#AI

5 items

Digital Folklore

May 20, 2026 · 17 min read

Most of the internet is talking to itself — the folklore of the ghost web.

Between 30 and 50 percent of all web traffic is not human. The internet built for people is now mostly machines talking to machines — and has been for years.

Read article

Digital Folklore

Apr 23, 2026 · 8 min read

The Invisible Persuader: How AI Became the Internet's Favorite Explanation for Everything

The belief that AI is secretly managing your opinions has become one of the internet's most viral ideas. It spreads through the same feeds it claims to be warning you about.

Read article

‘BLOCKADE’: The Right Is Using AI Content Scanners to Try to Supercharge Book Banning

Someone sat down, thought real hard, and came up with the acronym BLOCKADE — Blocking Lustful Overzealous Content, Keeping Away Depravity and Extremism — which tells you everything you need to know about the people running this operation before you read a single word of the actual article. The program feeds books into an AI and scores them against a list of roughly 300 words ranked by how offensive they are to conservative values, which is a definition the AI itself gets to decide, because apparently the humans involved ran out of opinions halfway through and handed the rest off to a chatbot. One school district in Texas has already pulled over 1,400 books from its shelves while the budget to buy new ones sits frozen, which is the kind of number that tends to get lost in the discussion about whether the algorithm is biased. The whole enterprise is being sold back to the same districts it helped overwhelm, at five dollars a scan, which is either the most efficient grift of the decade or proof that someone out there has been paying very close attention to how this country works.

This Company Is Secretly Turning Your Zoom Meetings into AI Podcasts

WebinarTV, a company apparently founded on the principle that nobody has suffered enough, has been crawling the internet for Zoom meetings and turning them into AI-generated podcast episodes — and before you ask, no, they didn't bother telling most of the people in those meetings. The organizer technically opts in, which is a real cute way of saying everyone else in the room gets volunteered without their knowledge, including people in therapy sessions and job interviews, two situations where "surprise, you're now a podcast" ranks pretty high on the list of things you didn't need. The AI spits out a synthetic host who recaps the whole thing in that relentless upbeat tone that makes you want to put your fist through a wall, and by default the finished product is available to anyone with the link. It's the kind of innovation that really makes you step back and appreciate how nobody asked for this.

Wikipedia Bans AI-Generated Content

Wikipedia has updated its editorial guidance to prohibit AI-generated text from being added to articles, citing concerns about accuracy, sourcing, and the erosion of the editorial voice that makes the project legible as human-made knowledge. The ban doesn't come with a detection system — it relies on editors to identify and remove machine-generated content the way they always have, through community vigilance and dispute. What's notable is the framing: Wikipedia isn't calling AI unreliable so much as calling it unaccountable. The argument is less about what AI gets wrong and more about who is responsible when it does.