Sound is a rich source of information that can enhance an artificial intelligence's situational awareness, providing insights into its environment and identifying the presence and behavior of specific biological entities, such as humans and animals. Current artificial intelligence approaches excel at identifying sounds within a scene and large language models are capable of articulating common knowledge. However, a gap exists between how humans assess a sound environment and evaluate specific bio-sounds, and how current AI systems do so. This project aims to close this gap by enhancing the biological plausibility of machine listening AI, leveraging our deep expertise in human hearing, sensory gating, attention, and short-term memory to develop new representations and in silico models. Oscillatory neural circuits will play a crucial role in this development. The final result will enhance human machine interaction, soundscape augmentation, and make cities smarter.