Salisbury University students on campus

ChatGPT and AI Tools

ChatGPT and AI Tools: Implications for Teaching and Learning

Seemingly overnight there has been an explosion in the area of AI tools that automate tasks like writing papers (ChatGPT), creating graphics (GPT DALL-E), or paraphrasing text (Quillbot). While there are a lot of promising aspects of these evolving technologies, the potential impact on academic integrity is concerning. While AI detectors are also emerging rapidly, the accuracy of them is not full proof. Here we provide information and recommendations on ChatGPT and will provide updates moving forward.

What is Chat GPT (Generative Pretrained Transformer) as described by Chat GPT?

  • ChatGPT is a natural language processing tool that uses machine learning techniques to generate human-like responses in real-time chat or messaging systems. It is based on a transformer-based language model, which is a type of neural network designed to process and generate text. ChatGPT is trained on a large dataset of human conversations and is able to generate responses that are coherent and contextually appropriate.
  • Read more about it on OpenAI where you can also create an account and experiment with it, or check out this ChatGPT crash course for beginners.

Concerns surrounding the use GPT or other AI tools:

  • Plagiarism: If students use GPT or similar tools to generate written work, there is a risk that they may submit work that is not their own, or that they have not fully understood and can't explain. This can lead to issues of academic dishonesty and plagiarism. More information on SU’s Academic Misconduct Policy and Procedure can be found on the Student Academic Misconduct Policy page.
  • Limited learning opportunities: If students use GPT or similar tools to complete assignments, they may miss out on the opportunity to practice and develop their own critical thinking skills.
  • Bias: ChatGPT is trained on a dataset of human conversations, which may contain biases that are reflected in its responses. This could lead to unfair or prejudiced treatment of certain groups of students or researchers.

It's important for faculty to set clear guidelines for how GPT can be used in their courses.

  • Consider updating your syllabus to reflect your expectations of how/if GPT and other AI tools may be used in the learning process.
  • This article written by Ryan Watkins at GWU provides a thorough list of things to consider as you develop your syllabi for the semester.
  • The Sentient Syllabus is a phenomenal resource for syllabus language and other considerations in the era of AI.
  • A crowd sourced collection of syllabi statement examples from different disciplines and institutions can be found online here: Classroom Policies for AI Generative Tools form on Google Docs.

Recommendations for course design and assessment of student learning

  1. Learn how GPT would respond to your course assignments. Create an account here to better understand its capabilities and limitations.
  2. GPT is best at regurgitating responses at the lower level of Bloom’s taxonomy. Focus student assessment/work on the higher levels such as analyze, evaluate, create.
  3. Formative assessment, which is used to assess learning while it is still ongoing, allows faculty to better learn their student’s voice, and see the progress of mastery. A good example of this is feedback on drafts of written assignments.
  4. Projects and presentations can be an effective way to assess learning, as they allow students to demonstrate their understanding in a more interactive and creative way.
  5. One limitation of GPT is that it is current only through about 2021. Using current topics, or classroom topics, in your prompt will limit the abilities of GPT.
  6. Create Authentic Assignments which prompt students to complete complex tasks that mirror what they would need to accomplish in a professional setting.
  7. Incentivize the learning process to focus on learning as the outcome. Examples of how to do this include specifications grading, standards-based grading, contract grading, and ungrading.
  8. Join the 2023 Teaching & Learning Conference MyClasses site and watch the panel session on ChatGPT implications for higher education.

Check out more resources on how to best assess learning from ID&D: https://kb.salisbury.edu/display/IDD/Assessing+Learning

Final Thoughts

  • Calculators were invented in the 1960s and have since become a widely accepted and indispensable tool for a variety of tasks, including scientific, business, and personal use. When calculators were first introduced, they were met with both enthusiasm and skepticism. Some people saw them as a valuable tool that could significantly improve efficiency and accuracy, while others were concerned about their potential impact on employment and the skills of the workforce. Similarly, ChatGPT and other artificial intelligence technologies are being developed and used in a variety of fields, including higher education. Like calculators, ChatGPT has the potential to improve efficiency and accuracy in certain tasks, but it also raises ethical and social concerns which need to be carefully considered. It is important for institutions and individuals to carefully evaluate the potential benefits and challenges of using ChatGPT and to consider any necessary safeguards or protocols to ensure its responsible and ethical use.

     

Interested in learning more? Here is a curated list of resources and information for teachers and faculty: AI Text Generators: Sources to Stimulate Discussion among Teachers, Compiled by Anna Mills and licensed CC BY NC 4.0.