>>10026121Depends. For most people in it for the money, it's mostly a tech degree for people who otherwise don't care about math and science. It's safe, it doesn't have the same requirements are engineering, and it's got good job prospects.
That being said, as a field, and for those who want to get into it and DO like math and science, it's a major that, if you pick your school carefully and double major in it with math or physics, can be an incredibly rewarding major and very interesting subject. In immediate undergrad, going deep in either systems (think OS design, scheduling problems, memory modelling, designing procedures that control the circuitry directly, etc.) and theory (advanced algorithms, theory of computation, quantum information theory and computation, numerical analysis, etc.) is a rabbit hole of interesting topics. You get to use CS theory (read: not just computational models to plug equations into) to study all sorts of things.
Grad CS tends to elucidate those. I'd say avoid mainstream ML because there's a lot of bad research that masks the good research since it's so heavily funded. ML theory and AI theory on the other hand are pretty great and interesting fields. Grad CS is also where you realize that it's just all math. Even your algorithms require mathematical analysis a good amount of the time, especially with randomized algorithms. You get situations where the continuous case gives you the intuition you need for your problem. Look up Alon-Cheeger Inequality in spectral graph theory. which came from an understanding of Reimannian manifolds no less.