The text discusses the current state of NoSQL databases, their data models, and scalability. It highlights four main categories: key-value stores, BigTable clones, document databases, and graph databases. The author argues that while key-value stores and BigTable clones excel in scaling to size, they often struggle with complexity due to simpler data models. In contrast, document databases and graph databases have richer data models but may face scalability challenges. The text emphasizes the importance of choosing a data model that fits the specific needs of an application, rather than relying on generic solutions. It also notes that NoSQL is not just about performance and scaling, but also about providing a good fit for modeling business domains.