AI Tools for Ethical Learning

How can we use AI tools ethically and responsibly as students and learners? This is an ongoing discussion and exploration.

Created for an Academic Literacy Workshop in September 2024, this guide is a launch pad for any student to using the free version of basic generative AI tools - ChatGPT, Gemini, and Copilot - to complete student tasks ethically.

Fig. 1. A cartoon image of a genie with outstretched hands. DALL·E, version 2024-09-17, OpenAI, 17 Sep. 2024, openai.com/dall-e. 

So, let's start with a quick poll.

Another poll for the ✅ YES ✅ crowd

We prompted ChatGPT (free version)

explain how to use AI technologies ethically and responsibly to college freshmen in a comic strip format

Here's what we got.

Sadly, ChatGPT (free) does not allow us to create a public link to a conversation that includes a generated image. Here's the list it generated for this comic explainer:

Consider: How well did the AI perform the task?

Decide:

Dive Deeper

Consider: What risks do we face using AI tools? What benefits do we gain?

Here's what our students said as analyzed by Gemini:

THE RISKS

• Academic Integrity: Plagiarism, lack of authenticity, and the risk of not learning are major concerns.
• Skill Development: AI could hinder the development of writing, problem-solving, and research skills.
• Dependence: Overreliance on AI could lead to a loss of independence and creativity.
• Bias and Misinformation: AI can perpetuate biases and spread misinformation.
• Security and Reliability: There are concerns about the security and reliability of AI tools.

THE BENEFITS

• Efficiency and Time-Saving: AI can save time on tasks like research, writing, and editing.
• Support and Assistance: AI can provide ideas, perspectives, and feedback.
• Creativity and Innovation: AI can help spark creativity and generate new ideas.
• Learning and Development: AI can help students learn new skills and improve their writing.
• Mental Health: AI can offer emotional support and reduce stress.

What would you add or remove from our list?

What specific tasks can AI help us with if we use it ethically?

Here's what our students said as analyzed by Gemini:

  KEY THEMES

• Writing and Editing: AI can help with proofreading, rewording, and structuring writing.
• Learning and Understanding: AI can explain complex topics, summarize information, and create study plans.
• Organization and Planning: AI can assist with tasks like time management, outlining, and meal planning.
• Research and Analysis: AI can help with research, finding information, and data entry.

SPECIFIC TASKS

• Writing: Proofreading, rewording, and structuring writing.
• Learning: Explaining complex topics, summarizing information, and creating study plans.
• Organization: Time management, outlining, and meal planning.
• Research: Finding information, data entry, and creating bibliographies in MLA format.
• Problem-solving: Breaking down complex topics into simpler terms and understanding math problems.

What would you add to our brainstorm?

 Think Critically about AI Output

Every free generative AI tool includes a disclaimer that it makes mistakes. Plan and use knowing the information could be wrong or completely made up. There are multiple studies on response accuracy, but most are highly specific to a discipline or type of information need. See this sample search for example, which is limited to peer-reviewed journal articles found in our library collection.

 Cite AI

Knowing AI makes mistakes and might be wrong, if you decide to copy/paste AI responses into your work, you must cite it. How you cite it depends on the citation style you are using. If you are allowed to include AI-generated content depends on your instructor, so review the class syllabus and check with the instructor directly.

How you give AI your task will dramatically impact the results you get. The task is called prompt engineering, and you can find many tip sheets on writing a good prompt.

For academic purposes, we suggest this prompt building strategy.

1. Assign a Role - Who should the AI be when responding to our question?
2. Define the Task - What do we need the AI to do for us?
3. List Requirements - What guidelines should the AI follow?
4. Add Context - What information will help the AI understand the task?
5. Converse - How can you follow-up with the AI's response to guide it toward your desired result?

Practice Prompt Building!

Our students wrote prompts for these scenarios if you want to give it a try

Scenario 1:

You are struggling to understand the concept of rational functions in your college algebra course. You’re unsure of how to graph them and find the asymptotes. 

Scenario 2:

You are assigned a text to analyze for your history class: President John F. Kennedy Jr.'s address during the Cuban missile crisis. The text discusses an event with varying viewpoints, and you need help understanding those perspectives. 

Scenario 3:

You have been instructed to write a research paper on a topic that interests you. You were told that you will need an introduction, 3 main points, and a conclusion.

Test Those Prompts

Fun things to try:

• Use the same prompt in different generative AI tools and compare the results.
• Pick a tool to prompt. Wait a day, and ask the exact same question again. Compare the results.
• Ask how many r's are in the word strawberry.

AI tools can be incredibly helpful, but they're not infallible. They can make mistakes or misunderstand things, just like a genie might grant a wish that goes awry. It's important to remember that AI is a tool, and like any tool, it needs to be used with care and understanding.

Think critically about AI's response to your prompt. Here are a few questions you might ask yourself.

• Was the output accurate and relevant to your task? 
• Did the AI provide any unexpected or useful insights? 
• Were there any biases, inaccuracies, or inconsistencies in the output? 
• How will you validate or verify the information provided by the AI? 
• What are the limitations of using AI for this task, and how will you address them? 
• Is this an acceptable use of AI, an unacceptable use, or a gray area? If it's in the gray, should I still use it?