This is not one of those “superintelligence is just around the corner” pieces, but rather asks if maybe we’re already learning too much about our own brains, demystifying thinking and its implications. You can’t know too much, can you? (James Somers, The New Yorker)

New paper: Which sources does ChatGPT provide in Germany when asked for the latest news? Main takeaway: Very different results depending on whether you use the customer chat interface or the API. Customers get more results from content partners, the API is more diverse or even fringe.

They speak our language, but it does not appear to constrain their thought: “AIs tend to express liberal, secular values even when asked in languages where the typical speaker does not share those values.” Fascinating research by Kelsey Piper.

The Journalism Benchmark Cookbook: We prototyped a benchmark evaluating the task of information extraction in journalism. (Charlotte Li, Jeremy Gilbert, Nicholas Diakopoulos, Generative AI in the Newsroom)

What if we let ChatGPT, Claude, and Gemini cosplay as famous authors? Would anyone notice? Would critics swoon? Would readers care? Spoiler alert: finetuning works really good, a new study finds. (Rosalia Anna D’Agostino, LinkedIn)

Generative AI and news report 2025: “We document rapid growth in the use of a new set of tools bound to impact the discovery of information, how people use it, and, by extension over time, changing competition for both attention, advertising, and the money people are willing to spend on media and content.” (Felix Simon, Rasmus Kleis Nielsen, Richard Fletcher, Reuters Institute)

Can AI solve the world’s problems, like clean drinking water? “A superintelligent machine may be able to understand these problems, but it will have no way of overcoming them,” writes Francis Fukuyama. “There are water mafias that buy water where it is cheap, and resell it at extortionate prices. They are armed and ready to use violence if you get in their way.”

AI Search, Users, and News: A trove of data from LM Arena offers a glimpse into user search behavior. A few sources garnered the majority of impressions. (Nick Diakopoulos, Generative AI in the Newsroom)

Werewolf leaderboard: GPT-5 is the best at bluffing and manipulating the other AIs in Werewolf. (Foaster Labs)

Users of German news site Süddeutsche who were shown a difficult quiz about AI-generated images afterwards trusted media less and visited the news site a little bit more often – and now everyone’s hoping that quality journalism still has a chance. (Sarah Scire, Nieman Journalism Lab)