[D] Are some classes of models considered "obsolete" now?

I am looking to do some self-driven training in ML and data science as a senior SWE. I am in the process of collecting and organizing courses, and I am wondering if some topics are now obsolete and I can safely skip them. This is primarily from the POV of building applications.

For example, take the Andrew Ng GAN specialization on Coursera. With stable diffusion models like Midjourney or OpenAI's vision models, do GANs still have a use case? In …

building courses data data science machinelearning process science swe them topics training