The Signpost

In the media

The battlegrounds outside and inside Wikipedia

German government IPs get nasty

The Frankfurter Allgemeine Zeitung reports (paywalled, in German) that government employees have made 17,000 edits mainly to the German Wikipedia in recent years. As summarized by golem.de (in German) some of the edits were traced to the IP addresses of the German Federal Office for Information Security (BSI) and contained vandalism and slurs.

In a very recent case Vice Chancellor Robert Habeck of the Green party was accused by an anonymous editor from a BSI IP of enriching himself with funds intended to fight COVID. A criminal investigation of Habeck was dropped in March and the edit was removed.

In June 2021 a government IP edited articles on "Illegal immigration" and "Illegal residence", trying to criminalize immigrants. In an article about journalists, a government IP described a professional society as "the rectum of the powerful" and described journalists as "press whores." These edits were also reverted by the Wikipedia community.

The Signpost can confirm these reports to the following extent:

  • The German language editing history of one IP range identified with BSI [1] extends from 2008 up to this week.
  • Another IP traceable to BSI has edited the English Wikipedia since 2015 through 10 March 2022. Their work includes edits to articles like two oversighted edits to Algirdas Paleckis in 2020. Paleckis is a Lithuanian diplomat and politician convicted in 2021 of spying for Russia. The conviction is currently under appeal.
The IP also edited in 2019 North Korean leaders' trains and in 2017 several minor edits to Mass surveillance in the United Kingdom. Similar edits include Hague Convention on the International Recovery of Child Support and Other Forms of Family Maintenance (2017), and Korean Demilitarized Zone (2017). While the typical edits tend to be small, many are about political and security related topics that could cause concern to citizens of the affected countries if it were known that BSI was secretly editing these articles.

Altogether there are about 1,000 English-language edits traceable to BSI, about 10,000 German language edits, and likely more in other languages.S

Pro-Kremlin disinformation outlets on Wikipedia

Do Kremlin-related sources publish disinformation on Wikipedia? If so, how well does Wikipedia deal with the problem? Some basic answers to these questions were provided this week by the European Union's East StratCom Task Force in its publication EU vs Disinfo in the story Pro-Kremlin Disinformation Outlets Referenced By Hundreds Of Wikipedia Articles which was republished by Stop Fake, a Ukrainian organization that fights Russian disinformation.

East StratCom researchers "ran a pilot study last year focusing on four well-known pro-Kremlin disinformation outlets – SouthFront, NewsFront, InfoRos and Strategic Culture Foundation. All four are linked to Russian intelligence services and are sanctioned by the US Department of the Treasury for attempts to interfere in the 2020 US elections." (wikilinks added)

They examined mentions of these websites, such as in references, over multiple versions, for at least 625 articles. Southfront.org accounted for the majority of the mentions (57 per cent), with news-front.info at 27 per cent. Twenty language versions of Wikipedia were represented, with five having the majority of mentions: Russian (136 articles), Arabic (70), Spanish (52), Portuguese (45) and Vietnamese (32).

The English-language version is not included in the tables for the sample period because a Reliable Sources Request for Comment deprecated southfront.org and news-front.info as reliable sources in 2019.

The article continues with an examination of whether the links were inserted intentionally as disinformation and what strategies could be used to limit these links.S

A good month for Annie Rauwerda's "Depths of Wikipedia"

Starting in 2020 as a "pandemic project" Rauwerda posted odd Wikipedia articles on her Instagram, Twitter, and TikTok accounts under the title Depths of Wikipedia. The project grew until now she has "about a million followers".[1] In March people really began to notice her with an article she wrote for Slate on Russians downloading complete versions of Wikipedia, followed by an article she wrote about Reddit, and the New York Times wrote [open-access link through subscriber referral] about her in a long article on March 31. The Times followed up with a student opinion on April 12. The New Yorker then gave her first try at a live comedy show a plug in their new issue. The recent show in Manhattan featured three comedians, Rauwerda as MC, and a sold-out crowd of 150. Add in a new Wikipedia article on Depths with a DYK on the main page. Several other newspaper articles piled on including one in the University of Michigan's The Michigan Daily. She'll be graduating from U of M with a degree in neuroscience in a couple of weeks. Check back here next month for more strange and exciting stories.

Rauwerda modestly told The Signpost "I'm still pretty surprised that any of this is news. All I do is take a bunch of screenshots. But hey, if another publication wants a feel-good story with goofy fun facts, I'll happily oblige. By the way, I think Wikipedians are the best (both on and off Wikipedia). Thank you all for always being so nice to me!" S

Did a Wikipedian shoot up a D.C. school?

A shooter opened fire on a school in Washington D.C. at about 3:20 pm on Friday, April 22, as classes were letting out for the weekend, injuring four people, including one minor. No deaths were reported until later in the evening, when a suspect was reported as having possibly committed suicide, according to several articles published in The Washington Post. The suspect was found dead in a fifth-floor apartment across from the school. He was surrounded by a tripod, ammunition, and several guns, which may have been used to shoot down at the school, according to the Post.

The Post also reported that the suspect was active on social media, 4chan, and Wikipedia, soon after the shooting began. He had previously edited Wikipedia articles on two other shooting incidents.

A thread at the administrators' noticeboard shows that the newspaper report had been noted; a user account, which was less than two weeks old, had been blocked and renamed; and several edits by the user had been removed from the article about the school.

The edits that were oversighted included one from four days before the shooting and three starting about 40 minutes after the shooting.S

In brief

Marysville, Washington, 25 miles (40 km) north of Seattle, known for its strawberry festival and, now, a certain Wikipedian

Notes

  1. ^ Sun, Michael; Gorman, Alyx (20 April 2022). "Mining for gold in the Depths of Wikipedia. Plus: what happens if you stop looking at your phone?". Saved for Later (podcast). The Guardian (Australia). Retrieved 20 April 2022. 8:15-35:20
  2. ^ Kavanaugh, Ryan (March 29, 2022). "Ryan Kavanaugh: Wikipedia Has Become Anti-Democratic – Op Ed". ExecEdge. Archived from the original on March 30, 2022. Retrieved March 30, 2022.



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


+ Add a comment

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.
  • (From Meta) Meta AI develops a novel dataset and model to help bring more representation to Wikipedia. Sorry, what? When did they ask the community whether we wanted their help and in what form? If we are to start using AI to generate baseline text to create WIR stubs from then I want it to be written by experienced Wikimedian volunteers, open source, and nothing to do with a for-profit company that engages in inordinate privacy violations and human rights abuses. (See also algorithmic bias.) — Bilorv (talk) 22:15, 24 April 2022 (UTC)[reply]
  • The paper is here. They do mention WiR a few times, but it doesn't sound like they made any attempt to get in touch. Overall, it does not seem particularly impressive to me − generative language models are good for a lot of stuff, they're very interesting, and they have a lot of potential, but wrapping them up in a soup of "AI" buzzwords and glowing generalities about "we made an AI to write articles" doesn't really convey what is going on. Essentially, their model does web searches and retrieves a bunch of information, which it kind of half-assedly stitches together into text resembling a Wikipedia article. But a lot of the information is nonsensical or incorrect (see page 7: "Overall, 68% of the information in generated sections is not present in the reference text"). Of course, basically all generative language models do stuff like this, Facebook has not done anomalously badly in this respect. But it'd be a lot more impressive if they weren't trying to pass it off like this thing was creating usable articles that didn't need human review. A lot of this stuff reminds me of when ASIMO would fall on its ass. jp×g 23:33, 24 April 2022 (UTC)[reply]
    • I'm reminded of the "Quicksilver" project from 2018 that kind of faded away. XOR'easter (talk) 15:35, 25 April 2022 (UTC)[reply]
    • Thanks for doing the deeper reading, JPxG. This is actually quite reassuring to me as it confirms that the Meta nonsense is just like their recent rename from "Facebook": a shallow facade of rebranding with no material substance behind it. They are simply exploring another avenue to piggyback off our labour at the expense of our reputation without any actual engagement in our movement—like the auto-generated Facebook pages they've had for many years that copy Wikipedia en masse with little or no attribution. Here it is wrapped in a laughable veneer of progressivism, with 32%-accurate articles made by an AI that is not actually available to Wikimedians, nor useful if made available, nor desirable even if improved based on better people that have tried and failed in this area. And this helps women how? — Bilorv (talk) 20:32, 25 April 2022 (UTC)[reply]
  • As for the "Wikipedian" shooting up a school, I see no reason to avoid mentioning the notable Edmund Burke School, which was the target of the attack as reported by the attacker himself, and countless reliable sources. I would not have chosen to call the suicidal gunman a "Wikipedian" because, to me, that term (which I rarely use myself), describes a person committed to Wikipedia's basic principles. Although this person's edits have been rightfully suppressed, it seems clear that they were in no way and at no time a good faith contributor to this encyclopedia. I think that "recently registered Wikipedia editor" would have been a more accurate and neutral description of the gunman. In the spirit of full disclosure, I have edited the school article to add neutral content about the shootings, and I was contacted in advance about this Signpost article, which I appreciate. Cullen328 (talk) 06:02, 25 April 2022 (UTC)[reply]
  • Honest question: Are Larry Sanger's statements still newsworthy? At this point, it seems that he bashes Wikipedia in some outlet every three weeks, usually repeating the same arguments over and over. IMO, it looks like Sanger is only trying to stay relevant instead of making an honest effort in improving Wikipedia or the Internet at large (not to mention some of his more dubious projects). Applodion (talk) 08:56, 25 April 2022 (UTC)[reply]

non-random break

  • Ryan Kavanaugh's complaints as described don't match well with what he actually said due to lack of context. His proof for his "fantasy world" is "They carry alternative titles, such as “Most Perfect Tutnum” and “Grand High Togneme Vicarus,” - which everyone knows are just jokes, no one takes them seriously (and yes, I have one, just for the fun of it). He also claims falsely that "The coterie of editors who gain more and more access by the number of contributions they have made to Wikipedia and the length of time they have been on the site. They are ranked on multiple levels, from “Registered User” to “Admin User.”" The only limitation to access is of course our extended confirmed protection policy. And of course it leaves out the fact that he has a grudge against us because we insist on reliable sources which has caused him problems verifying his age and other issues. His comments about paid editors are about his article again. He's claiming undisclosed paid editors are paid to dislike him. I loved his claim " You must counter by paying a more senior editor to fix the damage." Signpost, this is disappointing. Doug Weller talk 14:13, 25 April 2022 (UTC)[reply]
    • I wish I could get paid to dislike people; the world gives me plenty of opportunities to practice, so I'm awfully good at it. XOR'easter (talk) 15:18, 25 April 2022 (UTC)[reply]
      • If I had been paid for my Wikipedia contributions every time somebody accused me of being motivated by personal dislike, I'd be quite rich. But "personal dislike" is rather by-the-by when all I'm doing is reading sources and collating what they say (which is very rarely what I think). — Bilorv (talk) 20:32, 25 April 2022 (UTC)[reply]
    • @Doug Weller: your complaint IMHO boils down to "a lack of context" which I partly consider justified, but mostly consider to be a misunderstanding of what the "In brief" section of this column is trying to do. We can't give all the facts complete with context for all the interesting stories we see in the press about Wikipedia. Why did I find this op-ed interesting? And why did I suspect many of our readers would find it interesting? Well - he nails down the accusations that many folks who try to get an article in WP the first time. Common accusations about Wikipedia are an "exclusive club" of arrogent admins and other editors who live in a 'fantasy realm they have created" along with paid editing (or possible bribery), and the high barrier to editing. He very accuratly *listed the usual accusations*. Now how accurate are the accusations? Well, a lot of it has to do with the people coming into Wikipedia for the first time. With their expectations being completely off, these are probably the best explanations they can come up with for what the see. They'll almost never come up with the "true reason" - that they are just not notable. A large portion of the blame for this type of common reaction to Wikipedia lies with ourselves and especially with WMF communications. We need to let these misguided folks know what to expect ahead of time. If we don't there will inevitably be a lot of folks who get pissed off about how they are treated here, There's another group. IMHO possibly typified by Kavanaugh who seem to know what to expect, but want to get a free personal advertisement As far as I can tell our admins dealt with him in the proper way. Now, if you as a reader. want to learn more about this phenomena you can't expect The Signpost to say all this in the one paragraph we have available. But you can click through to the original article, which looks pretty transparent to me. You might learn something. Smallbones(smalltalk) 19:26, 25 April 2022 (UTC)[reply]
      As Offred says in The Handmaid’s Tale, “Context is all”. And of course I read the original article. Doug Weller talk 19:41, 25 April 2022 (UTC)[reply]
      I think part of what makes the general public so perplexed by our process is that "notable" to us has a narrowly defined, sometimes rigorously applied definition. Whereas to most people it means something rather different – maybe "accomplished in his/her/their field" for professionals, or "much talked about on social media" for popular subjects. Which definitely don't meet WP:GNG alone. In other words, our quasi-academic, textual-source-based and bureaucratic process is is alien and confusing to outsiders. The consensus-based stuff with distributed responsibility just adds another layer of confusion. ☆ Bri (talk) 03:25, 26 April 2022 (UTC)[reply]
      The general public are almost wholly unaware that we have any limitation in scope: this is mostly flattering, as it indicates that readers really struggle to find a topic that we don't have information about. Unfortunately, this assumption that all information is welcome on Wikipedia leads to misplaced efforts by newcomers who think that any piece of information will be appreciated. In the nicest case, this is somebody doubling the length of a plot summary of a movie that was already too long. In the most confrontational, it's somebody adding a fact about a politician and then assuming that the removal must be politically motivated. — Bilorv (talk) 15:56, 26 April 2022 (UTC)[reply]
      Well, in fairness, we do tell them it's "The encyclopedia anyone can edit". ☆ Bri (talk) 16:54, 26 April 2022 (UTC)[reply]
      When someone expresses a desire to be "in Wikipedia" my answer is, "Well, there are often difficulties and you appear to have one of the greater ones. Far as I can see you have, though no fault of your own, the misfortune of being a living person. For many purposes being a living person is a splendid thing, but for becoming the subject of a Wikipedia biography it's a handicap that most people cannot overcome." Two days ago, however, I found more effective the answer, "The most sure-fire ticket for becoming the subject of a Wikipedia article is an obituary in the New York Times. Not a paid death notice; a genuine obituary." Jim.henderson (talk) 16:10, 30 April 2022 (UTC)[reply]

















Wikipedia:Wikipedia Signpost/2022-04-24/In_the_media