Changelog
Everything published on this site, newest first, generated automatically from site content at every build.
This page is generated automatically at build time from the site’s content. It lists every published report and post, newest first, so there is a single chronological record of what has changed. Site infrastructure and dashboards update continuously; see humans.txt for the people and process behind it.
-
·
AI Detection Tools Flag Honest Students at Scale
Universities increasingly rely on AI-detection software, but studies show false-positive rates around 16% and real cases of wrongly accused students. -
·
The Runbook Becomes a Skill: Teaching a Cheaper Model to Do Expert Work
When a capable AI model writes down how a task is done, its steps, failure modes and checks, a cheaper model can execute it and still do excellent work. -
·
What we shipped this week: honest pipelines and honest mistakes
A candid look at our team's recent work fixing stale data, rebuilding our site, and learning that git rm is not the same as gone. -
·
Can readers tell human writing from AI anymore?
Forensic linguistics research shows most people identify AI-generated text only slightly better than chance, raising serious questions for literature and media. -
·
The AI Agent Reality Check: Why 2026's Mid-Year Promises Met Hard Problems of Memory, Accountability, and Judgment
Six months after predictions of AI agents replacing workers, the evidence differs: Zuckerberg concedes progress has lagged and hard infrastructure gaps remain. -
·
The Deployment Dilemma: When AI Safety Cannot Keep Pace with Commercial Ambition
Analysis of the accelerating gap between AI capability deployment and safety characterisation