Using Generative AI to View Questions Through Different Academic Disciplines – Communications of the ACM

Our last post introduced a novel interface with a facial expression drawn on the sides of a cube made of a sponge, to guide the tone of a Generative AI response. Showing a face to the camera on a phone or tablet prompted Generative AI’s response to come from a curious, thoughtful, skeptical, or confident point of view. Besides getting people to see how tone affects responses, the unusual interface might help people remember the experience well.

Tone of response is not the only way we might want to think about aspects of a question. The Epistemic Insight approach gets students to consider the questions from the point of view of different disciplines, like physics or philosophy. It is used in classes to help students to explore ways that different disciplines can work together to address Big Questions that are impactful and universal — and that stretch across several subjects. Looking at a question from a variety of disciplines helps students to discover more about the question. Looking at how different subject-disciplines address the same question can also make learning about each discipline more meaningful. 

The Epistemic Insight ‘Discipline Wheel’ was initially a graphic that could be printed onto postcards and posters. It includes 10 disciplines and a space in the middle of the wheel for a question. In some schools, students chose a series of disciplines to learn how each one addresses a Big Question. After they have talked about it from one perspective, say history, they discuss it from another, say theology, and then possibly physics. These lessons help students bridge siloed ways of thinking and prompt discussion of how different disciplines might work together. The design is also available as a template so that teachers can change the disciplines that appear on the wheel. A limitation with the paper version is that most secondary school teachers only know their own discipline well and it can take a long time for them to build up a resource that addresses the others. Even so, students and teachers report the use of the Epistemic Insight Discipline Wheel as transformative.

The arrival of GenAI opened the door for a new interactive version of the wheel. In the GenAI-powered version, students can work individually or in pairs at a computer and can freely choose which disciplines they want to explore. In a current prototype, they are limited to choosing between questions that the teacher has chosen to include. Teachers using the wheel can compose their own questions and use the wheel in a computer lab where students explore it for themselves, and/or to help them to plan a cross-curricular component to a lesson.  

Credit: Berry Billingsley

The use of the Discipline Wheel as an interface means that users are clear about what they are doing (choosing disciplines) and why they are doing it (to compare them and see how they work together). But anyone with access to Copilot, Gemini, or another large language model (LLM) can get a feel for how it works by writing a prompt that asks GenAI to give a range of disciplinary perspectives on a suitable Big Question. Here is an example:

Describe in a couple sentences each what it means to be alive from a historical, mathematical, and theological point of view.

ChatGPT said:

1. Historical Perspective: From a historical point of view, being alive is understood as the experience of human existence within the context of time, culture, and events. Life is shaped by societal development, technological advancements, and human interactions that influence how individuals and communities survive, evolve, and contribute to history.
2. Mathematical Perspective: Mathematically, life can be seen as a complex system governed by biological processes and patterns that can be quantified and analyzed. Equations and models describe growth, decay, probabilities of survival, and population dynamics, offering a structured understanding of life’s unpredictability and interrelated factors.
3. Theological Perspective: Theologically, being alive is often considered a gift from a higher power, where life is imbued with purpose and meaning beyond the physical world. It is viewed as a spiritual journey, with a focus on fulfilling a divine plan, moral development, and ultimately, the soul’s connection to the afterlife.

While it is obvious that the responses concern the same question, it is fun to see the way generative AI focuses on different domain tenets.  Reviewing an answer from a history, math, and theological viewpoint teaches us something about the question “What it means to be alive,” but also something about the fields of math engineering and theology. We see in the answers that the different disciplines use different vocabularies: for example, history talks of societal development and advancement, mathematics talks of complex systems and equations, and theology talks of purpose and meaning. Generative AI lets us freely choose a few domains to see different ways of analyzing a question. It opens up a large space for comparing points of view, learning about academic domains, and also learning about the relationships between them. 

We can use generative AI not just as a way to get an answer to a question, but also to gain epistemic insight into the ways that complex questions can be interpreted and investigated differently by different disciplines. As such, the aim of this blog is to continue our effort to show that a scenario to access knowledge can help us broaden our understanding and even compare different ways of understanding the world.

Like the faces on the sponge, the Discipline Wheel is designed to get people to use Generative AI to look at problems from different viewpoints. These tools inform prompts that help the Generative AI, but also our minds, see perspectives that can improve our understanding of where answers can come from. 

Generative AI instantly produces content for the Epistemic Insight Discipline Wheel. The answers from the Wheel are not intended to be the end of the inquiry, but rather a tool to encourage students — and teachers — to delve more deeply into how disciplines work.

This and our previous post have touched on ways the new AI can offer completely new ways to help you look at questions and answers to compare different points of view. In engineering or business experiences, we have many reasons for looking at things from different points of view as well. We look at problems from different points of view to look for simpler solutions, we do it to find bugs in our designs, and to flesh out the different parts of them. While there are many other ways to show how people might use Generative AI to view things from different points of view, our next post on perspectives will discuss the various persona and roles one might bring to solving a problem in creating a business technology solution. 

Ted Selker is a computer scientist and student of interfaces.

Berry Billingsley is an educator interested in philosophical “big questions.”