Role of AI in teaching and studying philosophy. Challenges and opportunities.

Imagine walking into a chemistry class where the teacher bans microscopes, or a math class that outlaws calculators. For philosophy teachers, the sudden arrival of AI felt less like a new tool and more like an existential threat to the very medium of thought: the written essay. If a machine can draft a passing paper on utilitarianism in five seconds, how do educators know if they are grading a student's developing mind or just a clever algorithm? This panic has forced philosophy teachers to move past simple bans and completely redesign how they teach and assess wisdom. ## The Practical Concerns: Plagiarism and the Death of the Draft The immediate crisis for philosophy educators is the death of the traditional take-home essay. When students outsource their writing, teachers lose their ability to diagnose confusion. Furthermore, AI threatens what educators call the **epistemic humility** of writing. In philosophy, we do not write to record what we already think; we write to discover *what* we think. As the philosopher and cognitive scientist [Daniel Dennett](https://en.wikipedia.org/wiki/Daniel_Dennett) noted in his work on how minds evolve, language is a tool for making our thoughts visible so we can criticize and improve them. When students skip the draft phase, they lose the ability to self-correct. ## The Solutions: Bringing Philosophy Back to Life Rather than acting as AI detectives, creative philosophy teachers are redesigning their classrooms to make cheating impossible—or irrelevant. ### 1. The Oral Exam and "In-Class" Philosophizing Many teachers are returning to the roots of philosophy by replacing take-home essays with oral examinations or in-class, handwritten bluebook essays. By stripping away technology, students must perform their thinking in real-time, defending their arguments in live, face-to-face dialogue. ### 2. Reverse-Engineering the AI "Hallucination" Instead of banning AI, some teachers assign it. In these exercises, students prompt an AI to write an essay on a complex topic, like [John Locke's theory of personal identity](https://plato.stanford.edu/entries/locke-personal-identity/). The student's actual assignment is not to write a new paper, but to ruthlessly edit, fact-check, and grade the AI's output. This approach turns the student into a supervisor. By identifying the AI's subtle misunderstandings and logical leaps, students develop deeper critical thinking skills than they would by writing a standard essay from scratch. ### 3. The "Process over Product" Portfolio Another strategy is grading the *evolution* of an idea rather than the final polished essay. Teachers require students to submit a portfolio that includes: - A hand-drawn mind map of their argument. - A audio recording of them explaining their thesis in plain English to a friend. - A log of how their ideas changed over three distinct drafts. By shifting the focus to the journey of thinking, the final product becomes less important than the visible footprints of the student's mind. > "The philosopher's job is to investigate the concepts that others take for granted." > — Mary Midgley, [*Philosophical Plumbing*](https://en.wikipedia.org/wiki/Mary_Midgley) (1992) By restructuring their courses, philosophy teachers are showing that while AI can mimic the *plumbing* of a finished essay, only a human mind can experience the genuine doubt and curiosity that drives philosophical inquiry.

If AI can now grade itself, has the take-home assignment breathed its last breath? It is tempting to think our only options left are a retreat to the stone age of bluebooks and pencils, or total surrender to automated grading. But there is a third way. We can design "AI-proof" homework by shifting our focus from **retrospective reporting** (writing about what happened) to **live, localized experiencing**. To save homework, we must design tasks where the student's unique, immediate physical and social world is the primary source material. ## The Friction of the Real: "Locally Grounded" Prompts Artificial Intelligence is trained on a static, global dataset. It knows everything about the world in general, but absolutely nothing about your specific life in particular. AI-proof homework exploits this blind spot by requiring students to ground universal philosophical questions in local, real-time data. Instead of asking students to write a generic essay on [Aristotle's virtue ethics](https://plato.stanford.edu/entries/ethics-virtue/), a teacher might assign this prompt: > Walk down your local main street or school hallway. Identify three physical objects or architectural choices that encourage community connection, and three that discourage it. Apply Aristotle’s concept of the *polis* to explain how your local environment shapes human character. You must include photo evidence and a reflection on your personal route. An AI can define Aristotle's *polis*, but it cannot walk down your specific street, observe the broken bench near the library, or connect that specific physical reality to your personal daily routine. By forcing the AI to work with highly localized, non-textual data, the student must act as the primary translator. ## The Socrates Bot: Co-Authoring with the Enemy We do not need to ban AI from homework; instead, we can turn the AI into an active sparring partner. This approach is inspired by the classic [Socratic Method](https://en.wikipedia.org/wiki/Socratic_method), where wisdom is co-created through relentless questioning. In this model, the homework is not a static essay, but a transcript of a debate. The student is instructed to adopt a philosophical position—for example, defending [Thomas Hobbes' view of human nature](https://plato.stanford.edu/entries/hobbes-moral/)—and command the AI: > "I want you to act as Socrates. Challenge my Hobbesian view of human nature. Ask me one tough question at a time. If my logic slips, call me out. We will debate for ten rounds." The student's homework submission is the unedited chat log, accompanied by a brief self-evaluation of where they struggled. Here, the AI is not a shortcut; it is a high-speed pitching machine. The student cannot copy-paste their way through this because the AI is reacting dynamically to *their* unique inputs in real-time. ## Philosophy as a Civic Lab In his classic work on educational philosophy, [*Democracy and Education*](https://en.wikipedia.org/wiki/Democracy_and_Education), the pragmatist philosopher [John Dewey](https://en.wikipedia.org/wiki/John_Dewey) argued that genuine learning only happens when theory is tested by action. > "An ounce of experience is better than a ton of theory simply because it is only in experience that any theory has vital and verifiable significance." > — John Dewey, *Democracy and Education* (1916) Following Dewey's lead, AI-proof homework can take the form of "ethical experiments." Students are asked to perform a micro-action in their daily lives and report on the psychological friction it creates. For example: - **The Kantian Truth Challenge:** Try to go 24 hours without telling a single lie, including polite "white lies." Document the exact moment you failed, what duty-based ethics ([Immanuel Kant's Categorical Imperative](https://plato.stanford.edu/entries/kant-moral/)) says you should have done, and how it felt. By shifting homework from passive summary to active, lived experimentation, we ensure that the student's own life remains the ultimate, un-fakeable classroom.