Concerned that you or your loved ones might be participating in a massive de-skilling event? Experiencing LLM-induced psychosis? Outsourcing cognitive and emotional functions to autocomplete?
Install SLOW LLM. (Sam Lavigne, via Matt)

Anthropic asked 81,000 Claude users what they want from AI. People in lower-income countries are more optimistic than those in Europe or North America. Worry about job loss is the clearest predictor of negative sentiment. Qualitative research at this scale was hard to do before. Whether Anthropic should be the one doing it is a question the report doesn’t ask.

Google is rewriting news headlines in search results. “Copilot Changes: Marketing Teams at it Again.” That’s not a headline a journalist wrote. Google did. Without telling anyone. (Sean Hollister, The Verge)

Journalist suspended for publishing AI-hallucinated quotes: A former editor-in-chief of NRC used ChatGPT, Perplexity, and NotebookLM for his newsletter, but didn’t verify quotes. He had warned colleagues about exactly this. (Dan Milmo, Guardian)

A 47-step AI tool that writes nothing: Bauer’s internal content briefing system pulls articles from competitors, looks at rankings, and identifies content gaps. The journalist still has to write the piece. “The premium is on expertise and authority. We wouldn’t ever want to do anything which dilutes that.” (John Rahim, The Media Stack)

Schibsted's Videofy is now open source: The tool pulls a published article, writes a script, matches footage, adds a voiceover, and hands editors a finished video.

Image Verification Assistant: A joint CERTH-ITI and Deutsche Welle tool for image verification, metadata analysis, and reverse image search. Alpha stage, open source.

Three roles for AI in journalism: source, colleague, assistant. A framework that’s more useful than another round of newsroom culture war. (Stephen J. Adler, Columbia Journalism Review)

The first white-collar job that AI can actually replace is the one that built the AI. Now coding is conversational. One dev is pleading with his chatbot: “Pushing code that fails pytest is unacceptable and embarrassing.” (Clive Thompson, New York Times)

Before you download a 35 GB model and watch your laptop give up: CanIRun.ai is a reality check for the local-AI-curious. Yes, there are options for an M2 with 8 GB RAM. Llama 3.1 8B is a “tight fit”, but Qwen 3.5 2B “runs great”.

Can you tell which passage was written by AI? This New York Times quiz is humbling either way. (Kevin Roose and Stuart A. Thompson)

Grammarly “cloned” Julia Angwin, Stephen King, and Neil deGrasse Tyson as AI editors, without asking. Angwin is now suing. The feature is gone, the apology is filed, and the AI Julia apparently gave bad advice. (Miles Klee, Wired)

34 projects, 24 contributors, 65 minutes: Hacks/Hackers held a vibe coding show & tell, and journalists shipped fact-checkers, salary surveys, power grid monitors, and a Discord clone.

An editor gets a promising pitch, starts googling, and finds a byline that exists everywhere and nowhere. A podcast episode about a scammer – and that some sources didn’t care about being faked. (Question Everything)

A young reader got so frustrated with the FT’s news product that she built her own using Claude Code. She’s paying for a subscription, she wants personalized news, and she couldn’t figure out how to get it. (Jodie Hopperton, INMA)