Skip navigation

ChatGPT and AI Technology

Resources and information related to ChatGPT, AI, and Large Language Models (LLMs) in teaching and learning.

What is ChatGPT? 

ChatGPT is a large language model developed by OpenAI. It uses machine learning algorithms to understand and generate human-like language in response to text-based inputs. ChatGPT was trained on a massive corpus of text data, including books, articles and websites, allowing it to develop a vast store of knowledge on a wide range of topics. Its main purpose is to converse with humans in a natural, conversational way, providing helpful responses to questions and engaging in discussions. While ChatGPT has had a lot of publicity in recent months, it is only one example of a wide range of generative AI tools being developed for a wide range of purposes. 

Large Language Models

A large language model is a type of machine-learning algorithm that is trained to understand and generate human-like language. It works by analyzing vast amounts of text data, such as books, articles, and websites, in order to learn the patterns and structures of natural language. This training data is used to create a complex neural network that can understand and produce language in a way that is similar to humans. 

The term "large" refers to the size of the neural network and the amount of training data used. Large language models typically use billions of parameters and are trained on massive datasets, which allows them to generate more sophisticated and nuanced responses.

Large language models are capable of a wide range of language-related tasks, including language translation, language generation, and language understanding. They are often used in natural language processing (NLP) applications, such as chatbots, virtual assistants, and language analysis tools.

Some Other LLM Generative AI Examples.

  • Open AI: ChatGPT, GPT-3
  • Google: Bard, PaLM
  • Deepmind: Chinchilla, Gopher, Sparrow

Limitations, Equity Concerns and Technical Prevention

  • As powerful as they are, large language models regularly produce inaccurate, misleading or false information (and present it confidently and convincingly).
  • ChatGPT responses will vary even when the same prompt is used or based on slight differences in the initial prompt. 
  • The archive used to train ChatGPT was compiled in 2021. At this time, it has limited knowledge of the world and events since then although it is likely to be updated in the future. 
  • There are biases inherent in the LLM technology and there are concerns with how the tools were trained to deal with explicit content. ChatGPT even states that it may occasionally produce harmful instructions or biased content.  
  • There is not currently a way to accurately block, prevent or detect AI writing in submitted work. While UCO's official plagiarism detection tool, Turnitin, has a preview version of its AI detection feature available in the similarity report, other detection tools are not supported by UCO OIT. 

Academic Integrity, Policies and Student Conduct

  • Syllabus Statements can help instructors set clear expectations for students on how AI and other technologies can be used in a particular course. These statements should be clear and focus on what is authorized and unauthorized use for the course in question. 
  • Get to know your students and have an understanding of their writing styles, skills and habits so that you are able to more easily identify when they are writing things in their own voice.  While doing this, also be aware of and examine your own biases and expectations for what you expect from your students. 
  • Remember, cheating isn't new so it will be important to focus on making it less appealing to students and create assignments that are more difficult for an AI to write. Things like reflections, personal stories, etc. are some good examples. 
  • Take a look at your institution's code of conduct and the parameters around both "cheating" and "plagiarism" as they are constantly developing. And engage in conversations with these offices to see how AI is being incorporated or accounted for in your institution's policies. 
  • The APA has released guidance for students and researchers on citing ChatGPT in their work. 

UCO's Office of Student Accountability & Conflict Resolution Statement 

Released April 6, 2023

Use Caution with AI Detection Tools for Academic Dishonesty

The UCO Office of Student Accountability and Conflict Resolution has received questions about how to detect cheating using artificial intelligence. 

Student Accountability and Conflict Resolution is aware of several AI detection tools, including a newly released function with Turnitin. However, the office cautions against using only these tools as their accuracy has been shown to be inefficient and spotty. 

Those concerned that a student has used AI in an unauthorized manner should email Student Accountability and Conflict Resolution for assistance to work through the issue.

21CPI March 2023 Panel Discussion

Have Humans Become Obsolete? Presentation Slides
March 22, 2023 Session Session Recording

AI in Education Settings

Writing Assignments with AI

Most instructors want to keep AI out of writing assignments; however, AI can play an incredibly helpful role when approached with specific intention. Students can use ChatGPT to assist with the brainstorming process, similar to how they might talk to a friend over coffee about the assignments, or even to assist with the early draft process by reviewing their initial work and providing feedback. An alternate AI resource such as Elicit.org could assist with research and summarizing resources, although it is highly recommended that students fact check any resources they are provided by AI! 

When using AI in writing assignments it is important to:

  • Keep an open mind when considering the best uses for your course;
  • Be open/honest with your students about what AI does and does not do well;
  • Encourage students to be transparent about their use of AI-generated content; and
  • Be clear about what is considered authorized vs. unauthorized use in your classroom. 

ChatGPT in STEM 

There are a number of ways that STEM-based instructors can take advantage of technology like ChatGPT in their classrooms. In addition to assisting with the research process, AI can help to explain and clarify complex concepts to students. It may even present the material in a different way that can help make connections for the student to other materials or concepts. Additionally, resources like ChatGPT may be able to assist when dealing with programming issues or lack of knowledge, as well as troubleshooting where a particular piece of code goes wrong. Another possibility is that students could ask the AI to provide practice questions that can help them practice and prepare for exams. 

It will be important for you to work with your students to understand the limitations of ChatGPT in your specific field. For example, ChatGPT is known to make common errors that a student might make, which can be misleading when a student is attempting to fact check. Additionally, it may not always understand the context of a question and could then provide incorrect or incomplete answers. 

Research and Scholarship Uses for AI 

There are many ways ChatGPT can be used to speed up processes, reduce busy work and handle a portion of the workload for those professionals who are already overloaded. Here are a few options when considering the use of generative AI tools: 

  • Generating conference presentation ideas/proposals and topic bullet points;
  • Quickly summarizing large amounts of content into a short synopsis;
  • Creating eye-catching titles for presentations;
  • Resolving writer's block for difficult-to-articulate concepts, emails, presentations, etc.; and,
  • Condensing content to meet word requirements.

Additional Resources

AI Introduction

AI in Education

Other Interesting AI Information