teens and twentysomethings today are of a very different demographic and have markedly different media consumption habits compared to Wikipedia’s forebears. Gen Z and Gen Alpha readers are accustomed to TikTok, YouTube, and mobile-first visual media. Their impatience for Wikipedia’s impenetrable walls of text, as any parent of kids of this age knows, arguably threatens the future of the internet’s collaborative knowledge clearinghouse.
The Wikimedia Foundation knows this, too. Research has shown that many readers today greatly value quick overviews of any article, before the reader considers whether to dive into the article’s full text.
So last June, the Foundation launched a modest experiment they called “Simple Article Summaries.” The summaries consisted of AI-generated, simplified text at the top of complex articles. Summaries were clearly labeled as machine-generated and unverified, and they were available only to mobile users who opted in.
Even after all these precautions, however, the volunteer editor community barely gave the experiment time to begin. Editors shut down Simple Article Summaries within a day of its launch.
The response was fierce. Editors called the experiment a “ghastly idea” and warned of “immediate and irreversible harm” to Wikipedia’s credibility.
Comments in the village pump (a community discussion page) ranged from blunt (“Yuck”) to alarmed, with contributors raising legitimate concerns about AI hallucinations and the erosion of editorial oversight.
Research has shown that many readers today greatly value quick overviews of any article, before the reader considers whether to dive into the article’s full text.
I’m in my mid 40s and this is true of me depending upon why I might be reading something. Sometimes, I want to do a deep dive; sometimes I just want to know roughly what
thingis.However, let me say: fuck AI summaries.
There’s alrrady simple.wikipedia.org. Why alienate editors with AI-summaries? Add the summaries from there?
deleted by creator
It always seemed to me that Wikipedia had a great and vigilant team. It never occurred to me they may need help. Can an idiot like me be of any help? I love Wikipedia and I want it to survive into the distant future.
Yes, you can do chores: https://en.wikipedia.org/wiki/Wikipedia:Maintenance
am i crazy or don’t most articles already have perfectly good summaries? i dont even buy the premise here.
Not at all. I just got out of something like long-covid (quickly tired and complex stuff is hard) and can attest that reading Wikipedia summaries is hard vs. some random blogs quick summary.
Are you feeling better now? I hope you are.
Yeah, only something like it, thanks. ☺️
Really? I’ve always believed that it is an amazing way to learn basic concepts, because of how articles are linked to each other; furthermore, I have not had any issues with reading entire articles in one sitting. God, what has happened to this world???
I used to complain that regular people can’t write a single page of text. Regular people nowadays can’t read a single page of text… our society is devolving, fast.
I read lots of Wikipedia articles.
I feel like this will end up being a cycle. The AI companies are pushing for AI summaries, which then leads to younger Gen Z and the new Gen Alpha being familiar with and preferring AI summaries. This, in turn, leads to other companies implementing AI summaries because they see that’s what the new generations are using, which leads to the new generations being even more accustomed to those summaries, which leads to more companies implementing them, and it’s just an endless cycle.
It’s damn near impossible to make any credible edits to any wikipedia page, anymore. I’ve just stopped all together.
That’s a real issue, but the article uses it as a jumping of point to get for AI slop.
I tried making a simple edit, and after creating my account I was told it was not happening because my IP was geo blocked.
Tried from office, same thing.
Well, keep your secrets then.
Username is relevant?
Wake up Lemmy, it’s time for your daily, Wikipedia should have more AI slop article.
Let’s make it 1400 words this time, and make sure to mention that younger generations watch Ticktok, but ignore that most TickTok slop is just people summerizing Wikipedia articles.
surely this article from a AI-pusher aligned with… checks notes …some eastern european business school with connections to MIT/Harvard, has the best interests of wikipedia/the public in mind. and isnt just using this in a thinly vieled attelpy at pushing the adoption of AIslop
The article is very biased - it basically suggests young people are unwilling to read, that AI is a good thing and that the wikipedia contributors are being unreasonable. It goes on to talk about how AI has “extracted value” from Wikipedia in an unquestioning way - no mention of compensation to the project, just talking about what a triumph Wikipedia is a source for AI to train on.
The “Simple Summaries” situation is less to do with the summaries and more to do with the risk of AI slop being introduced into Wikipedia unquestioned. The summaries were unchecked and unverified, which add a real chance that wikipedia started serving up inaccurate summaries and undermined it’s own reputation.
In addition that idea that younger generations don’t have the concentration span to “read a wall of text” is pernicious and patronising nonsense part of a general media bias against Gen Z and Gen Alpha. There seems to be this barely questioned narrative that they have short attention spans and are unwilling or even unable to read, just because they grew up in the era of social media like Instagram and latterly Tik Tok.
I’ll give a better hypothesis for why younger generations spend less time on wikipedia: the big tech giants like Google have stolen all the information people have put on there and serve it up in their own summaries on the search engine (preventing click throughs) or through their own AI slop engines. They don’t want people clicking through to Wikipedia, they want them clicking through to an ad. The problem is not Wikipedia, and the problem is not Gen Z or Gen Alpha; the problem - as is frequently the case - is the tech mega-corporations who steal everything (including wikipedia) and sell it back to us with ads or via AI slop.
Not reading allat 🥀🥀
Back in my day it was TL:DR! Get off my lawn!
Yes.
Yet behind the celebrations, a troubling pattern has developed: The volunteer community that built this encyclopedia has lately rejected a key innovation designed to serve readers.
But not that one, because rejecting AI 1) is not a generational rejection and 2) it is correct to reject it.
What I think is or will be the generational problem: the community that maintains it and decides what is being accepted or rejected is an “in group” that it is impossible to break into with conflicting ideas. For example, I do think the gaming, game mechanics and game development related pages can be vastly improved. But I don’t think the people responsible for those pages are interested in the changes I would suggest.
All the wikis for different games could just be on wikipedia. But they’re not, probably because they were rejected, because it’s “not relevant”. Well, some people decided they were relevant after all and they made their own wikis for those. The outcome is tribalism based fragmentation, because of differences in opinion of who values what and what should be preserved and what shouldn’t.
I’m with you on rejecting AI being sane, but the idea that gaming wikis should be integrated into wikipedia is kinda nuts. If I search “Iron” on wikipedia I’m looking for facts, not a thousand item long disambiguation cluttered with every game that has iron as a resource. Conversely, on a game wiki my search for “Iron” has an entirely different context and I’m looking for different info.
Not to mention game wikis have way lower editorial standards, their own tone (e.g. making jokes), versioning concerns, their own new user friendly homepages etc.
Wikipedia could tuck this all into a separate namespace, sure, but that’s effectively a separate wiki anyway and then it raises questions like “why is wikipedia hosting a mechanical guide for this porn game?” or “How long do we need to host the content for this game that peaked in 2012 and is now abandonware?” that are conveniently sidestepped by those communities supporting themselves.
deleted by creator
If I search “Iron” on wikipedia I’m looking for facts
Not what I meant.
The point is: there is an established group of editors, with established rules and preconceptions, an established interpretation on what good sources are and what a neutral perspective is and isn’t, and there is no chance of changing those and that is why I have no interest in interacting with wikipedia in any constructive way.
I could talk about politics too, I picked video games because I know those articles are also bad.
“designed to serve readers” [citation needed]
This was not, in fact designed to serve readers. No possible meaning of that is in anyway correct. It is “non-designed to serve non-readers”
They want to dumb down Wikipedia
Idk who “they” is. But from what I’ve seen, the administrators of Wikipedia tend to bias intake of new power-users and mods to people who have been with the project from inception (or, at least, the earlier the better). You get all sorts of justifications for why they’ve adopted this policy. But the bottom line is that Millennials and GenX make up the overwhelming majority of ranking users. And as they age out, they aren’t being replaced with people who were their age when they started using the platform.
This traditionalist base has done a lot to calcify how Wikipedia functions, even as variant communities have improved on the model.
The AI-summary shit is just the tip of the iceberg on the system’s problems. The website is filling up with dead links. The definition of a “trusted news source” is getting outrun by private sector buyouts of old media and unemployed journalists spinning up new media. A big chunk of the organizations’ resources have to deal with fending off legal threats and attacks on system vulnerabilities. The centralized hosting model is expensive to maintain. The rush to be “first to post” creates unnecessary drama among power users in popular niche fields. International language support is… meh (one area where AI would be a huge benefit, as LLMs really shine in this field).
This goes a lot farther than “they want to hurt my Wiki”. And if you bothered to read the whole article, you might see more of why. The Wiki Foundation has dragged its heels on automation and clustered around a handful of power-mods in a way that’s undermined its Open Editor model. Fighting over Simple Article Summaries is just the latest fumble by the leadership, a sizable commitment of resources that’s tossed in the dump almost as soon as its off the press.
Fighting over Simple Article Summaries is just the latest fumble by the leadership, a sizable commitment of resources that’s tossed in the dump almost as soon as its off the press.
It wasn’t off the press, it was announced and in the works but still not close to shipping. Maybe Wikimedia could’ve talked about this great innovative project with the actual Wikipedia community before investing so much money into it.
International language support is… meh (one area where AI would be a huge benefit, as LLMs really shine in this field).
What would international language support entail? Translating articles into other languages?
Which isn’t a bad thing. Wikipedia has for the last 25 years aimed at providing you with every bit of knowledge there is on a topic. That simply is not what people want when they look for information. No-one wants to read a full library’s worth of text when they want to figure out what happened in WWII. But Wikipedia lists all the minutae of every battle on every part of land, sea and air, including all the acting people from generals down to the lowliest private.
This is crazy. Articles already start with perfectly good summaries. look at your example:
https://en.wikipedia.org/wiki/World_War_II
the very first paragraph is this:
World War II[b] or the Second World War (1 September 1939 – 2 September 1945) was a global conflict between two coalitions: the Alliesand the Axis powers. Nearly all of the world’s countries participated, with many nations mobilising their resources in pursuit of total war. Tanks and aircraft played major roles, enabling the strategic bombing of cities and delivery of the first and only nuclear weapons ever used in war. World War II is the deadliest conflict in history, causing the death of over 60 million people. Millions died in genocides, including the Holocaust, and by massacres, starvation, and disease. After the Allied victory, Germany, Austria, Japan, and Korea were occupied, and German and Japanese leaders were put on trial for war crimes.
Why do we need any more summaries than that?
That’s an encyclopedia’s job.
Let’s consult Wikipedia (emphasis mine) [1].
Works of encyclopedic scope aim to convey the important accumulated knowledge for their subject domain, such as an encyclopedia of medicine, philosophy or law. Works vary in the breadth of material and the depth of discussion, depending on the target audience.
[1] https://en.wikipedia.org/wiki/Encyclopedia#Four_major_elements
Historically, general encyclopedias were limited by the physical amount of space they took up. Wikipedia is not limited by the page and volume counts of physical media and we shouldn’t treat it as such.
While I can agree that domain-specific encyclopedias should continue to limit the scope of their information to relevant topics, I see no reason that Wikipedia should follow suit. Who truly benefits from reducing and editorializing information, especially when the fundamental principle is the free and open flow of knowledge? Wikipedia could certainly benefit from writing on complex topics that is friendlier to the average joe, but that should never come at the expense of restricting the sum total of knowledge stored in its servers.
Wikipedia could certainly benefit from writing on complex topics that is friendlier to the average joe
simple.wikipedia.org exists for anyone who wants easier to understand articles.
…which seems like a much better way to generate summaries, honestly. Pull in human-written ones, and expand the simple version as necessary.
Simple.wikipedia isn’t a summary of regular Wikpedia, it’s a whole separate thing. It’s intended to convey the same data, just in a simpler way.
Simple English is for people who would like a simpler language. I’m advocating for reduced scope – or at least better organization of detail. Move stuff that’s irrelevant in the great scheme of things to subpages or pages with narrower scope, instead of writing one single compendium on a topic.
I feel like the English Wikipedia is already better at this. In the German, on the other hand, the first sentence sometimes contains multiple lines of etymological derivations of the article’s title before it even mentions what it’s about (as soon as I stumble upon one of these monstrosities again, I’ll report the example here).
Then go make one.
I don’t disagree that some articles could use better information hierarchy. Headings could make that experience way better. But to say that the info shouldn’t be there at all is short-sighted and ignores the point of an encyclopedia.
That simply is not what people want when they look for information.
Well, except for those who do. The problem is a use case mismatch. I’d argue, if anything, an encyclopedia should contain the minutiae. Unfortunately, there’s no huge compendium of brief but accurate and sourced synopsis of the same topics. To be fair, we’ve never really had one.
I agree with the editors that embedded AI summaries are not a good idea (at the moment, at least). Users can bring summarizers to the data set of that’s their want, or someone (maybe even wikimedia) will find a way to provide this in a way that preserves the underlying data’s validity. Stripping Wikipedia of its full context seems like a bad idea.
To all the people downvoting the above comment, when was the last time you’ve read a WP article of 10k+ characters from top to bottom?
A week ago, at most.
Wikipedia already has a simplified version. Literally simple.wikipedia… etc. For example, the page on the Vietnam War can inform you with a few paragraphs on each of the key points of the war. Harold Holt’s page has 3 paragraphs and an info box. It isn’t thorough by any means. It does, however, give the reader a chance to learn about something real quick with lesser chances of getting stuck in the mud and falling down wikipedia rabbit holes.
Somebody just needs to inform the simpletons that there is an easier to digest format already. No need to shrink a well of knowledge when there is a drinking fountain next to it for those who didn’t bring a bucket and rope.
That simply is not what people want when they look for information.
What? Is there anyone out there that prefers to find small bits of information lying around various sources over a concise summary followed by a solid fleshing out, all in one place? I honestly cannot imagine a use case where I would prefer that a source omits a bunch of information rather than just structure the information so that I can find what I’m looking for. Wikipedia does that. That’s why you have dedicated articles for all those battles in WWII, with their own table of contents and summaries to help you digest them. There has literally never in human history existed any source of knowledge coming even close to structuring and summarising this amount of information as well as Wikipedia has, and you’re advocating that they should make it… not that?
I would argue that there should be a simplified section and a classic ALL THE THINGS version under a “Would you like to know more?” button. I cannot tell you the hours I’ve spent following wikipedia rabbit holes and how rewarding that has been to me.
That’s what the info boxes on the side of the article are for. They’re the simplified, just-the-facts version. If you want to know more, you read the whole article, or look for the section that contains the info you need.
I don’t see why you would want to hide the hoard of knowledge that is a good Wikipedia pare behind a button. There’s already a summary at the top of the page and a table of contents for when you want more on some topic.
AI could help editors translate from other languages, but beyond that, it’s an inefficient mess that Wikipedia doesn’t need, plus given how much of AI is just regurgitating Wikipedia, It’ll give itself mad
cowAI disease.How do I download the full Wikipedia again?
Check out Kiwix: https://kiwix.org/en/applications/ Here’s their library of scrapped sites: https://library.kiwix.org/#lang=eng
Thank you!
The walls of text wouldn’t be so impenetrable if literacy rates were higher.
Sooooooo should we start creating an alternative to Wikipedia now ?
