What is wrong with it?

Unless you get a faculty position at a stellar university (highly unlikely nowadays), the teaching will be depressing. And your research will suffer as a result since you will be in no mood to do it.

Also, unless you plan to do everything yourself for research, you will need to get some funding. But whether you get that funding depends on whether your peers -- competitors actually -- like what you plan to do.

Finally, the fact that computer science tries to be a science severely limits your creativity -- at least if you want to publish in respectable conferences and journals.

Actually, the teaching is depressing even at stellar universities. I'm a postdoc at such a place. Last semester, I taught "Calculus for kids who don't care." This semester, I'm teaching "Poker and craps for kids who don't care."

I suspect the only way to avoid this is to work at an engineering school (not necessarily a top one).