When your news builds a memory
Pete’s personal newspaper turns accumulated coverage into a connected resource that his AI agents can consult for context without burdening him with another unread queue.
Developed from a conversation between Pete Winn, Anthony and Andy David
Pete moved his AI-generated newspaper out of a Flight Deck channel after the stream of unread stories began to feel like work waiting to be completed. The newspaper now appears in a separate feed where new items can surface like social posts. He can browse or dismiss them without every story becoming an obligation.
Behind the newspaper sits a graph database. When Pete’s pipeline finds and writes up a story, it identifies the topics, companies, individuals and other useful elements, then connects them to related nodes. After enough coverage has accumulated, a story about Donald Trump can lead to every other story involving him. Reporting on Nvidia’s AI development can reconnect with coverage of a particular data centre implementation.
The graph is intended for more than related-story retrieval. Pete’s AI agents can consult it when answering questions, checking what they already know about a subject and which earlier stories have interested him. That history gives later queries personal context drawn from material the agents have already read, while Pete remains free to treat the visible news feed as something to browse rather than a backlog to clear.