The recent developments in the research area indicate a significant shift towards leveraging advanced machine learning techniques, particularly large language models (LLMs) and geometry-aware neural networks, to address complex engineering problems. There is a notable emphasis on integrating these technologies into traditional engineering tasks such as parametric shape optimization and aerodynamics modeling. The use of LLMs for shape optimization demonstrates a novel approach to intelligent decision-making through natural language prompts, showcasing faster convergence and agreement with benchmark solutions. Meanwhile, geometry-aware message passing neural networks are being employed to enhance the modeling of aerodynamics over airfoils, effectively incorporating geometric structures into the modeling process. These advancements not only improve the efficiency and accuracy of existing methods but also open new avenues for future research in integrating machine learning with engineering design. Notably, the GeoMPNN framework, which incorporates a hybrid Polar-Cartesian coordinate system and message passing schemes, has been particularly successful, winning awards at major conferences. These trends suggest a future where machine learning and traditional engineering practices are increasingly intertwined, leading to more innovative and efficient solutions.