During our section on Existentialism and the meaning of life in the Introduction to Philosophy course that I just finished TA-ing this Fall, we read and discussed sections from Ecclesiastes. If you don’t know it, Ecclesiastes is a collection of Old Testament verses in which the eponymous title character discourses on the apparent meaninglessness of pleasure, accomplishment, wealth, politics, and life itself in the face of the infinitude of the universe and the absolute perfection of God. It is the source of many of our most clichéd phrases, such as “there is a time for everything” and “there is nothing new under the sun,” variations of which are repeated throughout the text as a sort of motif.
I am no Bible thumper, but I think Revelations and Ecclesiastes are the bible books for our time. The end is indeed nigh, but it ain’t nothing new under the sun. This could well apply to many features of our collective situation, but I want to apply it to the philosophy discipline’s despondency and collective freakout over the LLM problem and argue, pace Ecclesiastes, that we can and should try to do something new about it in our courses.
Rapturous memes, both of salvation and perdition, abound in the AI discourse, from last year’s creepy billboards around San Francisco declaring “Humans are so 2023,” the “AI 2027” scenario published last Spring by Kokotajlo et. al., to Marc Andreesen’s promises of private jets for chump change courtesy of super-intelligent computers. So far, these lurid prophecies haven’t panned out, and instead, AI is looking more and more like an inherently unreliable mediocrity machine that requires obscene amounts of electricity and other resources from an already ecologically degraded and resource-scarce world. In fact, the major public-facing and revenue-generating applications of the technology are turning out to be no high sci-fi futurism but pornography, mass surveillance, and cheating on one’s homework.
This last aspect of AI has generated its own subgenre of apocalyptica. For instance, Yanis Varoufakis recently claimed that in just a few years “higher education will be kaput” since, displaced by chatbots, “professors will no longer be the holders of knowledge.” More darkly, James Marriott in “The dawn of the post-literate society and the end of civilization” claims that “not only philosophy but the entire intellectual infrastructure of modern civilization depends on the kinds of complex thinking inseparable from reading and writing,” the final nail in the coffin of which, according to Marriott, AI is driving in now. This sentiment is echoed by Derek Thompson in “The End of Thinking,” which bemoans not the rise of super-intelligent machines but of super un-intelligent humans.
Indeed, according to my own informal surveys of my students, use of LLMs to summarize readings or “generate ideas” is widespread, and in the library at my University, one can observe students using ChatGPT to generate answers to their physics homework, solutions to their coding assignments, and responses to Canvas discussions. This use of AI does seem to be having negative effects. Many professors bemoan a decline in the quality of their students’ writing and thinking, and, even in elite colleges, students apparently can’t read books anymore.
But, I submit, none of this is new. I am just old enough to remember the tail-end of the golden age of the internet, when it promised to smash our television-induced torpor and bring the World together. My formative years, however, occurred in the internet’s “silver age.” What I am calling the internet’s “silver age” lasted from the mass adoption of Facebook in 2007-2009 to about 2020. In retrospect, this era was a halfway between the utopian promises of the early internet and the memeified nihilism of its current state. The old promise held on in things like MOOCS (Massive Open Online Courses), the association of social media with the Arab Spring and Occupy, and a proliferation of alternative ideas and marginalized information available on podcasts and blogs. I remember people saying then, like Varoufakis now, that widespread availability of information would spell the end for higher education’s monopoly on access to knowledge. Ironically, it has been true for some time now that there is absolutely no reason to pay for a college course to learn, say, Linear Algebra. If you know how to use a textbook and can look up any of the myriad high-quality lectures available online, you really can learn more today than at any other time in history. But, of course, that’s not what most people use the internet for most of the time.
The utopian possibility of the internet was always real, and, I believe, still is. Yet over the course of the 2010s, that utopian promise gave way to its opposite as the Internet became increasingly algorithm-ified, app-ified, extremist, and privatized. Instead of enlightenment, decentralization, and global cooperation, we got mass shooters, Donald Trump, super-surveillance, and epidemic loneliness.
We also got dumber.
Newspapers closed down or consolidated en masse, and instead of replacing them with grassroots journalism and global dialogue we replaced them with our crazy uncle’s rants on Facebook and clout-chasing twitter threads. Books became an element of décor. Socialization was replaced with scrolling and liking on Instagram. Our attention spans eroded, our literacy withered, and our echo chambers became more bizarre.
And yet, if we’re honest, we can see that this story is not just that of the decline and fall of the internet. It is longer than that. In another essay, Derek Thompson channels Neil Postman and argues that “everything is television now.” He’s right. Television was a really new form of media. Instead of capturing your attention by focusing it on a thing for a period of time, as does a lecture, book, or film, television is on in the background, all the time, smothering and diffusing (defusing, too) your surplus attention in a continuous stream of diversionary and anesthetic “content.”
Television was a powerful force of pacification and atomization. Rather than socialize through conversation in local group activities such as dances or fairs, people increasingly plopped alone in front of the tube. Television’s ability to define truth and insert official narratives subliminally into the minds of millions proved to be of immense use to states and advertisers. Indeed, the earliest adopters of mass television programming were the Nazis. Television also changed our politics and our intellectual life. It made politics turn on spectacle and narrative management, and people raised on television came to think increasingly in scenes and cuts rather than propositions and paragraphs. All of our media today is just this basic paradigm but faster, more targeted, and less avoidable. Its effects have been of correspondingly similar direction and greater magnitude. Really, there has been a conspiracy against literacy, thought, and sociality afoot since at least the mass uptake of television after World War Two. Perhaps it is not a coincidence that the last really great Philosopher, Wittgenstein, hails from just before that time.
And now here we are, at the final end of literacy. MOOCS and online textbooks didn’t replace College, but AI is displacing the learning that used to happen there. Many of our students use ChatGPT for basic tasks such as composing emails, and in our courses we have had to scramble to combat the existential threat to the skills we try to teach that are posed by AI-generated text output. The centrally controlled and thought-eroding median content of television has been perfected as students today dutifully transcribe and present as their own the transmissions of the new aleatory Ministry of Truth. This is the logical conclusion of the tele-visualization of culture, not the sudden intrusion of a new technology.
Seeing the current conundrum in this context can help us to stop freaking out about AI specifically and to grasp the magnitude of the long-term cultural transformation that it is a part of. We should be more horrified about the big picture, but calmer in the immediate context. Thereby, we may think a little more clearly about how to respond in our limited capacity as philosophers and as educators. If the vocation of the gadfly is to quietly nurture a lingering question mark after all things and to cultivate an ineradicable taproot of logic against the tillage of official narrative and common sense, then we need to proceed with cunning and daring against the foe that is the final tele-visualization of culture. Our classrooms are the field of our quest, and the problem of students turning in AI slop essays is our most immediate task.
Currently, most university teachers I know are attempting to force acquisition of the same skills in more or less the same way by finding work-arounds that are less easy to cheat. The most obvious and widely adopted solution has been to replace the traditional essay with in-class blue book exams. However, this comes with serious downsides. It deprives students of the opportunity to develop their own ideas over a sustained period of time, replaces the deep and considered thought of a take-home essay with frantically scribbled summaries, is absolutely no fun for the students or for the teachers, and involves a great deal of inscrutable handwriting. This amounts to a major diminution of the significance, value, and enjoyment of a philosophy course. If I had only ever written in class blue book exams in my undergraduate courses, I doubt I would still be here.
This is also, I think, not a very cunning approach. Complex goals are rarely achieved directly via a simple opposition of force against force, especially when you are the weaker force. What is needed instead is creative thinking about how to do philosophy in different formats besides the essay and to think carefully about the core skills we teach in philosophy and their relevance to the world that our students actually inhabit.
To the latter question, I think there is actually some good news. I think it is relatively easier to explain the value of philosophy now than it has been for some time. If in the past it was hard to justify “just thinking about stuff” since what mattered was “actually knowing things”—especially things such as computer programming—then, today, that formula has largely inverted. AI does in fact have impressive capacities to make the lookup and summary of information much quicker and to expedite tasks such as building a website. However, due to its inherent unreliability and inability to generalize outside the mean of its training data’s distribution, AI must be utilized with a great deal of thought and caution if it is to be valuable. For example, I tried asking ChatGPT a series of questions about some of the finer points of Marx, Hegel, and Kant. It will generally give a rather good overview of what these thinkers said and how their views differ from the views of other philosophers. However, if you push it to answer more nuanced or detailed questions (especially about topics where there is little literature or no published consensus), it is unable to situate the significance of the question in a broader context, often contradicts itself between answers, and will tell you misleading or very controversial things as fact. Hence, there is a great premium on having a rich and systematic understanding of a content area and on being able to “just think” critically about stuff in general.
I am sure that this applies not only to topics in 19th century German Philosophy but also to subjects closer to our students’ interests, such as computer programming and financial forecasting. Beyond the public-facing applications we are familiar with, AI algorithms trained on specialized data will likely see a good deal of application in non-public, bespoke tasks. For example, a friend of mine tells me that her lab is working on getting an AI model to narrow down possible molecule folding structures for further analysis by the more accurate (but more expensive) computer they currently use. The skills we teach in philosophy will be very valuable to students entering into industries where these kinds of models are being trialled or utilized, and we should not obscure this by digging in our heels and subjecting our students to nothing but blue books and sermons about AI usage. We could instead, for example, develop projects where students put together critical analyses of AI output or study its potential application in a given domain. Perhaps we could even partner with Computer Science Departments to have students train their own open-source models and then present philosophically informed critical evaluations of their model’s output, potential, and limitations.
But what, then, is to be done about the problem of the essay? The philosophical essay has hitherto been the crucible in which all of the above-mentioned virtues are forged. And, if you believe James Marriott, the written treatise is quite literally the foundation of modern civilization.
An obvious yet radical alternative is the oral exam and other forms of in-person, verbal evaluation. This is an obvious proposal since the oral exam has been around for a long time in German universities and lives on in the dissertation defense. There is a kind of prestige and ceremony to the oral defense: a serious work deserves public scrutiny, and if you really know your stuff, you should be able to defend it live and in person. Nevertheless, it is a radical proposal precisely because of the long equation of serious thought with written reflection and of veracity with the permanence of text.
Moreover, the essay has been a cornerstone of modern education since it had direct transfer to public intellectual life and to professional life. Even if our students never wrote or read about Descartes after graduating, they would have to be able to read, and perhaps write, elevated prose about complex topics if they were to stay informed and to participate in culture and politics. Moreover, it had direct transferability to professional skills such as writing a brief or a cover letter.
Today, neither of those apply. Our increasingly post-literate culture has largely dispensed with the essay and its relatives as the medium of popular intellectual and cultural life. Instead, articles and essays are increasingly a means of specialist communication in academic disciplines. In place of written forms, the culture has become dominated by oral and visual forms that range from the TikTok reel to live-streamed debates, video essays, and podcasts. It is simply not true that fluency in the essay is a requisite for participation in public debate and culture as it exists now. One may well lament this fact, and one may argue that public debate has been degraded as a result, but the point nevertheless stands. What is more, the status of text as the prestige medium of record and scrutability may be on its way out. When we are awash in AI text of dubious veracity and origin, text loses its connection to law, commitment, and verifiability. Not only text, but the recorded image is being dethroned as prestige media of record. Neither is trustworthy or verifiable any longer. Perhaps, then, we will see a return of oral presence and rhetorical skill as the imprimatur of veracity.
In the workplace too, writing is not as important as it used to be. The writing of routine briefs or note taking will be and right now already probably is being done almost entirely by AI. Writing a brilliant cover letter is also now an almost completely meaningless skill. No one reads cover letters anymore, and no one writes them. AI bots do both. On the other hand, the ability to present oneself eloquently, cogently, and persuasively in person just might impress someone and get you in front of the crowd in the way a great cover letter may have in the past.
I know that many philosophers will recoil at the suggestion to move away from the essay. It is true that writing is more than a mere container for thought that may be swapped out for another. Writing is an act of thinking, and the complete loss of the ability to write would be an unconscionable sacrifice. However, if what I’ve said above is true, then it doesn’t much matter if the written treatise is the foundation of philosophy and of “Civilization” because the essay, as a mass form, is going away regardless. The post-textual society is upon us. Perhaps this will mean that real literacy and textual thought will have to become a rarefied enterprise, the purview of a small, cloistered elite like the monastic transcribers of Aristotle in medieval Europe. Be that as it may, we should maintain our cloisters and tend them jealously, but if we want thinking, logic, and reasoned persuasion to live on in the broader culture, we need to figure out how to transfer those things to an oral format. This is a civic duty, but it is also essential for the long-term sustainability of our cloisters that we educate and produce in a way that can bring people into our cloisters that are not already in them. We should think of ourselves as a guerilla force waging resistance against a much greater adversary, and clinging to the territory of the essay may get us kettled and killed. We may need to strategically cede that territory so as to entrench our resistance within the enemy’s own territory.
I believe that we can do this. I believe we can conduct and teach rigorous thinking in formats besides the essay. I also think that a move towards oral forms of examination can have benefits that the essay doesn’t. For example, it can instill vital social skills and experiences in our students better than the essay. In an era of social isolation and intense atomization, this is a significant benefit. It may also lead to philosophical expressions that are more moving and beautiful than the current journal article norm. Furthermore, it may bring out the best in our students by enabling them to express themselves in a form that they have greater native fluency in. I have found that many of my students are better able to express themselves in conversation than they are in writing; all those TikTok reels seem to have made them pretty witty and fast with comebacks. For educators, it can also be an exciting opportunity for experimentation and classroom innovation. In the class for which I’m TA-ing this winter, we will be trying a format where students independently develop an argument, as they would in an essay, present it without notes as a brief 5-minute speech, and are then questioned about what they said for ten minutes. Other experiments may be possible, for example, debates between students or graded close readings. There is no need to be limited to a dry or predictable format.
Finally, I am skeptical that “philosophy and Civilization” are really so tied to text. The mass literacy we know now is a relatively recent phenomenon. Orality, in fact, has been the norm for most of human existence, and there have been many great civilizations that had little literacy. Orality does not have to be limited to the TikTok reel or YouTube short. Orality can be beautiful and profound, like Nina Simone or Cicero. As for philosophy, a case could be made that the soul of philosophy is not in the text but in the conversation. Western philosophy’s founding documents are dialogues, not treatises, and even today philosophy as a discipline maintains a much higher regard and space for oral debate and conversation than perhaps any other discipline. One might even argue that it is the text that is philosophically suspicious since it forces thought into the rectangular confines and graven assertion of certitude that is the written word.
Perhaps, then, we can look at this not as a calamity but as a new return to our roots. Maybe this new oral era can combine with our textual inheritance to produce a second philosophical renaissance; it doesn’t hurt to dream a little. Either way, I’m excited to grade my students’ oral exams rather than their blue books this Winter.
Gabriel Schat
Gabriel Schat is a PhD student at Northwestern University. He studies Marx , Hegel, and Process Ontology. He is writing a dissertation about the relationship between the universal and particular in Hegel and Marx.
