GLM-5 is a new model from Z.ai, boasting 754 billion parameters and a size of 1.51 terabytes, making it twice as large as its predecessor, GLM-4, which had 368 billion parameters and was 717 gigabytes. This MIT-licensed model reflects a significant advancement in the capabilities of large language models (LLMs). The article highlights the emergence of the term "Agentic Engineering," recently championed by figures like Andrej Karpathy and Addy Osmani, to describe the role of professional software engineers who build applications using LLMs. The author shares a firsthand experience with GLM-5 by running a prompt to generate an SVG image of a pelican riding a bicycle. The result was a well-rendered pelican but a less impressive bicycle frame, illustrating both the model's strengths and limitations. This hands-on test hints at the practical applications and ongoing challenges in utilizing such advanced AI models for creative tasks. The article also references recent developments in the AI field, including articles on OpenAI's evolving mission, the introduction of tools like Showboat and Rodney for demoing agents' creations, and insights from StrongDM's approach to software development without direct code interaction. These examples reflect the dynamic nature of AI and its growing integration into software engineering practices.