Research Guides: Generative AI: Gen AI for Teachers

What is Generative AI?

Generative AI is a type of artificial intelligence that uses algorithms to create new content such as images, music, and text. It involves training machine learning models on large datasets to learn patterns and generate novel content that is similar in style and structure to the data it has been trained on. In other words, it allows machines to produce new outputs that resemble human creations.

On March 15, 2022, the Office of the Provost released interim guidance for using these and other generative AI tools for academic work.

Learn More: Nate Chambers, of the USNA Computer Science department, provides an accessible overview of ChatGPT on YouTube: https://www.youtube.com/watch?v=fMiYNrjDPyI.

Syllabus Language

These resources can help you guide students who are using or considering use of generative AI tools.

Recommended Gen AI-Academic Integrity Statement for Syllabi and Sakai
A general statement with links to further guidance meant to complement current academic integrity/Honor Code statements.

Being Clear with Your Students: Syllabus and Sakai Statement Options for Faculty
A wide range of example language and guidance to fit your requirements and preferences, plus links to more information from teaching and learning centers and the NPS Generative AI Resource Hub.

How Can I Use Generative AI for Teaching?

Here are a few ideas to consider.

Activities or Assignments

Generate novel scenarios or cases and ask students to apply theories or models to them. Scenarios can be the basis of discussion or assignments.
Ask ChatGPT to generate text on a subject students are learning. Then, ask students to assess the text—how well do they think the AI did? What did it miss? What kind of understanding isn’t reflected in the output?
ChatGPT might be used to support wargaming exercises (perhaps generating scenarios, decision support, after-action, or supporting roleplaying and immersion).

Administrative

Draft, revise, or adjust tone in emails to students. Create boilerplate emails for common student questions or requests.
Draft or revise text to use in syllabi, lesson plans or assignment prompts.
Generate boilerplate comments for common feedback on papers or assignments.
Draft sample text for writing rubrics that demonstrates characteristics of different grading levels.
Create test questions or test question answer options.
Find trends and summarize qualitative student feedback in course evaluations.
Test your assignment prompts to learn how well ChatGPT answers them.

Explore Further:

A group of higher education researchers have asked educators to submit a collection of ideas for use of AI in education.
- https://creativehecommunity.wordpress.com/2023/02/02/creating-a-collection-of-creative-ideas-to-use-ai-in-education/
Writing instructor Anna Mills presented "Rethinking Writing for Assessment in the Era of Artificial Intelligence" (full video) at the UCSD Academic Integrity Office's recent virtual symposium.
Also at UCSD, the Halıcıoğlu Data Science Institute, Institute for Practical Ethics, and San Diego Supercomputer Center hosted "Chatting GPT" (full video). The final panel discusses implications of ChatGPT for teaching programming.

“Generative AI: A Case Study on Classroom Applications”

Dr. Lucie Moussu of Royal Military College presented during the Writing Center Consortium for Grad PME's winter workshop series.

You can review her slides here; a video link is forthcoming.

DESCRIPTION: Post-secondary institutions are entering a deeply transformative era with the advent of Generative AI (GenAI). This workshop will delve into the possible uses of GenAI in reshaping course development processes, including the design of syllabi, assignments, activities, and assessments. This workshop will explore how this technology can simplify course creation, foster innovative teaching approaches, enhance student engagement, and encourage ethical and effective uses of GenAI. Participants will gain hands-on experience with AI tools and learn strategies to prepare students for a future where AI will be an integral part of their lives.

Citing and Referencing Generative AI

The DKL Citation LibGuide now offers specific guidance about what to cite and when to disclose use of generative AI in academic work. Examples for each citation style are forthcoming.

Research Guides: Citation Guide: Generative AI

libguides.nps.edu

Learn how to cite articles, books, reports, theses, government documents, etc. for NPS theses, papers, and publications.

How Can I Deter Unacceptable AI use?

Can We Detect Generative AI?

Detection software tools, while rapidly developing, remain inconclusive in their results. We do not recommend relying on these tools at this time.

How Much Do Assignment Prompts Matter?

Adapting assignments offers some potential. However, given the rapid rate of AI development, current deficiencies of AI are unlikely to persist. Current forms of tools like ChatGPT are less immediately adept with prompts requiring close discussion of class materials, in-depth analysis, and clear citation and integration of source material.

What Can We Do in Class?

Model appropriate use of generative AI tools in class or through clearly guided assignments. Create opportunities for students to assess and critique generative AI output, encouraging them to engage with it critically.
Have conversations with students about the role writing plays in your discipline. How does writing support processing, deep learning, synthesizing, inspiration, and conversation with other texts?
Incorporate activities that are multi-modal and/or in-person (or synchronous), like posters, video, podcasts, presentations, or debates.

How Can We Leverage Communication and Course Climate?

Make clear your expectations about use of generative AI tools. Maintain open lines of communication—students may need to reinforce their understanding of what you want them to do.
When students feel it is critical to produce a “perfect paper"—when stakes feel high—it is more tempting to cheat or cut corners. Emphasizing learning over performance, lowering the stakes of individual assignments, and fostering intrinsic motivation can help.
Try applying a “process, not product” ethos with scaffolded assignments that build on each other and incorporate drafting, feedback, and revision.

Writing Center Consortium for Graduate-level Professional Military Education (WCCG)

By faculty and for faculty, the WCCG-PME hosts an annual multi-week series of hands-on workshops, offering faculty at military schools an opportunity to share practices that improve student writing outcomes. Three of six 2024 workshops focus on gen AI:

Generative AI: A Case Study on Classroom Applications, by Dr. Lucie Moussu, Royal Military College of Canada, 31 Jan 2024
- Slides
From Classroom to Command: AI Literacy in Higher Education and PME, by Chloe Woida (ctr), Naval Postgraduate School, 15 Feb 2024
- Slides
- Video
How to Use Generative AI to Teach Critical Thinking, by Carolyn Stoermer (ctr), Air Force Institute of Technology, April 10, 1100–1230 Pacific
- Register by April 8

Problems and Promise of Generative AI in Higher Education

Some of the major concerns educators have around generative AI include:

Academic integrity: ChatGPT can rapidly generate complete essays from provided prompts, which may tempt some students to use it inappropriately.
Educational impact: In some disciplines, writing tasks are also viewed as critical cognitive tasks that support and enable learning; the more those tasks become automated, the less students benefit.
Insufficient information literacy: Students may not have the skill set needed to assess the quality of information and sources offered by generative AI tools. They may assume information is accurate, despite the possibility of false data, bias, or misinformation.
User data and privacy: OpenAI, which offers ChatGPT, is a for-profit company and collects user data; the full extent of how this data will be used remains unclear.
Ethical and legal issues: Issues persist around copyright, the ethical implications of OpenAI's use of human labor in classifying graphic content, and environmental consequences of technology with significant computing demands, among others.

However, there are many reasons to be excited about generative AI's potential, including its potential for teaching:

Building AI literacy: Military AI strategists emphasize the need for AI literacy throughout service levels, which can be supported by incorporating and teaching effective use of ChatGPT and similar AI tools in PME contexts.
Reinforcing information literacy: Engaging directly with the limitations of generative AI output can serve as a testbed in which to build data and information literacy of students and the force.
Shifting cognitive load: Tools like ChatGPT may help both faculty and students redistribute cognitive load, streamlining lower-level tasks and enabling them to apply higher-order thinking where it counts.
Educational equity and personalized learning: Generative AI tools may facilitate personalized learning, helping faculty adapt material easily to individual learner needs and leveling the playing field for diverse learners.

Explore Further: UC Berkeley's Center for Teaching and Learning offers an "Opportunities and Threats" analysis with further reading on their page, "Understanding AI Writing Tools and their Uses for Teaching and Learning at UC Berkeley."

Generative AI