Damn. This is amazing. £325 per week, paid monthly, for 3 years - and the result was a profit for the Irish economy:
www.citizensinformation.ie/en/employmen...

thanks to @jonathancardy.bsky.social for the interview

“I’m a lycanthropic spell checker.” That’s how Wikipedia editor WereSpielChequers explains his username.

In my latest newsletter, we talked about easily confused words, dead links, and the election process for the Wikimedia Foundation’s board of trustees.

I've written before about Wikiracing, but it's not the only wiki-based game.

The WikiGameJam is a WikiHaus hackathon happening this weekend (Oct 3-5) at Hex House in Brooklyn: wikigamejam.org

They'll be playing WikiAsteroids, WikiTrivia, and hopefully designing the next gen of wiki-powered games.

The biggest Wikipedia news this week? Amazon sold out of “The Editors.” Don’t worry, it’ll be back in stock soon and in the meantime, you can grab a copy other places.

Wikipedia editors have put together a guide on signs of AI writing:

-Tone: vague attribution of opinion ("has been described as")

-Style: the now infamous em dashes; excessive use of boldface

-Communication intended for the user: "I hope this helps," "You're absolutely right," sycophantic chatter

no one tells you that a huge part of being a Wikipedia beat journalist in 2025 is people emailing “the Wiki page for Tylenol is really weird right now,” detailing the debate over disambiguation links, then acting surprised you haven’t landed the Wikitylenol story on the front page of The NY Times

Altman’s enthusiasm for spiral notebooks is so endearing that it almost obscures the irony. His job is to promote the company’s AI tools to billions of potential customers, promising the transformative benefits AI. But when he does his own deep thinking, his style is decidedly analog.

Q: Why print photos (or anything else)?

A: A print is self-worth. It’s a statement that you’re putting on an image that you’ve created to make it manifest... It isn’t a thing you’re creating for likes and comments... It’s just exactly what it is. --Chris Gampat, The Phoblographer

"If we can't agree on basic facts, we can't self-govern. And that is what authoritarians count on." @rachelegoodman.bsky.social

The scariest part isn't the fake book titles generated by AI. It is the hostility patrons show librarians when they cannot find them. People now trust their sycophantic LLMs more real people.

www.404media.co/librarians-a...

Thomas Mallon with an unexpected benefit of keeping a diary.

“The strategy is clear: erode trust in Wikipedia and other independent knowledge institutions. Because once those collapse, the party in power can declare the Truth, no citation needed.”

My latest for Slate:

Excellent @stephenharrison.com piece about Wikipedia, breaking news, and the right wing outrage machine that's kicked into high gear around articles about Charlie Kirk and the killing of Iryna Zarutska.

slate.com/technology/2...

Thank you, Chris, for sharing my piece.

P.S. You've built such a cool community of readers, interested in both data-driven analyses and music history. Well done.

www.cantgetmuchhigher.com/p/the-surpri...

Because cynicism is the easier path, people assume Wikipedia editors title articles based on feelings.

But titles follow policy. Different rules apply depending on facts, like whether there’s been a conviction.

H/t to @molly.wiki for laying out the reasoning and pushing back on the bad-faith takes

Music writer Chris Dalla Riva has a theory that late ’80s pop songs left a weaker cultural footprint than early ’80s hits.

His proof? The size & page views of their Wikipedia articles.

I wrote about his investigation, KPop Demon Hunters, and the Wiki editors who write pages for #1 hit songs.

Knowledge acquisition used to be an automatic consequence of doing or researching anything, argues @jackraines.bsky.social, but thanks to AI, the end result is now decoupled from learning about what you've built.

What should you do? Write it down. Document the whole process for future record.

This article from @biblioracle.bsky.social explains why AI summaries can't match the power of reading the original.

Each of us is a "unique intelligence," and when that collides with the unique intelligence expressed in an author's full work, the collision of unique intelligences throws off sparks.

Last week, the leaders of Meta, Alphabet, OpenAI, and Apple took turns praising Trump at a White House dinner.

Not in the room: the leaders of Wikipedia, the volunteer editors who curate and maintain the internet's knowledge.

"Wikipedia has become the next target, and a more stubborn one... its contributors are difficult for any particular state to persecute... there's no government funding to cut off or advertisers to boycott. And it's so popular and useful that even highly repressive governments hesitate to block it."

quick notes from @biblioracle.bsky.social’s book:

-automation ≠ artificial intelligence
-writing is thinking
-writing is feeling
-framework for thinking about AI: resist, renew, explore
-remember that we are embodied creatures, not machines

Good point- the contributor must be thoughtful & tech-savvy to appropriately bulk upload to Commons.

I was thinking of my early, somewhat clumsy coverage of prolific contributors like Ser Amantio who use automated tools. That’s why I appreciate how clearly Vera describes automation in her post.

Appreciate Vera's honesty here - Of the 86K media files she has added to Wikimedia projects, most came from bulk uploads of sets that are in the public domain/freely available.

But 7,500+ of the files are Vera's own work. She goes out with her press pass, documenting the world for Wikipedia.