In general, a professor teaches basic, intermediate and advanced courses.
The content of basic and intermediate courses is generally based on textbooks. Advanced courses may require recent results. But a course might still have a fixed reference material to teach content from.
Apart from this, the challenging task is to guide PhD researchers. Research progress, in terms of quantity, in some domains is like never before. Programming is also like never before. In order to understand and appreciate the work by researchers, a professor needs to update her familiarity on her student's research topics on a regular basis.
I am presuming that in some domains like machine learning, the workload of a professor is increasing at a greater rate, mainly in guiding her researchers. Is my presumption true?
Note (update): My intention is not to say that teaching is an easy task to do. It may be difficult in the initial days. Once if a person gets enough on a particular course then it is easy to manage the same course for coming years.