Imagine a world where AI could design complex systems, from intricate microchips to innovative spacecraft. Large Language Models (LLMs) are one step closer to realizing that vision, but they face a unique challenge: understanding causality. LLMs excel at generating human-like text, but they often struggle to grasp the underlying "why" behind engineering principles. How can an AI truly design if it doesn't comprehend the cause-and-effect relationships that govern how things work? Researchers have been exploring this puzzle using a model of causality called SAPPhIRE, a framework for representing how systems function. The key lies in providing LLMs with the right kind of knowledge. A recent study delved into how providing LLMs with different types and levels of technical "context" impacts their ability to generate designs relevant to the SAPPhIRE model. The results are fascinating: the type and detail of background information significantly influence an LLM's understanding of the scientific principles it needs to generate accurate designs. Some contexts work better than others, like giving an AI the perfect textbook before asking it to solve a problem. Interestingly, certain ways of phrasing engineering concepts resonated more effectively with the LLM, demonstrating the importance of not just what information we give an AI but how we frame it. This research opens exciting doors for AI-assisted design. Imagine engineers partnering with LLMs to quickly generate and test countless design options, grounded in a true understanding of physics and causality. While further refinement is needed, this work marks a critical step toward unlocking the full potential of LLMs in engineering and design. The future of product development may be a collaboration between human ingenuity and the analytical power of artificial intelligence.