The field of medical natural language processing is experiencing significant growth, with a focus on improving the accuracy and reliability of medical question answering and dialogue generation systems. Researchers are exploring innovative approaches to grounding large language models in high-quality medical knowledge, enabling more targeted retrieval and improving the detection of medical hallucinations. Notable advancements include the development of curated corpora of medical query-response pairs, which have been shown to improve the accuracy of medical question answering benchmarks. Additionally, multi-agent generated datasets are being utilized to advance medical reasoning capabilities, with promising results in fine-tuning strategies for large language models. Particularly noteworthy papers include: MIRIAD, which introduces a large-scale curated corpus of medical QA pairs that improves the accuracy of medical question answering by up to 6.7%. ReasonMed, which presents a multi-agent generated dataset for advancing medical reasoning, achieving state-of-the-art results on medical question answering benchmarks.