IMO this would track, applied mathematics (even e.g. data science, though perhaps calling that applied math is a bit generous / insulting to more serious applied math) is in some ways more exciting now because it is far easier to surface complex / appropriate methods for the task at hand, and you can more confidently explore these methods because the AI sort of "has your back" in catching some of the more obvious beginner errors you make during these explorations. Plus, applied math feels roughly more results- than process-focused, compared to pure math.
IMO the divide here between pure vs. applied math feels a lot like the divide between those who enjoyed coding for the understanding it led to, i.e. the writing itself was the joy, vs. those that primarily coded for the results. I enjoy the creative part of coding, the thought of software jobs just devolving into writing specifications and doing code review very much kills it for me.