Just to add my comments after reading this thread while taking a break from deciding on whether I should use circular dependency for the design of my next program. (yes, I am in a computer science related field)

I have a feel that the OP does know what research at a graduate level is like. For all the others who comment on how it would be a time of drudgery, non-productive work and a less linear pursuit of knowledge, I feel they themselves are only passing what others told them post-graduate work is like.

While those who commented are probably only educated to a bachelors level, there is some truth to their words. What I see is that the OP probably already knows this, further evident from how he is already contemplating about this prior to undergrad, and is simply questioning the salary of a job as a professor while he endures the journey in unraveling the wonders of computer science.

My take: Do a PhD only if you are interested in research. While you right now do seem to have PhD as a goal, reassess this after say two years in undergraduate. The difference in undergradate and PhD is glazed by few of the viewers. I'll add some detail with which I am confident will paint a more vivid picture of what PhD students do.

Simply put, you need to conceptualize, develop and produce something original. What I mean by that is something which is not covered in the textbooks, let alone the publications were written by other professors.

Many students fall into the trap where they think they know for sure that they want to do a PhD because they can get A's and A+'s in their first and second year. Great grades could be an indicator of whether you want to do a PhD but it is more of an indicator of whether you can do a PhD. You wanting to do a PhD may be more accurately assessed by the following:

1. Do you attempt to ask more questions once you found a solution to a hard problem. I.e., consider another data structure of solving an algorithmic problem but reduces the run time.

2. Do you subconsciously think of computer science problems amidst your usual routine in life. This sounds a little lame but I'm quite sure that every CMU, MIT and Stanford bound person wouldn't mind having a discussion of Java vs C++ whenever wherever.

3. Do you view programming as fun or as a chore. Put differently, would you skip LAN sessions to go home and run Visual Studio 2010.

These are some questions that when answered truthfully will give you a better idea on your pursuit of a PhD.