Imagine a world where AI can handle the complexities of business processes, from predicting the next step in a workflow to identifying automation opportunities. Recent research explored this very possibility, examining how Large Language Models (LLMs) perform on core Business Process Management (BPM) tasks. The study pitted several open-source and commercial LLMs, including the heavyweight GPT-4, against each other in a series of challenges. These included recommending the next activity in a process, identifying tasks ripe for robotic process automation (RPA), answering questions about process descriptions, and extracting process constraints from natural language. Surprisingly, the results revealed that bigger isn't always better. While GPT-4 demonstrated a consistently strong performance, smaller, open-source models like Llama 3 proved remarkably competitive, even outperforming GPT-4 on certain tasks. This suggests that businesses might not need to invest in the most powerful (and expensive) LLMs to gain AI-driven process improvements. However, the research also highlighted the importance of careful model selection. Different LLMs exhibited distinct strengths and weaknesses across the various BPM tasks. For example, Claude 2 excelled at activity recommendation but lagged in other areas. This underscores the need for businesses to carefully consider their specific requirements and choose the LLM that best aligns with their needs. While this research provides encouraging evidence for the potential of LLMs in BPM, it also acknowledges limitations. The study focused on a specific set of tasks and models, and further research is needed to explore a wider range of applications. However, the findings clearly demonstrate that LLMs are poised to play a significant role in transforming how businesses manage their processes, offering a glimpse into a future where AI-driven insights optimize efficiency and decision-making.