TickerNews is a swipe feed for news and learning: world headlines alongside bite-sized micro-lessons in machine learning, physics, finance, philosophy, and literature, regenerated every day.
TickerNews is built and run by Sean Lyons as a solo project.
The site runs on an automated pipeline. News cards are collected from public RSS feeds; every card names its source and links to the original article.
Educational cards are generated by a small open-weights language model (Phi-4-mini) grounded on Wikipedia extracts: the model writes from a source extract and runs a factual self-check against it, and claims the check can't support are dropped. The educational feeds are refreshed daily.
Generated content can get things wrong. If you spot an error, email hello@tickernews.ai and it will be fixed.
TickerNews is free to read. A tip-jar link will live here soon — if you'd like to support the project, check back shortly.