Hybrid perovskites have emerged as a group of semiconductors that can solve key problems involving efficiency and production in optoelectronics and spintronics research. Over the past decade, this field has evolved to a point where the literature contains an enormous volume of chemical and physical information. The dispersed nature of the large, rapidly growing body of hybrid perovskite materials data poses a barrier to systematic discovery efforts, which can be solved by materials property databases, either by high-throughput or by systematic, accurate human-curated efforts. This article discusses the necessity, challenges, and requirements of building such data libraries. The level of detail captured in scientific publications still eludes the full grasp of generalized AI, but combining general AI with detailed curation may eventually and drastically help speed up data accessibility in the hybrid perovskite community and beyond.