AI Radar

Noteworthy AI conversations and writing, summarized from publisher notes.

About AI Radar

AI Radar tracks what the people shaping AI actually say. It follows a short list of lab leaders and researchers across 31 podcast, video, and newsletter sources, and publishes a summary whenever one of them is a real guest, speaker, or author.

How items are selected

Every hour, a worker reads the configured feeds. A model judges each new item against one rule: a followed person must take part. Mentions, third-party analysis, news roundups, and colleagues at the same company do not qualify. A local check then confirms the named person appears in the publisher’s own title, byline, or notes before anything is published.

How summaries are written

Summaries are written by AI from publisher notes, such as episode descriptions. AI Radar does not use transcripts, so a summary describes what the publisher says an item covers and is not a verbatim record. Every summary links to the original, which is the source to rely on. Items with sparse notes wait until richer notes arrive.

What gets published

New items appear on the front page the day they are found. Each day is kept in the archive, and a daily RSS edition goes out in the evening. There are no ads or tracking scripts.

The code and data are public on GitHub.