One more small piece of advice: try to get on whatever mailing lists you need to to start attending the research talks at your college or university. (Some of this might be harder in the age of Zoom/economic collapse, but still...) This can vary based on what type of college you are attending, but if you are at a research university, there can be a lot of these: departments or research groups may have seminar series/colloquia, and then every doctoral student is going to have a public defense, and every job candidate who wants to be a professor is going to do a public "job talk," and at some places, every junior-faculty member who wants to get tenure has to do a "tenure talk," etc. These are almost always just open to anybody who wants to come.
So there may be a lot of opportunities for you to see a diverse range of talks (many of which will be intended for a general CS audience, at least for most of the talk) and get a sense of what people find exciting/interesting/achievable on various frontiers of CS research in different subfields. For the job talks at least, the candidate is sometimes giving the best talk they will ever give in their life, so it's possible to see something compelling in a subfield that you might not otherwise glimpse just by taking a course or something. And (this is more applicable to real-life talks) if you like you can usually go up to the speaker afterwards to ask them something privately and usually presenters are flattered to have somebody interested. Or if somebody in the audience asks a question that you find interesting, same thing. This might seem awkward and I know it might not seem this way, but it's hard to truly embarrass yourself by showing interest or enthusiasm -- you're in college! (As a student, I once found myself in a conversation with some random dude about the Gamow/Stern/Knuth elevator results after we both waited for the elevator for a long time after attending a talk at MIT, and when we were done I was like, "By the way, who are you?" and he looked a little surprised and was like, "Uh, I'm Butler Lampson.")
If you're not at a research university, that may decrease the number of random research talks you can show up to, but it also suggests the faculty will be much more eager to collaborate with undergraduates and involve you in a material way in their research projects.
So there may be a lot of opportunities for you to see a diverse range of talks (many of which will be intended for a general CS audience, at least for most of the talk) and get a sense of what people find exciting/interesting/achievable on various frontiers of CS research in different subfields. For the job talks at least, the candidate is sometimes giving the best talk they will ever give in their life, so it's possible to see something compelling in a subfield that you might not otherwise glimpse just by taking a course or something. And (this is more applicable to real-life talks) if you like you can usually go up to the speaker afterwards to ask them something privately and usually presenters are flattered to have somebody interested. Or if somebody in the audience asks a question that you find interesting, same thing. This might seem awkward and I know it might not seem this way, but it's hard to truly embarrass yourself by showing interest or enthusiasm -- you're in college! (As a student, I once found myself in a conversation with some random dude about the Gamow/Stern/Knuth elevator results after we both waited for the elevator for a long time after attending a talk at MIT, and when we were done I was like, "By the way, who are you?" and he looked a little surprised and was like, "Uh, I'm Butler Lampson.")
If you're not at a research university, that may decrease the number of random research talks you can show up to, but it also suggests the faculty will be much more eager to collaborate with undergraduates and involve you in a material way in their research projects.