About

Weekly Content Gathering

Every Wednesday morning, an automated pipeline scans the web for new content about AI in education using the Brave Search API. It runs 19 targeted queries across four source types:

• ▸Broad topic searches — general queries like “AI in schools” K–8 policy and “AI tools” students elementary middle high school, including content from .edu and .gov domains
• ▸Trusted mainstream outlets — targeted searches of Education Week, Chalkbeat, The 74 Million, EdSurge, Edutopia, KQED, NPR, the New York Times, the Washington Post, the Hechinger Report, The Atlantic, CNN, AASA, ISTE, NAIS, ASCD, Common Sense Education, and K-12 Dive
• ▸Policy and government sources — the U.S. Department of Education, the White House, NASBE, CCSSO, NCSL, the Education Commission of the States, and New Jersey state government sites
• ▸Independent and practitioner voices — Substacks and blogs from educators and ed-tech thinkers who are often ahead of the mainstream conversation, including Eric Hudson, Stephen Fitzpatrick, Leon Furze, Stefan Bauschard, Johnny Chang, Tom Daccord, Michael B. Horn, Lance Eaton, Claire Zau, Matt Miller, Connected Learning, MSA Evolution Lab, Wes Fryer, Ben Williamson, and Audrey Watters. Know another good one? Submit it here.

Each weekly run returns up to 380 candidate articles, which are then passed through an AI filtering layer.

AI Relevance Filtering and Classification

Every candidate article is evaluated by Claude — Anthropic’s AI model, the same family that powers Claude.ai — before anything goes into the database. Claude reads each article’s title, URL, and description and answers two questions: Is this genuinely about AI in K–12 education? And if so, how should it be classified?

Claude assigns each passing article:

• ▸Category — National News, Local/Regional News, School AI Policies, Tools & Products, Analysis & Commentary, Perspectives, or School Spotlights
• ▸Geography — Local NJ, Northeast, National, or International
• ▸Grade level — PreK, K–8, High School, or General
• ▸A short summary of what the article covers

Articles that don’t pass the relevance check are discarded, keeping the database focused and free of tangentially related content.

AI-Generated Summaries

After each weekly import, Claude synthesizes the newly classified sources into two types of summaries:

• ▸Category summaries — For each of the seven content categories, Claude distills the top articles into 3–5 bullet points covering key themes, trends, and examples. These appear as summary blocks on the homepage.
• ▸Overview summary — A cross-category synthesis, structured in sections, that provides a bird’s-eye view of where AI in education stands right now. This is the first thing you see when you visit the site.

Both summary types are regenerated each week alongside each new import, so they always reflect current sources.