Ask an Expert: The Intersection of Liberal Arts and Artificial Intelligence
CLA faculty share how they work with and research AI
From smart assistants, like Siri and Alexa, to chatbots that write essays or complex computer code, new technologies look straight out of science fiction. Opinions on these range from alarmist to optimistic. We sat down with faculty across the college for their take on artificial intelligence (AI) and how they work with or research it within different academic disciplines.
Brandiff Caron
Lecturer, Interdisciplinary Studies in the Liberal Arts
Area of Expertise: Democracy Theory and Technology Design
My primary field of study is science and technology studies (STS). For the last 10 years, I have been working with engineers and computer scientists to develop creative ways to incorporate considerations of justice, equity, diversity and inclusion into their design processes, including artificial intelligence. Recently, I have shifted my focus from engineering design to thinking more about the role of non-engineers and non-computer scientists in the larger conversation about these technologies.
The Interdisciplinary Studies in the Liberal Arts (ISLA) Department at Cal Poly offers a number of unique opportunities to teach students from across the university about these technologies. And it is in ISLA 303: Values and Technology where my new pedagogical experiment is unfolding.
One of the most challenging aspects of thinking about ethical technology design is the difficulty in knowing what a technology is going to do prior to it becoming widely used. And once a technology has become widely adopted, it is famously difficult to “put the toothpaste back in the tube,” so to speak.
In ISLA 303, we have taken deep dives into the many social challenges created through the use of AI. Students are asked to engage in the creation of “useful fictions” and speculate as to what potential negative and positive effects on society there may be.
With these “useful fictions” in hand, students are able to go back and examine the AI technologies as they are today and can point to specific design choices that could be made in ways that address some of these consequences.
Students have enjoyed being tasked with this creative project alongside the sociology, philosophy, political science and history that make up the rest of the course. I am excited to be developing plans for a short story collection of Cal Poly students thinking about the future of AI through these useful fictions.
Deb Donig
Assistant Professor, English
Area of Expertise: Anglophone Literature
I’m the founder of the Ethical Technology Initiative at Cal Poly. I teach a course on ethics and technology, and I also host the “Technically Human” podcast — a global podcast about ethics and technology. Across those three things, I spend a lot of time thinking about AI.
As an English professor, I always have to explain to people why I’m interested in ethics and technology to begin with. And what I say is that before we can build anything, we first have to imagine it. What do we want this technology to do?
The majority of problems that we really need to solve in human society are not problems of how airlines can use AI to better respond to customer service complaints — that’s not a human society problem. Human society problems are how do we create better public housing for people in dense urban areas where the rents are more than two thirds of somebody’s income? How do we eliminate or reduce bias in medicine?
These are things that AI could be solving right now. The problem is that these issues are incredibly valuable on a social level and less valuable on an economic level. And where AI is developing the quickest is not where there are vital problems that need to be solved, but rather where there is financial opportunity.
My hope is that we can spread the word and demand some sort of reform to the economic and profit structures undergirding AI production and innovation, and that we can redirect funding and mandate legislative requirements that commit or incentivize AI to solving real-world problems.
Be part of the conversation. Check out Donig’s podcast “Technically Human” and learn more about ethical technology from experts in a multitude of fields.
Jason Peters
Associate Professor, English
Area of Expertise: Composition and Rhetoric
I coordinate the first-year English composition program, and when I heard about ChatGPT’s release last winter, I really found myself scrambling to get access to it so that I could provide support to the dozens of faculty members who teach college writing courses.
I’m not an expert in writing with digital technologies like other folks in my field are, so I’m not working with ChatGPT in any research capacity, at least not yet. My role has largely consisted of putting ChatGPT into a larger context, encouraging teachers and students to think about it beyond just the concern about whether using it constitutes plagiarism, and inviting us to think in more expansive terms about what the technology means for us and how to best make use of it in the context of writing instruction.
Additionally, I would really like to ease anxiety around this topic and encourage people to remember that writing is a technology. You can’t do it without tools. When I was in my first year of college, less than 30 years ago, I wrote my first college essay on an electric typewriter and corrected typos with Wite-Out. By the time I was writing my senior honors project four years later, I was using an OCR text scanner to scan my old typewritten drafts into Microsoft Word in Windows 98 and copying and pasting digitized text into my senior project. I remember copy and paste being controversial because students didn’t need to type out every word of their essays by hand.
Better writing-assistive technologies enable us to devote less labor and intellectual energy to the writing process and more of that labor and energy to the rhetorical purposes underlying our writing. ChatGPT is a new development in a long line of writing-assistive technologies that, at one time or another, seemed disruptive and destabilizing to a status quo, but which we have since learned to integrate into our writing processes.
Learn more from Peters about ChatGPT’s writing limitations and teaching potential at bit.ly/4505Bv3.
G. Andrew Fricker
Assistant Professor, Anthropology and Geography
Area of Expertise: Geography and Spatial Ecology
As someone who works with a lot of earth observation imagery, AI represents a major advance in extracting valuable environmental information from satellites, aerial imagery and unmanned aerial vehicle imagery. New deep learning techniques, AI methods that teach computers to process information in a way inspired by the human brain, are as effective if not more effective than human analysts. Deep learning, in particular, is a quantum leap in terms of the precision and accuracy of geospatial imagery prediction and classification compared to traditional methods.
To study this, computer science Professor Jonathan Ventura and I started a research lab that focuses on using deep learning to solve geospatial environmental problems. These projects give students opportunities to work on interdisciplinary teams and with faculty geographers, biologists and computer scientists on applied research. The lab is currently working on three major projects using deep learning to:
- Quantify eelgrass cover in the Morro Bay estuary and seaweed aquaculture in Southeast Asia.
- Count trees and quantify tree cover in urban forests in California as part of a Cal Fire Grant.
- Identify palm and cacao agroforestry as part of a NASA/Jet Propulsion Laboratory grant.
These tools seem extremely complicated to use, but that’s changing rapidly. I’ve taught numerous students how to use deep learning models on geospatial remote sensing imagery, with no coding experience whatsoever. For students who are trying to learn to code, AI tools like ChatGPT are a really helpful resource in learning to write code. It’s an exciting time to be in the AI and remote sensing/spatial ecology field, and it’s only getting easier and more accessible for the average user.
Download a a pdf of IMPACT magazine.