The field of engineering design and optimization is witnessing a significant shift towards the adoption of differentiable programming techniques. This movement is driven by the need for more efficient and scalable optimization methods, particularly in high-dimensional design spaces. Recent developments have focused on replacing non-differentiable components in traditional computer-aided engineering workflows with surrogate models, enabling end-to-end differentiable pipelines for tasks such as shape optimization. Noteworthy papers in this area include the Surrogate-Based Differentiable Pipeline for Shape Optimization, which demonstrates the potential of using 3D U-Net full-field surrogates to replace meshing and simulation steps. Another significant contribution is the Differentiation Strategies for Acoustic Inverse Problems paper, which showcases the use of automatic differentiation and randomized finite differences for acoustic shape optimization, achieving substantial energy reduction at target frequencies. The Fast 3D Surrogate Modeling for Data Center Thermal Management paper presents a vision-based surrogate modeling framework for real-time temperature prediction in data centers, achieving significant speedups and energy savings. These advancements have the potential to transform the field of engineering design and optimization, enabling faster, more efficient, and more accurate design processes.