The field of human-AI collaboration is rapidly evolving, with a growing focus on designing interactive systems that support creative tasks and research writing. Recent developments suggest that reframing research paper writing as a serious game and using role-playing mechanics can enhance motivation, creativity, and intellectual growth in novice researchers. Additionally, there is a growing concern about the homogenization of human thought through repeated exposure to similar AI-generated content, highlighting the need for scalable methods to evaluate language model output diversity. Studies have shown that human-led interaction modes can improve idea quality while preserving diversity and ownership, whereas model-led modes can improve idea quality but reduce diversity and ownership. Furthermore, the use of divergent and convergent thinking personas in LLMs can shape human machine creative problem-solving and provide design principles for creativity support systems. Noteworthy papers include: GAMER PAT, which introduces a prompt-engineered AI chatbot that reframes research paper writing as a serious game. Artificial Hivemind, which presents a large-scale study of mode collapse in LMs and reveals a pronounced Artificial Hivemind effect in open-ended generation. Partnering with Generative AI, which compares human-led and model-led interaction modes in human-AI co-creation and finds that human-led modes improve idea quality while preserving diversity and ownership. Scaffolding Creativity, which develops a coach-like LLM environment that embodies divergent and convergent thinking personas and provides design principles for creativity support systems.