Generative Artificial Intelligence (GenAI) has gained popularity in software development, including in software architecture (e.g., for generating architecture documentation or exploring architectural decisions). We argue that GenAI not only impacts how architects work and the artifacts they produce, but also affects (i) skills required to effectively and efficiently use GenAI for software architecture and design activities, and (ii) skill development and assessment (e.g., how to move from junior to senior engineer without actual hands-on experience due to the use of GenAI or the risk of "de-skilling'' of senior architects by relying on too much on GenAI without critical reflection). We therefore propose a new research direction to understand and support software architecture and design skills in the era of GenAI. We also outline several questions and problems for future research.