Software is not only taking a shot at writing essays but also grading them and providing instant feedback on student work in progress, analysis that is well beyond grading multiple-choice quizzes. These programs still need to work out some bugs (a clever student can game them with coherent-sounding nonsense), but they are much further along than we had been expecting five or ten years ago.

(Emphasis added.)

That’s from Tyler Cowen’s Average is Over.

A surprisingly large number of papers and books in the humanities, as well as grant proposals, are filled with “coherent-sounding nonsense,” and at least in humanities papers I’ve read a lot of incoherent-sounding nonsense (which may help explain declining enrollment in humanities majors). The market for coherent-sounding nonsense is surprisingly robust.

EDIT: Relatedly, much later Cowen writes of the way that in economics “Newly minted PhD candidates are extremely proficient with data, but a lot of them don’t have much microeconomic intuition. [. . . ] Overall, the profession is producing more first-rate empiricists than before, yet theory hasn’t progressed much in twenty years or more. Theory is increasingly ignored” (225). If one could make a similar statement about English the field was a whole would be better. In some ways, perhaps one can: the growth of MFA programs and undergrad writing classes is some in sense the move from a theory of literature to the practice of it.