Live data from Hacker News

Generative AI and Wikipedia editing: What we learned in 2025

wikiedu.org

71–80 of 132 posts

Re: Generative AI and Wikipedia editing: What we learned in 2025

#71

> That means the article contained a plausible-sounding sentence, cited to a real, relevant-sounding source. But when you read the source it’s cited to, the information on Wikipedia does not exist in that specific source. When a claim fails verification, it’s impossible to tell whether the information is true or not. This has been a rampant problem on Wikipedia always. I can't seem to find any indicator that this has…

Linkrot is a problem and edited articles are another. Because you can cite all you want, but if the underlying resource changes your foundation just melted away.

Re: Generative AI and Wikipedia editing: What we learned in 2025

#76
post #36

> That means the article contained a plausible-sounding sentence, cited to a real, relevant-sounding source. But when you read the source it’s cited to, the information on Wikipedia does not exist in that specific source. When a claim fails verification, it’s impossible to tell whether the information is true or not. This has been a rampant problem on Wikipedia always. I can't seem to find any indicator that this has…

LLMs can add unsubstantiated conclusions at a far higher rate than humans working without LLMs.

True, but humans got a 20 year head start and I am willing to wager the overwhelming majority of extant flagrant errors are due to humans making shit up and no other human noticing and correcting it.

My go too example was the SDI page saying that brilliant pebble interceptors were to be made out of tungsten (completely illogical hogwash that doesn't even pass a basic sniff test.) This claim was added to the page in February of 2012 by a new wikipedia user, with no edit note accompanying the change nor any change to the sources and references. It stayed in the article until October 29th, 2025. And of course this misinformation was copied by other people and you can still find it being quoted, uncited, in other online publications. With an established track record of fact checking this poor, I honestly think LLMs are just pissing into the ocean.

Re: Generative AI and Wikipedia editing: What we learned in 2025

#77

> That means the article contained a plausible-sounding sentence, cited to a real, relevant-sounding source. But when you read the source it’s cited to, the information on Wikipedia does not exist in that specific source. When a claim fails verification, it’s impossible to tell whether the information is true or not. This has been a rampant problem on Wikipedia always. I can't seem to find any indicator that this has…

The problems I've run into is both people giving fake citations (the citations don't actually justify the claim that's being made in the article), and people giving real citations, but if you dig into the source you realize it's coming from a crank. It's a big blind spot among the editors as well. When this problem was brought up here in the past, with people saying that claims on Wikipedia shouldn't be believed unle…

A common source of error is in articles for movies where it gives plot summaries. The plot summaries are very often written by people who didn't watch the movie but are trying to re-resemble the plot like a jigsaw puzzle from little bits they glean from written reviews, or worse just writing down whatever they assume to be the plot. Very often it seems like the fuck ups came from people who either weren't watching the movie carefully, or were just listening to the dialogue while not watching the screen, or simply lacked media literacy.

Example [SPOILERS]: the page for the movie Sorcerer claims that rough terrain caused a tire to pop. The movie never says that, the movie shows the tire popping (which results in the trucks cargo detonating). The next scene reveals the cause, but only to those paying attention; the bloody corpse of a bandito laying next to a submachine gun is shown in the rubble beside the road, and more banditos are there, very upset and quite nervous, to hijack the second truck. The obvious inference is that the first truck's tire was shot by the bandit to hijack/rob the truck. The tire didn't pop from rough terrain, the movie never says it did, it's just a conclusion you could get from not paying attention to the movie.

Re: Generative AI and Wikipedia editing: What we learned in 2025

#79
post #62

Earlier quoted context omitted.

That thing is "the main competitor to Wikipedia" in the same way I'm the main competitor for the Olympic 100m race. I mean, both I and the winner have legs so it's going to be a close race, right?

It’s on its way to becoming more popular and a clear competitor to it. Just a matter of time.

Is that "more popular" in the sense of McDonald's popular?

Re: Generative AI and Wikipedia editing: What we learned in 2025

#80

> That means the article contained a plausible-sounding sentence, cited to a real, relevant-sounding source. But when you read the source it’s cited to, the information on Wikipedia does not exist in that specific source. When a claim fails verification, it’s impossible to tell whether the information is true or not. This has been a rampant problem on Wikipedia always. I can't seem to find any indicator that this has…

Linkrot is a problem and edited articles are another. Because you can cite all you want, but if the underlying resource changes your foundation just melted away.

Pretty much every citation added to wikipedia is passed on to web archive now, either by the editor or automatically later on.

For news articles especially the recommendation now is to use the archive snapshot and not the url of the page.

It’s not a perfect solution, but it tries to solve the link rot issue.

Post reply on HN