Evolution Strategies (ES), a powerful class of optimization algorithms, are like skilled navigators charting a course through complex landscapes to find the best possible solutions. But they often operate as a "black box," making their decisions opaque and difficult to understand. This lack of transparency can limit their use, especially in critical applications where understanding *why* a solution was chosen is as important as the solution itself. New research is shedding light on these enigmatic algorithms by integrating them with the explanatory power of Large Language Models (LLMs). Imagine having an AI assistant that not only finds the optimal solution but also explains, in plain English, how it got there. This is the promise of "Explainable ES" (XES). Researchers are using self-adaptive ES algorithms, which cleverly adjust their search strategies on the fly, to tackle challenging optimization problems. As the ES navigates the search space, a detailed log of its journey is recorded, including its steps, successes, and setbacks. This log is then fed to an LLM, which acts like a translator, converting the technical data into a clear, human-readable summary. The LLM identifies key aspects like how quickly the algorithm converges to a solution, the quality of the best solutions found, and whether it got stuck in any suboptimal traps. In a case study using the notoriously tricky Rastrigin function, known for its many local minima, this approach showed promising results. The LLM successfully summarized the ES's performance, highlighting the best and worst fitness values and pinpointing moments of convergence and stagnation. This ability to understand the ES's behavior offers valuable insights for further optimization, like fine-tuning its parameters to avoid pitfalls. While still in its early stages, this research opens exciting doors for making complex AI systems more transparent and trustworthy. Future directions include interactive analyses, where users can ask questions about the optimization process, and even agent-based systems that use the LLM's insights to automatically improve the ES's performance. This work brings us closer to a future where AI not only solves our problems but also explains its reasoning, fostering trust and enabling more effective collaboration between humans and machines.