I think this is just a matter of what position you want. A classmate of mine wrote one paper on machine learning (he basically barely know how to run the Python to train some ancient neural net), did a short 1 year postdoc with his advisor and got a job as a professor at some no-name university in Japan.
I'm sure your classmate could get a professorship somewhere if he wanted. If his goal is to be a professor at Caltech then it's a different story