�԰�����

At �԰�����, the liberal arts help us navigate a changing future with well-rounded perspectives that span diverse disciplines and viewpoints. Hear from four voices across campus about what excites, concerns, and inspires them about artificial intelligence.

Now, meet the Panelists.
��

KYLE GILLETTE ’01, PH.D.

Special Adviser to the Provost for Expression and Civil Discourse Author, longtime theater professor, and former acting dean of the humanities, Gillette is now charged with advancing civil discourse on big topics at �԰�����.

RONNI GURA SADOVSKY, J.D., PH.D.
Philosophy

A researcher into social norms, Sadovsky has taught courses on the ethics of artificial intelligence.

BRITTON HORN ’09, PH.D.
Computer Science

Horn conducts research on game design, procedural content generation, and artificial intelligence in games.

ALTHEA DELWICHE, PH.D.
Communication

Delwiche, a longtime researcher into digital media, big data, political propaganda, and virtual worlds, has recently launched a lab studying media coverage of artificial intelligence.
��

Q: HOW IS ARTIFICIAL INTELLIGENCE (AI) CHANGING YOUR FIELD?

KG: I love that the advent of AI requires faculty to rethink the point of pedagogy for writing. Why does the act of writing matter? If you answer that question like most faculty do—that learning to write is about learning to think—then you have to start interrogating more deeply the contexts and stakes of composition. Computers can generate certain kinds of generic texts quickly and

without grammatical errors. But they can’t interrogate or unpack or anticipate other minds; they can’t think associatively or metaphorically.

In that sense, AI is no more a threat to writing than the invention of photography was to painting. Photography didn’t make painting obsolete. It made it much, much more interesting. From impressionism and cubism to abstract expressionism, painters reimagined the medium and emphasized the subjective, human basis of art.

The same thing will happen to writing. Universities don’t have to fear machines that write like humans if they don’t teach humans to write like machines. As President Beasley has put it, in a world of technologically-generated answers, the people who ask the most interesting questions make all the difference.

RS: The primary change I see is a change in the relationship between philosophy and the wider world. Like many earlier technological revolutions, this one has generated a renewed demand for philosophical expertise and a recognition of its importance. Philosophers can help answer questions about when (if ever) we should think of AI systems as sentient, as holders of beliefs, or as partners in knowledge creation. Philosophers can also help practitioners in other fields think through the ethical problems raised by specific applications of artificial intelligence—including high-stakes applications in finance, criminal justice, medicine, and education. Many philosophers have been hard at work trying to answer these questions and help practitioners of other fields investigate their own questions with more clarity.

BH: AI techniques in my field have been used for decades. Even though there are numerous ways to generate content, my research area (procedural content generation for games) has seen a dramatic increase in focus on using AI for generating content such as game levels, music, and art. In teaching, I have found that students who use AI to help create homework solutions don’t fully understand what they are learning. Because of this, I have created more scenarios where they must comprehend and trace code that I produce rather than creating code from scratch on their own.

AD: For the past 80 years, communication researchers have relied on a method known as quantitative content analysis to detect systematic patterns in media content. The rise of AI is enabling something known as large language model-assisted content analysis. This semester, I’m working with an interdisciplinary team of undergraduates to analyze popular press articles about the rise of artificial intelligence.

On the teaching front, AI changes everything. We are being forced to rethink prompts and

expectations for all of our assignments and assessments. We should all be asking ourselves questions like, “Why are we asking our students to do this?” and “How can we clearly differentiate tasks that can be delegated to AI from tasks that must reflect unaided human skill and judgment?”

Q: CAN YOU TELL US SOMETHING YOU LOVE AND SOMETHING YOU HATE ABOUT AI?

BH: I love that AI has caused more people to understand how influential computer science as a whole is to higher education and our world in general. However, I hate that some people view computer science as only AI, even though there are a plethora of topics that AI does not begin to address.

AD: Generative AI tools such as GPT-4, Pi, Gemini, and Claude can function as interactive tutors when attempting to master any new field of inquiry. AI models are wonderful creative and strategic collaborators and can come up with surprisingly good suggestions about how to refine questions and methods in the context of complex research projects. I love this about AI.

I hate that so many people are approaching AI models as if they are universal answer machines. Instead of iterating and guiding the AI model to more sophisticated and nuanced outputs, many people simply accept the first answer from the model and move on (the equivalent of uncritically clicking on the first Google search result).

RS: One thing I love about AI is, as I mentioned before, the way it draws people into philosophical questioning—questions about the nature of language and cognition, about values like privacy and fairness that we hope to preserve in our evolving institutions, or about the value of creative works that AI can now mimic.

In the context of higher education, one thing I hate about AI is that it tempts our students to avoid productive frustration. For high-achieving students like ours, the experience of feeling truly stuck or hopelessly lost may be one that they never had before college.

These experiences are unpleasant, but they are also incredibly important to the learning process. Students become self-confident problem-solvers by building up a tolerance for the feeling of being stuck; they need to discover for themselves that there are strategies for finding their way or getting unstuck, so that they don’t give up when they are utterly confounded by a hard problem in the future. But with ChatGPT offering them a way to avoid that frustration, it becomes very challenging for us as teachers to create assignments that will facilitate that kind of learning.

Q: DO YOU SEE AI AFFECTING THE WAY TRINITY STUDENTS AND FACULTY CAN COLLABORATE AND DISCOVER ACROSS DISCIPLINES? WHAT DOES AI MEAN FOR THE FUTURE OF THE LIBERAL ARTS IN GENERAL?

AD: AI and the liberal arts are like peanut butter and chocolate: two great tastes that go great together. One of the most exciting elements of AI is the fact that it enlarges our understanding of how to program computers and invites liberal arts majors to take a seat at the table as institutions design, train, and deploy new AI systems. In fact, our students are uniquely well-prepared to participate in these conversations.��

RS: AI presents complex ethical and political problems that cut through our traditional disciplinary divisions. We will make no progress on those problems if we allow disciplinary boundaries to restrict our thinking. So, I am hopeful that this new technology will give us opportunities to collaborate across disciplines.

��

��