Current discussions of platforms as cultural infrastructure emphasise extractive, exploitative and toxic corporate systems in which readerly practices are captured, processed and monetised through platform governance and generative AI systems. This focus obscures the existence and potential of community-built or public social infrastructures that organise and enable cultural engagement and interpretation in different ways including through care-based, non-market and collectively-governed models, highlighting public access and value over extraction, surveillance and monetisation. Taking digital reading platforms as an example, this paper asks how algorithmic reading transforms books and their social interpretive environments into datasets, reshaping authorship, cultural circulation, and cultural value. It theorises the connection between reading and scraping to describe how cultural engagement is reconfigured as socio-technical practices involving readers, platforms, data infrastructures and AI systems, and examines how competing infrastructural logics redistribute interpretive authority within contemporary digital arts and culture environments.