I am a scientist who just applied machine learning to my research (JEPA/Representation/Geometric branch) and it did wonder! Allowed me to see so many papers that I am still struggling to write up. From what I see, there are clearly a million possibilities not done yet, e.g., industrial data, pattern
