Context engineering, how to ask, and AI agent interfaces
Read on the web This newsletter contains: how to organize instructions for LLM agents (and why less is more), two classic communication anti-patterns, and a...
Preparedness paradox, research advice, and open source coding agents
Read on the web Three things this time: why good prevention ends up looking wasteful, a compilation of research advice, and a recipe for running a coding...
Measuring disruption and Claude Scholar plugin
Read on the web Two things this time: a new paper on measuring scientific disruption, and a Claude Code plugin for academic workflows. Measuring disruption...
LLMs and inferential privacy, How to do peer review, cmux, etc.
Read on the web Here's a round-up of some recent things. LLMs Can Infer Political Alignment from Online Conversations — A new paper from our group. We show...
Migration & updates
Hello! 👋 This is my first email from Buttondown. I've just migrated from the Substack, for various reasons. If you're reading this, your subscription carried...
