I'll echo this. This kind of laser focus narrows options a lot after university as well. The dirty secret about most "ML" jobs is that a lot of the work looks l

I'll echo this. This kind of laser focus narrows options a lot after university as well. The dirty secret about most "ML" jobs is that a lot of the work looks like the things professors delegate to grad and undergrad students: data collection, data cleaning, automation, etc. The actual ML is a fraction of the work at best. Same goes for computer science as such. Most of the work is grunt work in testing, debugging, peer review, etc. As you advance, there may be more time available for pure math work, but that means the amount of communication just goes way up: writing, peer reviews, presentations, management, managing up, etc. Point being, don't be afraid to develop your application skills. Things like teamwork and communication matter a lot. Computer science is the common major, but most demand is for coders. Of course, what companies need is typically software engineers and (good!) software project management.