Skip to content

A conversation between ChatGPT and Mark Gstohl on teaching, learning, and artificial intelligence.

ChatGPT is an artificial intelligence (AI) chatbot developed by OpenAI. It's been in the news a lot since its launch in late 2022.


Dr. Mark Gstohl is CAT+FD's Associate Director for Programming. He's also an associate professor in the Theology Department who has taught at Xavier since 2000.


Links for this episode:

Transcript:

...continue reading "Conversation #116: ChatGPT interviews Mark Gstohl"

computer artificial intelligence

If you have been following the news lately, you have heard a lot about generative Artificial Intelligence (AI) programs like ChatGPT. The advancements in AI writing technology have created a sense of unease among educators, who fear that students may use AI writing tools to plagiarize their work.

Last week Turnitin announced the launch of AI writing detection capabilities in Turnitin Feedback Studio. Turnitin stated that their technology is capable of identifying both AI-generated and AI-assisted writing, including ChatGPT.

An AI writing indicator has been added to Turnitin Feedback Studio. This indicator shows possible AI-generated content. Students will not be able to see the AI writing detection indicator.

AI writing detection indicator
AI writing detection indicator

Example of AI Writing Detection report
Example of AI Writing Detection report

If you intend to use the Turnitin AI Writing Detector, it is important that you understand the results may NOT be reliable.

A Washington Post article, by Geoffrey A. Fowler, reported that Turnitin’s ChatGPT detector flagged an innocent student.

Tools that use artificial intelligence are developing rapidly. False positives are not just a possibility, but an inevitability. Consequently, Turnitin’s AI Writing Detector tool cannot be fairly used as a way to assess whether a student’s work may have been written by an AI tool. We strongly recommend that you use this functionality as a guideline, not a grading metric. The results should not be used as the sole basis for adverse actions against a student. Treat any flags the tool may raise as a reason to review the student’s work further.

Bart Everson and Elizabeth Yost Hammer led an "AI²: Artificial Intelligence and Academic Integrity" workshop recently. ICYMI, you can find the workshop recording and resources referenced in the workshop on our wiki.

Want more information?

Brightspace Tip #276: Turnitin Feedback Studio
Turnitin AI Writing Detection
Turnitin AI Writing Detection FAQs
Guide for Approaching AI-generated text in the classroom
AI Misuse Rubric
AI Misuse Checklist
Brightspace Tip #277: Turnitin - Quick Submit

View all the Brightspace training recaps
Instructors Quick Start Tutorial
Continuous Delivery release notes
Brightspace Known Issues
Request a sandbox course
Sign-up for Brightspace training sessions
You can find Brightspace help at D2L's website.
Join the Brightspace Community.
Try these Brightspace How-To documents.
Visit our Brightspace FAQs for additional Brightspace information
or schedule a one-on-one session, email, or
call Janice Florent: (504) 520-7418.

Note: Are you doing something innovative in Brightspace or perhaps you've discovered a handy tip? Share how you are using Brightspace in your teaching and learning in The Orange Room.

Image credit: image by Alexandra_Koch from Pixabay

This is a guest post from Mary A. Guillory, Remote Assistant Librarian for Xavier University Library.

It is interesting to have lived to see the day that an AI chatbot became an A-list celebrity—paparazzi and all. Its name has been smeared in the tabloids, discussed in the news, and has sparked fear in academics internationally. Students and professionals hoping to do less work rejoiced, while simultaneously coping with the frustration that comes with obtaining an account and getting the short end of the stick on a traffic-spike-plagued database website. All those things aside, the real proof of OpenAI’s ChatGPT success is that it has made it into the headlines of The Onion three times thus far. Like any assistive technology it makes learning more intriguing and allows users to more easily do things they might have struggled to complete without the tool. 

So, what does this librarian consider the best use for ChatGPT? Its ability to provide critical thinking practice, enhance coding education motivation, and act as a study buddy. Since artificial intelligence is able to infer so much from big data, it excels at helping people to think and grow intellectually. The following are my elaborations on these three ideas: 

1. Critical Thinking Practice 

Blind trust is never higher than with a computer. We enforce our own perception that the algorithm behind {insert whatever web tool name you use here} is always right on a daily basis because it is able to prove a high confidence rate to us by balancing crowd sourced human behavior data with our personal patterns and preferences. The most surprising thing about ChatGPT is that it is an AI still in training, so it gets things wrong. It operates from a dataset that hasn’t been updated since 2021 and is not connected to the internet in a way that allows it to take advantage of Google’s strides in the search engine arena. Though technology advanced in increments over the years, Google search AI has been collecting live human data to deliver better results and present them in answer form for over a decade.  

What might come of having students think critically about the answers offered by ChatGPT? The AI chatbot does not cite its sources of information, which means that to some degree it is speaking as an authority. There is power in the cognitive dissonance created when a human fact checks a podcast like “Exploring Afrofuturism with AI: A Librarian Interviews ChatGPT” and finds issues with ChatGPT’s infered answers—factual issues. The best part is that these types of exercises can be customized to student interests and created by students for students.  

2. Coding Education Motivation Enhancer 

Coding in any language whether it be for the web, mobile applications, or computer software is a valuable skill. It can even come in handy for using low-code or no-code platforms when customizations are desired. The issue usually reveals itself in the time and discipline required to develop coding skills to a useful level when students might be starting at zero and have little interest in code beyond what it can do for them today. Need some HTML or CSS to spruce up a blog post or website? ChatGPT can help. Need a basic Python program? ChatGPT can do that too. Need to work with the PowerApps language Microsoft Power Fx you’ve never heard of before? No problem, ChatGPT even speaks Klingon. The best part is that it can take students straight into the editing and trouble-shooting process of coding, which many may find more intriguing than writing the same little boring calculator or joke generator over and over again. Having to learn the hard way why the basics are important upfront to make the code work for a real-world need is way more fun than memorizing them with vague hopes of creating something someday. 

3. Study Buddy 

People get tired of answering questions and sometimes don’t feel like discussing certain topics. Throw scheduling or COVID-19 into the mix and ChatGPT might make its way to the top of the study buddy list. Students can practice discussing any topic, answering interview questions, or get instructions and tips on how to complete a desired task. It is also quite good at suggesting study resources and plans. 

Want to Learn More About Artificial Intelligence and ChatGPT? 

  • Register for CAT+FD’s hybrid “AI2: Artificial Intelligence and Academic Integrity” workshop on February 9th
  • Email me for a link to the AWS Machine Learning University’s monthly Friday webinar on February 3rd or sign up to receive emails about future sessions from Amazon. Taught by one of Amazon’s data scientists, this month’s topic (the first in the series) will focus on “Responsible AI”. Students and faculty are welcome to join these sessions.  

I know everyone reading this post is waiting on the answer to the big question so here it is—no, this CAT FooD was not prechewed by ChatGPT.