My main newsletter, Complex Machinery, includes a section called "In Other News..." It's where I list one-liners about interesting articles that didn't fit into any segments.

You can think of this list as a version of In Other News, but with a wider remit than Complex Machinery's "risk, AI, and related topics."

Above the fold

1. Last week Anthropic made headlines by announcing its new model – dubbed "Mythos" – but refusing to release it due to safety concerns. Apparently it's been finding all kinds of vulnerabilities in software. While the claims of Mythos's capabilities don't seem too far-fetched, one wonders whether the announcement doubled as a well-planned PR stunt. (The Guardian, Le Monde 🇫🇷, Der Spiegel 🇩🇪, JP Morgan Private Bank, New York Times, The Register)
2. The Stanford 2026 AI Index report is out, and it highlights important trends in the genAI space. (Stanford HAI, Sherwood News, MIT Technology Review)
3. If you enjoy reading – and you clearly do, as you're reading a newsletter that points you to other things to read – you probably have a firm preference on paperback versus hardback books. And if you prefer paperbacks, you probably wonder why they arrive more than a year after the hardcover version of the same book. One UK bookseller did the legwork. (Tom Rowley)
4. Remember when federal law limited oil companies' liabilities in the event of a spill? OpenAI is walking that same path, trying to limit their liability in the event a model causes widespread harm. (Wired)
5. Some people assign a race to humanoid robots, based on the work being performed. (Scientific American)
6. The venerable Economist's heretofore faceless, pseudonymous writers may soon appear on video. (New York Times)
7. This is a tale about a company doubling down on credit card fraud. But it also reads like a lesson for genAI companies. (Ars Technica)

The rest of the best

• Lloyds of London – yes, that Lloyds, the massive insurance exchange – has let genAI into the board room. (The Times UK)
• Anthropic's Claude is getting worse. And end-users aren't happy about it. (The Register)
• Google's AI Overviews search summaries don't always pull from the most trustworthy sites… (New York Times)
• … which might be why Google's they're wrong. A lot. Even if they're only wrong a small percentage of the time, that still adds up to a lot of misinformation. (Futurism)
• Generative AI companies pour money into policy papers in an attempt to improve their image. (The Guardian)
• One estimate: genAI will impact half of jobs. (Les Echos 🇫🇷)
• In which companies learn a hard lesson about operations: the mountain of code coming out of genAI bots is overwhelming the human developers tasked with reviewing it all. (New York Times)
• Meta CEO Mark Zuckerberg is getting his own AI clone. (FT, Ars Technica)
• Someone allegedly tossed a molotov cocktail at Sam Altman's home … (The Guardian)
• … but at least one outlet offers a plausible reason. (The Onion)

Did I miss anything?

Have something I should read? Send the link my way.