Developer Lyra Rebane has created Xikipedia, a web app that transforms Wikipedia entries into a social media-style feed. The app pulls information from Simple Wikipedia and presents it in a scrollable format, allowing users to like posts and customize their feeds based on categories. The goal is to offer an alternative to potentially negative social media experiences.

Xikipedia's algorithm learns user preferences without collecting or sharing data, running locally and deleting data upon refresh. Users can explore content from various categories, and liking a post increases the likelihood of similar content appearing. The app aims to provide a curated, engaging experience for discovering information, similar to how platforms like Reddit or Twitter work.

While Xikipedia offers an intriguing way to explore Wikipedia, it's worth noting that it may contain NSFW content due to the random article selection. Also, there might be a delay while the 40MB data loads. This project is a demonstration of how a basic algorithm can tailor content based on user interaction.

Ultimately, Xikipedia presents a fresh approach to consuming information, possibly reducing the negative aspect of doomscrolling. It’s a testament to the power of open-source knowledge and algorithmic curation. The future may hold more niche social media experiences focusing on curated content.