I'm still hoping for AI agents to mature to a point where they can be universal scrapers for my RSS. Have a headless client, scraping, interpreting websites in the background... burn excess cpu cycles and dead dinosaurs to replicate universal RSS dream.