The field of molecular optimization and design is moving towards leveraging reinforcement learning and generative models to efficiently explore chemical space and discover novel molecules. Recent developments have focused on improving sample efficiency and optimizing molecular properties while preserving structural similarity to the original lead compound. Notable advancements include the integration of reinforcement learning with molecular fingerprints to generate chemically relevant molecular structures and the use of active learning to prioritize top design candidates in generative AI workflows. These innovations have significantly improved the generation of valid and realistic molecules, reducing the frequency of structural artifacts and enhancing optimization performance. Some particularly noteworthy papers include: POLO, which achieves superior sample efficiency by fully exploiting each costly oracle call, and Steering an Active Learning Workflow Towards Novel Materials Discovery via Queue Prioritization, which significantly increases the number of high-quality candidates identified by the generative model. Guiding Evolutionary Molecular Design: Adding Reinforcement Learning for Mutation Selection is also noteworthy for its integration of reinforcement learning with evolutionary algorithms to guide molecular mutations based on local structural context.