The field of artificial intelligence is witnessing a significant shift towards the development of multi-agent systems that can facilitate complex decision-making and creative tasks. Researchers are increasingly exploring the potential of large language models (LLMs) and multi-agent interactions to improve various applications, including storywriting, requirements engineering, and harmful content detection. A key trend in this area is the use of multi-agent debate strategies to enhance the accuracy and robustness of LLMs. By enabling different agents to collaborate and debate, these strategies can help reduce bias and improve the overall performance of the system. Another area of focus is the development of frameworks that can support zero-shot learning, allowing systems to detect and respond to new and emerging threats without requiring extensive annotated data. Overall, the development of multi-agent systems is enabling more effective and efficient solutions to complex problems, and is likely to have a significant impact on a wide range of applications in the coming years. Notable papers in this area include: Constella, which supports storywriters' interconnected character creation through LLM-based multi-agents, and MIND, a multi-agent framework for zero-shot harmful meme detection that outperforms existing zero-shot approaches.