By Thomas Jilk

When David McHugh joined the Information School as a teaching faculty in late 2022, the release of ChatGPT was beginning to inject uncertainty and trepidation into conversations about the future of education. As AI use on campus became widespread, it was clear to McHugh that students wanted to know if, when, and how to use the tools.  After all, “Using AI tools is not the same as using them competently or thoughtfully,” he said. 

To that end, one of McHugh’s first big projects at the iSchool was creating a non-technical course around generative AI, a perfect opportunity to engage students on the thorny but important questions in the space. How can students use AI tools in educational settings without degrading the learning process? What is the future of digital privacy in the AI era? How are employers incorporating these technologies into their organizations? 

McHugh’s unique professional background across instructional technology, digital librarianship and games for learning gave him a valuable perspective for analyzing and critiquing interactive software tools, which carried over nicely to teaching about AI. In his new course — LIS 408: Generative Artificial Intelligence: Strategic Application, Evaluation and Critique — McHugh offers practical lessons to help students navigate the technology with clarity and judgment, including: 

• Be a thoughtful information analyst 
• Protect the learning process 
• Consider your privacy 
• Understand ingrained bias 
• Prepare for an uncertain work future 
• Prioritize human connection 

The course is intentionally designed for students who will use AI in their academic, professional, and personal lives, rather than those building the algorithms. In this way, it aligns closely with a core purpose of the new College of Computing & Artificial Intelligence (CAI). By enabling students to approach AI with curiosity, intention, and a healthy dose of skepticism, it prepares them to responsibly shape the future. “We help students develop durable habits, like asking better questions of AI systems, being discerning information practitioners, and making informed judgments about when these tools support learning and when they automate too much,” McHugh said.  

While the course today serves largely Information Science undergraduate majors, its lessons speak to a much wider audience; nearly half of Americans are now using AI chatbots, even as many express discomfort about the technology’s impacts. Below are six key takeaways McHugh shared from LIS 408, which hold relevance beyond the classroom for anyone navigating the AI age. 

1. Be a thoughtful information analyst 

Digital literacy begins with understanding how information is created and shared online. “Students quickly see that AI can surface sources that look credible but are not, and it can misrepresent what real sources actually say,” McHugh said. “They practice tracing claims back to their origins, asking whether a piece of information is in conversation with other sources or standing alone.” 

They also learn to separate hype from data, a nuanced skill that is more relevant than ever. Headlines about AI can exaggerate either capabilities or risks, so students build the habit of examining common marketing tactics and evaluating whether a claim is supported by evidence. In one hands-on exercise, students work in Google AI Studio to build their own media-credibility analyzer, defining three to five criteria, such as source attribution or emotional language, and writing system instructions to enforce them. 

Building these skills is useful for anyone online today, because interrogating information rather than accepting it at face value has never been more important, McHugh said.  

2. Protect the learning process 

A central theme of the course is that AI can make learning feel easier in ways that bypass important cognitive work. It can essentially eliminate the space between asking a question and finding an answer, McHugh said. “Automating that space too often leads to what I call ‘the vibes of learning,’ or the feeling of progress without the substance,” he explained. 

For example, one of the useful ways to use AI without undermining learning is to ask the tools for feedback on a written draft, McHugh said. On the other hand, using AI to brainstorm may actually restrict what you can accomplish. “Many people lean on AI for early-stage idea generation, but if AI supplies the initial ideas, your thinking narrows around whatever the model suggests,” McHugh said. That’s because defaults are powerful; in any system, most people stick with whatever appears first. As a result, what you think is opening your mind to ideas may actually be limiting your possibilities. 

In general, McHugh added, “For long-term learning, if it helps you save time, it may be a bad use case. If it takes longer but leads to higher quality, that’s probably a good use.” 

3. Consider your privacy 

Students today live in an information environment where data is more permanent than ever. AI systems complicate this further. “Once information enters a model, it can’t realistically be surgically removed,” McHugh noted. Even if regulations like those in the European Union require companies to erase someone’s data on request, there is no equivalent delete function for generative AI training data, because the information is spread out across parameters. 

This phenomenon raises practical questions: What do we want AI to know about us? What should stay out of these systems entirely? How do we protect our own and others’ data in this new world? “In the course, we talk in detail about the tradeoffs between convenience and privacy, and why protected, institutionally supported tools matter,” McHugh said. 

4. Understand ingrained bias 

Students in LIS 408 also learn how AI models inevitably reflect the biases of the people and data underlying them. And because AI blends countless data sources, biases can interact in unpredictable ways. One example comes from AI-assisted grading in a K-12 context. The educator Leon Furze tested how an AI system would grade the exact same student assignment if only the student’s name on the paper was changed. The results ranged from a 78 to a 95. (“Such inconsistency completely undermines the fairness and reliability of the grading process,” Furze noted.) 

In another class exercise, students design a prompt to evaluate a sample resume against a job posting, then, in a second round, try to nudge the AI’s score to 100% (and then toward zero) using the fewest possible edits. Watching a resume’s fate swing on cosmetic changes rather than substance drives home how easily these systems can be gamed, and how little that has to do with a candidate’s actual qualifications. 

Bias can emerge from language patterns, cultural markers, dialects, or even the way a name is perceived. Understanding this helps students calibrate trust and distrust, McHugh explained, an essential skill in any information profession.   

5. Prepare for an uncertain work future 

With many unanswered questions lingering around the future of work, at least one thing is certain: Students entering the workforce will encounter AI, whether they seek it out or not. “Many will be asked to ‘figure out the AI stuff’ simply because they are young or took a class on the topic. They need to be ready for that moment,” McHugh said. 

That means knowing when AI is appropriate, when it may be harmful, and when it is simply unnecessary for their specific roles. It means being able to articulate risks to colleagues, evaluate tools, and push back when automation undermines the goals of a project or the values of an organization. This mindset can apply not just to work, but to their personal lives as well. 

6. Prioritize human connection 

Students need to examine how technology, including AI, is impacting their daily lives, McHugh said. “We have increasingly good tools for many tasks, but we also live in an attention economy where distraction is constant,” he emphasized. “AI accelerates this, making it easy to automate not just tasks but entire learning experiences.” He worries that as generative AI continues to permeate society, people could spend “increasingly more of our lives staring at and communicating with computers, and less with humans.” Thus, he said, it is now extra important to figure out the “right-size” amount of technology in our lives to enhance learning, relationships, and attention. 

McHugh’s course is intentionally human‑centered for this reason. “We use whiteboards, move around the room, collaborate, and create artifacts together,” he said. It’s an ongoing exercise in sharpening judgment around when to engage with the tech and to continually engage fellow learners.  

Today, chatbots are excellent at generating adequate answers. At CAI, “Our job is to focus on the questions,” McHugh said. If students ask good questions of the technologies, and of each other, they will be equipped to thrive in an AI-shaped world and emerge as the next generation of AI-savvy professionals and leaders. 

---