Imagine having an AI assistant that could sift through mountains of scientific papers and produce a perfectly structured literature review. Researchers from The Open University of China are working to make this a reality with their innovative prompting technique. Literature reviews, a critical part of any scientific endeavor, are time-consuming to create. This new research aims to automate the process, making research more efficient. Their method guides Large Language Models (LLMs) step-by-step to generate a literature review's title, abstract, headings, and even the main content. This structured approach, according to the researchers, helps maintain coherence and relevance. The team achieved an impressive third place in the NLPCC 2024 Scientific Literature Survey Generation competition, demonstrating the potential of this approach. Their system, using the Qwen-long LLM, was highly effective at producing logical structures and demonstrated high recall for matching relevant headings. Importantly, the method also proved to be cost-effective, making it a practical solution for researchers. While the results are promising, the team acknowledges that challenges remain. Currently, the system doesn't incorporate the actual *content* of the referenced papers, which could lead to inaccuracies. Future research will focus on incorporating this crucial information to create even more reliable and comprehensive AI-generated literature reviews.