April 24, 2023 - Editor’s Note: There are two versions of this story. The first was written by an Inside Marist staff writer; the second was written by ChatGPT as an experiment to compare AI-generated content to human-generated content. 

Story Version #1
By Michelle Eggink

Data science and analytics student Christian Sarmiento '25 teamed up with Dr. Amanda Damiano, Director of the Center for Social Media and Assistant Professor of Communication, to explore what the Marist community thinks about ChatGPT and its possible impact in the classroom.

The artificial intelligence chatbot developed by Open AI has been leading conversations across news platforms since its launch. In response to the growing ChatGPT buzz, Sarmiento surveyed Marist students, faculty, and staff to learn more about how people feel about its impact on academia. 

The survey revealed mixed feelings about ChatGPT. While many of those surveyed stated their apprehension about the use and output of ChatGPT, most trusted or didn’t question the output of ChatGPT’s responses when the AI-generated content was presented to them, even when some of ChatGPT’s responses were incorrect. 

The survey also noted some interesting findings about the assumptions and unknown implications of ChatGPT’s use in academia. For example, when respondents were asked if students would use ChatGPT to plagiarize, most people said yes. But when they were asked, “Would you personally use ChatGPT to plagiarize,” the majority said no. The researchers hypothesized that this disconnect could be due to a social desirability bias, which is when people taking a survey answer in a way that they think will be viewed favorably by others. 

Overall, the research points to a need for schools to address the broader issue of AI alignment and ethics and to consider institutional guidelines for ChatGPT that can be created in the context of curriculum.

“Schools should assess their goals so they can outline where and how they want to implement these types of technologies,” said Sarmiento. “In addition to creating some sort of regulation and awareness plan, schools should be encouraging critical thinking and how to ethically and efficiently use software like ChatGPT as a tool, not a replacement for work or original thought.” 

Sarmiento has been interested in the power of computers since he was young, building his own computer when he was in the 7th grade. An aspiring data scientist and machine learning engineer, he has been working to gain as much research experience during his undergraduate years as possible, tackling big topics like ChatGPT.

“It’s pretty exciting to look at ChatGPT and where it’s going,” said Sarmiento “ChatGPT is a progressing technology that can be incredibly helpful, but in terms of implications, there’s a lot of unknown. ChatGPT isn’t going away anytime soon, so it’s good to get prepared by increasing awareness about it and improving data literacy so we can effectively utilize it and interpret it.”

Christian Sarmiento and Dr. Damiano in front of ChatGPT research findings. Photo by Alexandra Rossi '24.

Sarmiento worked alongside Dr. Damiano to analyze relevant research, conduct interviews with faculty members from both the School of Computer Science & Math and the School of Communication and the Arts, and survey the Marist community across all schools and age ranges. The survey focused on exploring how the accessibility of a large-language software like ChatGPT affects learning for both students and educators and how the community feels about the technology being available for public use.

“It’s been an ultimate pleasure working with Christian with his tireless energy, enthusiasm, and interest in undergraduate research,” said Dr. Damiano. “While Christian is new to Marist, he hit the ground running, making so many connections from doing this project and helping forge ongoing collaborations between the School of Communication and the Arts and the School of Computer Science and Mathematics.” 

As the current President of the Marist Analytics Community, Sarmiento also recently organized an on-campus AI & Language Models panel of educators and field specialists. The event was moderated by Eitel Lauría, Professor of Data Science & Information Systems and Director of Graduate Programs, and took on topics like AI alignment, AI ethics, and the larger implications that come with the development of these models.

Marist Analytics Community’s AI and Language Models Panel. Photo by Michelle Eggink.

"The foundation models as GPT-4 and applications like ChatGPT are a spectacular feat of science, technology and engineering that do not cease to amaze,” said Lauría. “While they bring about unprecedented progress and revolutionize countless fields — in that regard the research work by Sarmiento and Damiano is very relevant — we cannot ignore their dark side, from biases and disinformation to the potential stifling of human creativity. As humanity inches up toward Artificial General Intelligence, we must grapple with the existential risks and ethical complexities that arise from creating superintelligent systems.”

Lauría worked with Sarmiento and Damiano during their research, helping them choose software for sentiment analysis and think about how the use of ChatGPT aligns with individual goals of specific users (direct alignment) as well as the goals of broader society (social alignment).

Sarmiento presented his findings at Marist’s annual Celebration of Undergraduate Research, Scholarship, and Creative Activity (CURSCA) on April 19.

 Sarmiento presenting his research at CURSCA. Photo by Zoe Fernandes '25.

The Marist Institute for Public Opinion has also conducted national research about what Americans think about ChatGPT. You can read the survey results here.

Story Version #2
By ChatGPT 

ChatGPT has a broad range of capabilities. We asked it to write a news story using just this prompt: “Write me a story about a student at Marist College named Christian Sarmiento doing a study on ChatGPT.”

Christian Sarmiento was a curious and dedicated student at Marist College, with a passion for exploring the latest advances in artificial intelligence. He had heard about ChatGPT, a powerful language model created by OpenAI, and was eager to learn more about its capabilities and potential.

With the support of his professors and fellow students, Christian embarked on a comprehensive study of ChatGPT, analyzing its algorithms, data inputs, and output patterns. He spent countless hours poring over technical papers and conducting experiments, trying to understand how the system worked and how it could be improved.

As he delved deeper into the world of ChatGPT, Christian began to realize just how powerful this technology could be. He saw the potential for ChatGPT to transform countless industries, from healthcare and finance to education and entertainment. And he knew that he wanted to be a part of this revolution.

With his research complete, Christian presented his findings to a group of industry experts and fellow students at Marist College. He demonstrated how ChatGPT could be used to generate natural language responses to a wide range of questions and prompts, and showcased the system's ability to adapt to new data and contexts.

His presentation was met with enthusiastic applause and praise. He knew that he had made an important contribution to the field of artificial intelligence, and that his work would have a lasting impact on the world.

As he packed up his notes and left the conference room, Christian couldn't help but smile to himself. He knew that his journey with ChatGPT was far from over, and that there were countless new discoveries and challenges waiting for him in the years ahead. But for now, he was content to bask in the glow of a job well done, and to revel in the limitless potential of this remarkable technology.