Prompting as a Skill: Should Universities Teach AI Literacy?

My computer science professor told us on the first day that knowing how to code matters less than knowing what to ask the machine to do. I thought he was joking until I watched half my class fail the midterm not because they could not write Python but because they could not break down the problem into steps a program could follow. That moment stuck with me through sophomore year when ChatGPT arrived on campus and suddenly everyone was asking machines to write essays, debug code, and summarize articles without understanding what made a good question.

Universities should require AI literacy courses because students need to learn how to guide these tools effectively, recognize their limits, and use them without sacrificing critical thinking. Teaching prompting as a skill does not mean encouraging shortcuts but preparing students for a workplace where collaborating with AI will be as common as sending email.

AI literacy means understanding how language models work, what they can and cannot do, and how to communicate with them clearly. A prompt is not just typing a question into a search bar. It requires context, specificity, and often several rounds of revision. I learned this the hard way last semester when I asked an AI to help outline a history paper on the Treaty of Versailles. My first attempt was vague, something like "explain the treaty," and I got a generic summary I could have found on Wikipedia. When I revised it to ask for the economic clauses that led to German resentment, then followed up asking how those connected to political instability, the responses became useful starting points for research. That difference between lazy asking and strategic prompting is what students need to learn before they graduate into jobs that assume they already know.

One reason universities resist adding AI literacy is the fear that it legitimizes cheating. My English professor banned all AI use after she caught three students submitting essays that sounded identical. I understand her frustration, but pretending the technology does not exist just pushes its use underground. In my own writing process, I have started using AI to generate counterarguments after I draft my thesis. I type out my position, then ask the model to argue against it. Reading those responses forces me to think harder about weak points in my logic. That is not cheating; that is using a tool to sharpen my thinking the same way I might talk through ideas with a roommate. A required course could teach students this distinction, showing them how to use AI for brainstorming and feedback rather than as a replacement for their own ideas. Without that guidance, students either avoid the tools entirely or misuse them because no one taught them better.

Another argument for AI literacy is that prompting has become a workplace skill employers actually want. My cousin graduated two years ago with a marketing degree and now spends part of every day writing prompts for tools that generate ad copy, analyze customer feedback, and suggest campaign strategies. She said her company assumed new hires would know how to do this, but most of her cohort had no training and had to learn through trial and error. If universities treated AI literacy like writing or data analysis, students would enter the job market with a skill they will use regardless of their major. A biology student might prompt an AI to help design experiments or interpret genetic data. A business student might use it to model financial scenarios. An education major might use it to create personalized lesson plans. The specific applications change, but the underlying skill of asking clear, purposeful questions remains constant across fields.

Some argue that focusing on AI distracts from teaching students to think for themselves. A guest speaker in my philosophy seminar worried that relying on AI would make students intellectually lazy, accepting whatever answer the machine provided instead of wrestling with hard questions. I take that concern seriously because I have felt that temptation. When I was struggling to understand Kant for a midterm, I asked an AI to simplify his arguments, and for a moment I considered just memorizing its summary instead of reading the original text. But then I realized the AI explanation only made sense because I had already spent hours confused by the reading. The tool helped me organize thoughts I was already forming; it did not do the thinking for me. A good AI literacy course would teach students to notice that difference, to use AI as a mirror that reflects their ideas back in new forms rather than as a shortcut past the hard work of learning.

Universities have always adapted their curriculums to prepare students for the world they will enter, not the one their professors remember. My grandparents learned to type on manual typewriters in college; my parents took required computer literacy courses in the nineteen eighties when universities worried students would graduate without knowing word processing. Teaching AI literacy is the next version of that same responsibility. Students will use these tools whether we teach them or not, so the question is whether they will use them well. A required course could cover how to write effective prompts, evaluate AI output critically, understand bias in training data, and recognize when a task requires human judgment instead of machine assistance. I wish I had learned these skills in a classroom instead of piecing them together through mistakes and late-night frustration. The students coming after me deserve a curriculum that acknowledges the tools they will actually use and teaches them to use those tools wisely.

Did you like this example?

268

11