Knowledge graphs are powerful tools for organizing and connecting information, but building them can be a tedious process. Imagine having to manually extract facts from countless documents—it's a knowledge engineer's nightmare. Could large language models (LLMs) offer a solution? Recent research explores whether LLMs can automatically populate knowledge graphs, potentially revolutionizing how we build these complex networks of data. The study focused on the Enslaved.org Hub Ontology, a knowledge graph documenting the narratives of enslaved people. Researchers experimented with different LLM prompting strategies, including text summarization and a technique called Retrieval-Augmented Generation (RAG), where the LLM accesses an external database for relevant information. The results were promising. When guided by a modular ontology, the LLMs extracted almost 90% of the correct information, demonstrating their potential to automate this traditionally laborious task. While LLMs are undeniably faster than humans, the research raises critical questions: Does speed compensate for occasional factual errors? And how can we optimize the structure of knowledge graphs to make them easier for LLMs to understand? This research highlights the exciting potential of LLMs to streamline knowledge graph creation, opening doors to building more comprehensive and interconnected data resources. While challenges remain, this study represents a significant step towards automating a crucial but often bottleneck process in knowledge engineering.