestrip above. I agree that this is a very important set of question:
  • Do Wikipedians want to do this kind of work (make/triage videos for Wikipedia), and/or are there people outside the current Wikipedian/Commonist pool who would want to join our communities to do it?
  • How can we test that in a lightweight way, before investing tons of time and money into new technical solutions that may or may not see sustained uptake?
I came up with the idea of a "Featured Video" project OTOMH above, but that would be one potentially very interesting signal –  i.e., if there were a critical mass of motivated Wikipedians who wanted to sift through existing Commons media and/or make new videos and pilot this as a new WikiProject dedicated to finding and adding videos to more Wikipedia articles, that would be very cool! Maryana Pinchuk (WMF) (talk) 20:46, 5 April 2024 (UTC)[reply]
We could leave updating of some graphical content to folks like Our World in Data and simply pipe their amazing material in. (Basque)
For video we have VideoWiki which is an automated system that builds videos from mediawiki based scripts. This makes updating easy (when the tool is not broken). Doc James (talk · contribs · email) 22:33, 16 April 2024 (UTC)[reply]


The Signpost · written by many · served by Sinepost V0.9 · 🄯 CC-BY-SA 4.0