To teach in the time of ChatGPT is to know pain
Summary
A college instructor describes the demoralizing impact of large language models on asynchronous online teaching, citing difficulties in detecting AI-generated work and the time sunk into grading. The piece argues for higher-quality, non-cheatable assignments and outlines tensions between preserving online access and maintaining academic integrity, suggesting the friction of AI adoption will shape pedagogy for the foreseeable future.