Everyone blames screens for the decline of reading. Nobody blames the consolidated corporate publishing industry that made books pricier, more generic, and harder to find. jacobin.com/2026/10/books-pu…

You can spend ONE afternoon with: • Gottlob Frege • Bertrand Russel • Ludwig Wittgenstein • W.V.O. Quine Who are you choosing?

I’m starting to get increasingly uncomfortable with the economics of AI as it is currently happening It seems fundamentally wrong that a small number of people with access to a huge amounts of VC money can vacuum up all the contributions of humanity, often paying nothing for it, then turn around and get enormously rich while the many people who put all the work into creating the training data get nothing I am generally pro-AI, but the way it is happening is seeming increasingly parasitic to me

We don't have a reading crisis. A lot of people are still reading, but they tend to read romantasy porn or self-help books—or, perhaps, Sally Rooney novels. What we have, my friends, is a seriousness crisis. There is, indeed, a difference between reading profound classic literature and reading contemporary slop, and it’s time we admitted it.

Prediction: no one will care about journal names in 5 years.

Can you tell if the ideas in a piece of writing are from a human or an AI? As AI writing becomes more commonplace, emerging policies on AI use increasingly focus on who came up with the ideas in the text rather than who wrote the text. Modern AI detectors such as Pangram answer the latter question but not the former. Introducing IdeaLens, a detector of AI ideas!

We represent documents as outlines of their ideas. Starting from 1M web-scale documents with prose authorship labels from Pangram 3, we construct WildOutlines, a paired corpus of documents and their outlines. We train IdeaLens on these outlines after paraphrasing to fit the Pangram labels. Paraphrasing removes any remnant stylistic cues from the source document forcing IdeaLens to focus on the underlying ideas and not surface-form. As a control, we also train ProseLens with the same backbone but on the raw document texts.