The following essay is a more readable version of an academic paper I will present at the 18th annual International Conference of Education, Research and Innovation (ICERI2025) in November. It extends ideas from my previous article about vibe-coding here on The Augmented Educator, exploring what happens when this problematic metaphor migrates from software development into educational practice.
Coined by AI researcher Andrej Karpathy in February 2025, the term “vibe-coding” described a crucial element in the early adoption of generative AI among software developers. The phrase described a programming approach where developers guide AI through natural language conversation, selecting outputs based on intuitive assessment rather than traditional line-by-line code composition. Within weeks, the term had spread throughout the software development community, its casual tone perfectly matching the seemingly painless way AI could generate functional code.
The appeal of vibe-coding’s effortless promise soon reached classrooms. Teachers and students, facing their own pressures to integrate generative AI, began adopting comparable language. Terms like “vibe-teaching” and “vibe-learning” emerged to describe how teachers and students were using AI tools.1 A teacher might “vibe” their way through lesson planning by prompting ChatGPT until something usable appeared. Students might “vibe-learn” by iterating with AI until assignments felt complete. The metaphor suggested that educational AI use could similarly be approached through feel and intuition.
But here’s the issue: as I have already pointed out in a previous essay, the vibe metaphor mis-characterizes effective practice in both domains. Quality software development, like quality education, requires directed search and purposeful iteration, not casual browsing. When developers create robust applications, they engage in systematic exploration guided by clear objectives. They test hypotheses, evaluate outputs against specifications, and refine approaches based on feedback, which differs from the undirected trial and error that “vibing” implies. This becomes even more apparent in education, where effective practice demands intentional design and careful calibration of challenges to promote learning.
What “Vibe-Teaching” Actually Looks Like
Consider how vibe-teaching would manifest in practice. A pressed-for-time teacher needs materials for tomorrow’s lesson on photosynthesis. They open an AI assistant and type: “Create a lesson plan on photosynthesis for 9th graders.” The AI generates something that looks reasonable: an introduction, some activities, discussion questions, a worksheet. The teacher skims it, makes minor adjustments, and considers the task complete.
What’s missing from this interaction? The teacher hasn’t considered their specific students’ prior knowledge or misconceptions. They haven’t thought about how this lesson connects to yesterday’s discussion or next week’s lab. The AI cannot know that three students in second period have reading difficulties, or that the class became fascinated by a question about plant respiration last Tuesday. These contextual factors that expert teachers constantly consider vanish when teaching materials simply emerge from generic prompts.
Students might engage in similar practices. Facing an essay assignment, they could turn to AI seeking a finished product. “Write a five-paragraph essay about the causes of World War I.” When the result seems too simple, they refine: “Make it more sophisticated.” “Add another historical example.” They evaluate outputs through surface features rather than understanding. “Does it sound academic enough? Is it the right length?” The actual learning that should occur through researching, evaluating sources, and constructing arguments never happens.
It has to be noted that these practices exist on a spectrum. Some teachers might thoughtfully adapt AI suggestions based on pedagogical expertise. And some students could use AI for brainstorming while maintaining intellectual ownership. But the problem is that the vibe metaphor provides no framework for distinguishing beneficial from harmful uses. Its emphasis on intuitive, affect-driven interaction offers no criteria beyond subjective and instant intellectual satisfaction.
Why Vibing Fails Education
The incompatibility between vibe metaphors and educational purposes reveals itself across multiple dimensions. First, effective pedagogy begins with clear learning objectives that inform every subsequent decision. Teachers engage in what educators call “backward design,” starting with desired outcomes and working backward to create aligned experiences. This process requires deliberate planning, not intuitive browsing. When teachers vibe their way through lesson creation, they replace intentional design with chance. Even when resulting materials appear adequate, they lack the coherent progression that characterizes expert instruction.
Second, psychological research consistently shows that struggle and effort are essential components of learning. Concepts like “desirable difficulties” highlight how cognitive challenge drives retention and transfer. When students create AI-generated essays without engaging in the writing process, they bypass the necessary struggle. Writing involves analyzing sources, organizing ideas, and finding language for complex thoughts. Each activity strengthens intellectual capacity. The AI-generated essay may earn a grade, but it leaves the student’s capabilities unchanged.
Third, the vibe framework masks essential professional labor. Teaching expertise develops through years of practice and refinement. Expert teachers possess deep content knowledge and finely tuned abilities to read student understanding. When these teachers use AI tools properly, they apply this expertise in evaluating and adapting outputs. They recognize when suggested examples might confuse rather than clarify. They understand how to sequence activities for their particular students. But if teaching involves simply generating materials that feel right, what distinguishes the professional educator from anyone with AI access?
The systemic implications extend further. Educational institutions adopting vibe approaches risk creating environments where students with access to thoughtfully integrated AI receive enhanced education, while those whose teachers rely on vibing receive generic instruction. The apparent democratization of content through AI could paradoxically increase inequality. Meanwhile, routine submission of AI-generated work threatens the meaning of assessment itself. If credentials no longer reflect actual learning, the entire educational enterprise collapses into an empty performance.
Keep reading with a 7-day free trial
Subscribe to The Augmented Educator to keep reading this post and get 7 days of free access to the full post archives.