Teaching Students about AI

Ensuring students have a shared understanding of GenAI basics

  • Keep things simple. Students likely already have some experience with GenAI models and knowledge of how they work. 
  • Ensure students are provided with key definitions for any terms you plan to refer to in your own course (e.g., “large language model,” “machine learning,” “GenAI”). 
  • Make clear to students that LLMs and other GenAI tools use probabilistic algorithms to create outputs; they do not process or “think” about information the way a human does and they cannot assess or verify their outputs for accuracy or appropriateness. 
  • Briefly discuss how LLMs and similar systems are trained, and how an LLM can only “know” what it is able to reference in the data to which it has access.

Resources you can share with students: 

Resources you can request to help support students in class:

Drawing student awareness to how AI can support or undermine their learning

  • Discuss how AI use affects the learning and skills students are expected to develop in your discipline. In what ways can AI be used to support or enhance learning? How might using it undermine learning? 
Examples

Example: UW student surveys about AI use indicate that students often view using GenAI to summarize readings as acceptable and report engaging in this practice regularly. However, instructors have discovered that relying on GenAI summaries can negatively affect student learning by limiting students’ ability to comprehend and critically engage with topics, parse complex texts, and notice key information that the summaries omit or misrepresent.

  • Discuss issues/concerns students should be aware of when using AI tools for learning or productivity purposes.
Examples

Example: Statement Permitting AI Use. Adapt this assessment policy from STAT 341 which permits GenAI use, and which explains when and how it is most beneficial for students to use GenAI, and includes the expectation that students document and cite their GenAI use. In particular, consider the “Additional Clarification” section.

Resources you can share with students: 

Explaining your course policy on AI use  

  • Explain your thinking around the learning you want students to do in your course and what it is critical for them to be able to do without the assistance of AI, if anything. 
  • Be explicit if interaction with other humans supports the learning you expect students to do in your course and explain how AI use may inhibit or support that learning. 
  • Discuss how researchers and others in your discipline are using AI in their work, the opportunities and benefits this use is bringing to the field, and what you hope students explore or accomplish using AI in your course. 
  • Discuss the ethical concerns around AI use that are common in your discipline, as well as your own personal concerns and those that students may bring with them. Common concerns include: 
    • Environmental impacts (energy, water, and land usage) 
    • Bias in GenAI outputs 
    • Risks of social harm (misinformation, fraud, AI psychosis, worker exploitation). 
    • Concerns about the practices of the businesses that create and maintain AI tools.
Examples

Example: Embedding Ethical Considerations for AI Use in Image Generation. Consider this course assignment structure from DAC 329/COMMST 329 where students are encouraged to use AI while developing their understanding and application of AI ethics, from consent and copyright to environmental costs.

  • Identify how job-market expectations in related fields may be shaping your course decisions. 
  • If permitting only certain uses of AI in your course, clearly explain how and when students can use AI and why you have made these choices in your course assignment descriptions. Let students know that if they are unsure about specific AI use cases, they can ask you (or their TA where appropriate) for further assistance.  
Examples

Example: Statement Permitting AI Use for Specific Purposes. Adapt this example of an assessment policy from REC 100 which includes instructions and examples for students to document and cite their AI use. In particular, consider how the instructor explains her rationale and describes prohibited and permitted uses of AI.

  • Define expectations for permitting use of GenAI for translation by multilingual and international students as there are varying and sometimes much more liberal GenAI practices in other countries' education systems. If AI use is permitted to translate course materials, be clear with students which materials are included (e.g., lecture notes, slides, assessment questions), and which tool would be acceptable for them to use.  
  • Provide resources students can use to help guide their use of AI in your course (see resources below). 

Resources you can share with students: 

Preparing students for AI integrations in your course

  • Model the type of AI use you want students to become familiar with or engage with.
Examples

Examples:

  • Identify which tools are best for various discipline-specific tasks and the benefits they provide.

  • Invite students to explore AI with a discipline-specific lens, and model how they can apply what they are learning to develop better prompts, agents, or outputs.

  • Share and deconstruct what low-quality or unreliable AI output looks like in your discipline, how to recognize it, and how to avoid producing it.

  • Demonstrate how AI biases reveal themselves in your discipline and what you can do to counter them. 

  • If designing an activity that has students use an unlicensed AI tool, remind students to examine the tool’s Terms and Conditions and ensure they understand if and how their privacy will be impacted while using the tool. Ensure they understand that by using these AI tools, they will be entering into an agreement with the company that governs how they are permitted to use the tool and how the company will be collecting the student’s data. 
Important

Important note: Ensure that students can opt out of using unlicensed AI tools. You cannot require students to accept a third party’s terms or to relinquish their personal data or intellectual property. If the tool is part of a required activity or assessment, you should provide a viable alternative for students who object to using it.

  • If a course learning outcome is for students to develop AI literacy skills, it is reasonable to require students to use AI. However, depending on the expectations in your program about which courses are to include AI literacy skills, you may choose to offer an alternative assessment or a variation on an activity if a viable option is possible (e.g., you provide AI-generated output for students to analyze). 
Examples

Example: When possible, create activities and assessments that allow students to achieve AI literacy outcomes without using AI tools themselves. Ask yourself what you are assessing: Students’ ability to use AI, or their knowledge of AI and ability to critique it?

  • Remind students that they are responsible for their decisions about when and how to use AI, and for what they produce with AI.
  • Provide students with guidance and resources to help them appropriately disclose, document, and attribute their AI usage (see resources below). 

To learn more: Read

Resources you can share with students: