This module is part of a Micron blog series “Understanding Artificial Intelligence (AI) concepts through the lens of memory.” The featured article, “What does it mean for AI to ‘know’ something?” explores how large language models (LLMs) generate responses by calculating mathematical relationships between words encoded in high-dimensional embedding space. Readers learn foundational concepts like tokens, embeddings and embedding space to understand how AI systems approximate meaning based on statistical relationships in language. By understanding this distinction, readers can develop more nuanced approaches to AI-generated outputs. The article also touches on the key role of memory systems as part of the hardware architecture that shapes what AI systems can store and process. This memory perspective introduces valuable context for understanding both the capabilities and constraints of modern AI systems. Ultimately, readers are invited to engage with AI more thoughtfully by illuminating the gap between statistical pattern recognition and genuine knowledge. A quiz is also included to support educators incorporating this material into their curricula.