AI is useless in many Humanities courses. At least 30% of the final grade is class participation, in my case with the added wrinkle of having to be able to participate in it in Spanish. What is AI going to do for you in this situation?

Soon enough, Humanities degrees are going to be the most valuable ones because there will be a certainty that AI didn’t do the entirety of the work for the student.

So far, very few students in my courses even tried to engage in AI cheating. Not because they are unusually honest, but because they discover soon enough that it doesn’t work. Humanities are not about finding one correct answer. They are about the beauty, the enjoyment, and the importance of the process of looking for answers. You can’t short-circuit the process because the process is the whole point. Being in the presence of beauty does not have a purpose. It is the purpose.

5 responses to “Good News for the Humanities”

  1.  Avatar
    Anonymous

    Anything with a real-world component ought to survive.

    “Ought to” is doing yeoman’s work in the preceding sentence: Medicine, engineering, computer programming degrees et al. ought to have been retained value.

    Yet just a few years ago actual nurses and doctors insisted on point of retaliation that scraps of fabric and paper, worn by people untrained in PPE use stopped community transmission of a corona virus. Some still do. DIE-engineer bridges collapse. And anyone using MS software knows how well it used to work in the Before Times.

    Americans who value knowledge and learning have watched in sadness and disgust as the academic-government-industrial complex, pace Eisenhower, inexorably termite-ed our universities and colleges.

    AI may speed the burnt fool’s wobbling finger back to the fire of reification. I.e.: “If Credentialed Academic B says the person participated in class, in comprehensible Spanish, and Credentialed Administrator D stamps the paperwork conforming it, Student D has Languages Degree.”

    Or it may not. The problem, O Academics, lies not in our LLM stars but in yourselves.”

    Interesting times.

    Like

  2. Physpostdoc Avatar
    Physpostdoc

    “Soon enough, Humanities degrees are going to be the most valuable ones because there will be a certainty that AI didn’t do the entirety of the work for the student.”

    This is completely backwards from what we are seeing around in several universities including mine (top private R1), not to say that either your hope or actual version of events is a happy development.

    The humanities courses are the worse hit by any measure with rapidly improving abilities of LLMs to generate, analyze and critique text, complete with “humanizer” options. Since most of the humanities coursework heavily relies on subjective analysis and writing, all of this has been rendered redundant in one fell swoop with the advent of AI combined with unfortunate propensity of students to cheat. A direct indication of this is the number of humanities departments and admissions that have been severely downsized/frozen across the board in the past year (see, for example, Syracuse University’s recent closure of admissions in 84 majors— it is the most recent and extensive, and not the only one https://www.nytimes.com/2026/04/01/nyregion/syracuse-university-degrees-eliminated.html ).

    You have mentioned before too about language courses being protected against AI due to strong verbal component in instruction — which, even if true (in the face of multimodal AI development) is hardly a consolation for higher education in these disciplines, since most of the courses where strong verbal in-class engagement is involved are traditionally service or lower introductory courses (like in sciences where teaching faculty can try pop quizzes and oral tests etc in Physics 101, Calculus 101 etc). As students progress through junior year into specialized and senior UG courses, and then into Master’s and PhD classes, the primary mode of study as well as evaluation almost exclusively shift to homework or self-study based work. And here the situation has already turned into a tech-dystopia novel for humanities disciplines!

    In STEM courses the analytical and objective problem solving is still somewhat impeding this free march of AI. Even though the complete rout that we are seeing in humanities is unlikely in mathematics and hard sciences, the changed approach due to outsourcing of several research skills, such as writing papers and coding, to AI is already leading to a lot of uncertainties and re-evaluation of research methodologies. It is a brave new world, one for which scarcely any of us is ready.

    Like

    1. Clarissa Avatar

      I should probably take this post back, given the recent discussion about a Harvard literature professor who claims not to be able to understand a novel by Toni Morrison. With professors like that, I don’t know what we expect.

      As for AI, it improved significantly in its research capacities. It hallucinates less often, and it provides mostly good, helpful bibliographies. I use it almost daily for these purposes.

      But it’s utterly impotent at writing texts that anybody would want to read. It definitely doesn’t generate any ideas about works of literature. I sometimes place a paragraph I have written into AI and ask it to complete the paragraph. Then I weep with laughter because it’s genuinely entertaining how impotent it is to do anything other than take it all back in a circle to the beginning of the sentence.

      In what concerns the use of AI in the workplace, let’s take an example of a business that sells first and foremost text. Such a business used to hire 15 writers and 2 editors. The 15 writers are now gone because of AI, but instead of two editors, there are four because their work has become more important than ever. Nobody is going to put out an unredacted AI text because it’s simply unreadable.

      AI doesn’t create or think. It generates masses of grammatically correct, plausibly sounding text. The dirty secret of undergraduate education is that professors very rarely read student essays. They kind of skim, but they don’t read. They will have to revise that approach and start reading. As soon as they do so, AI writing will become obvious.

      Like

    2. Random reader Avatar
      Random reader

      It is possible to do in class assessment for advanced STEM courses as well. I completely switched to having pop- quizzes every class in my graduate level course. I still assign homework, but I don’t bother to grade it as it is pointless in the age of AI. However, I do have a short quiz on the beginning of each class that tests students on concepts and my exams will all be in person. I think it is up to us to understand what AI can do, how it is being used and change the way we assess students accordingly. The times of relying on homework or take home exams are gone. Students still have to do the homework in order to learn, but I am not grading it.

      Like

      1. Clarissa Avatar

        Same here. I’ve been doing quizzes at the beginning of every class to test if the students have done the readings for years. Everybody fails the first couple of times and then they start doing the readings.

        Like

Leave a comment

A conservative literary critic and college professor