Knowledge engineering, the art of capturing and structuring information in a machine-readable format, has always been a meticulous process. Think of it as building a vast, intricate library, where every book, shelf, and category needs careful consideration. But what if you had an AI assistant to help you categorize, cross-reference, and even suggest new additions to your collection? That's the promise of using Large Language Models (LLMs) in knowledge engineering. A recent study explored how knowledge engineers are using these powerful AI tools to transform their workflow. Researchers observed knowledge engineers during a hackathon, and through interviews, they uncovered both the exciting potential and the complex challenges of this new frontier. One key takeaway is the importance of "prompting" – crafting specific instructions to guide the LLM. It turns out, asking the right questions is crucial for getting valuable results. Imagine asking your AI librarian for a specific historical fact versus asking it to summarize a complex philosophical concept. The way you phrase your request drastically impacts the quality of the response. However, even with the right prompts, evaluating the output of LLMs poses a significant hurdle. How do you ensure the AI is not just making things up or perpetuating existing biases in the data it was trained on? Traditional metrics like precision and recall are not enough. Researchers suggest we need new benchmarks and evaluation strategies specifically designed for this task. One intriguing proposal involves using "adversarial algorithms" – essentially, another AI that tries to break the first one. If the knowledge-building AI can withstand this kind of scrutiny, we can have more confidence in its output. But ultimately, human oversight remains essential. LLMs are powerful tools, but they are not infallible. Knowledge engineers need to develop new skills, not just in prompting, but also in critically evaluating the AI's contributions and ensuring they align with ethical considerations. This research paints a picture of a field in transition. LLMs offer the potential to automate tedious tasks, accelerate knowledge gathering, and even uncover new insights. But realizing this potential requires careful attention to evaluation, bias detection, and responsible AI practices. The future of knowledge engineering looks bright, but it will be a collaborative effort between humans and their increasingly intelligent AI assistants.