In his 2008 book Physics for Future Presidents, Prof. Richard Muller emphasizes that basic physics can and should be taught to students of varied backgrounds such that they can have a conceptual understanding of topics of universal interest such as solar energy without overly relying on the underlying mathematics. The underlying philosophy of the book is that we can be better citizens and leaders if we have a clearer understanding of how the physical world around us actually functions. However, the world has changed quite dramatically in the intervening sixteen years: John Hopfield and Geoffrey Hinton-pioneers in the field of Artificial Intelligence - have been awarded the 2024 Nobel Prize in Physics. In this article, we will see how AI is the new physics and that non-science students will benefit greatly from developing a conceptual understanding of how AI tools work rather than using them as a black box.

AI is clearly a driving force today in the science, technology, engineering, and medicine (STEM) fields. Technological developments ranging from driverless cars to pharmaceutical drug discovery have been accelerated due to recent developments in AI. However, AI is not just transforming technology sectors; it is reshaping diverse non-STEM fields such as art, law, journalism, and education. Artists employ AI algorithms to generate innovative pieces, blending human creativity with machine precision. Legal professionals use AI for predictive analysis and help with case evaluations and legal research. Even marketing, media, and communication professionals may rely on AI for data analytics and even content generation, blurring traditional notions of human creativity and authorship. This article itself - perhaps naturally, given the subject - has benefited from brainstorming ideas with modern large language models.

My main contention is that students will find it easier to be critical and judicious in their use of AI tools if they are broadly aware of how these tools work. They would be less likely to treat AI-generated output as the absolute truth if they are trained to think of AI models as software programs extracting and generating information from a large but finite set of training data than as a black box, which is an omniscient source of infinite knowledge.

To ensure that students have such an outlook, we must ensure that they have access to training that equips them with a conceptual understanding of how modern AI models work. To be sure, there is no dearth of information and training available on AI technology in this day and age. However, a vast majority of such training focuses on the mathematical and coding aspects of machine learning and AI. This quantitative and technical focus makes these resources inaccessible to students from non-STEM backgrounds.

It is my strong belief that academia should also develop resources that will allow students from non-STEM backgrounds to appreciate the conceptual underpinnings of modern AI technology, such as LLMs, and prepare them to engage with these tools in a way that takes advantage of the opportunities that this technology provides while equipping them to recognize and alleviate the pitfalls that accompany their use.

It is in this context we recently piloted a course to empower students from all disciplines, including non-STEM students, to have a conceptual understanding of how modern AI tools work. The course design deliberately breaks away from conventional AI education models, adopting a "no-code, no-math" approach that makes complex concepts accessible to students from diverse academic backgrounds. Rather than focusing on technical complexities, students engage hands-on with tools like ChatGPT while exploring their implications across various fields—from literature and the arts to business and economics. This course received a strong response from our students, and we hope that with further refinements, it will contribute to our students being able to use AI in a responsible and healthy fashion.

My interactions with students in this course have drawn my attention to two contradictory emotions that non-STEM students feel about AI: they are excited about how developments in AI can add new frontiers to their field, but they are equally worried about the negative impact that AI can have on the job market in diverse fields. We hope to provide these students with resources that can help them derive the greatest benefits that AI can provide while alleviating the risks to their future by making sure that they are empowered to deal with AI as an ally rather than as a competitor. I believe that this is only possible when students are equipped with the knowledge and experience to prompt AI tools with precision, evaluate their output with healthy skepticism, and build upon and extend AI capabilities with creativity and originality.

Author: Prof. Kaushik Gopalan, Faculty of Computer Science & Director of the Centre for Interdisciplinary Artificial Intelligence, FLAME University.