New Year, New Beginnings and Old Thinking on the Role of Scientific Intuition in the Age of AI

As we welcome the new year, we do so at a time when artificial intelligence (AI) is rapidly reshaping our lives, including STEM education. AI can solve problems, generate explanations, and even design models in seconds—but it cannot replace the scientific intuition, curiosity, and deep conceptual understanding that come only from learning physics thoughtfully and rigorously. This is the reason why I think knowledgeable and inspiring physics teachers will be even more important now. We must work even harder to help students develop scientific intuition; otherwise, they risk surrendering their thinking to AI, which will do the reasoning for them. Therefore, as a physics teacher educator, I will do my best to support future physics teachers in acquiring the Pedagogical and Content Knowledge needed to become such educators, while learning about technologies that can support student learning, including AI.

While I use AI daily, I am fully aware of many challenges it might bring to education in general and STEM education in particular. I have experienced the challenges brought by AI first hand while teaching a STEM education course this summer. While preparing my lecture notes, I asked ChatGPT draw solutions to the following question:

Figure 1: Lit a light bulb activity (illustrated by ChatGPT-4)

This is a standard question proposed by Lillian McDermott and her team decades ago. I have been using it ever since. So this is what ChatGPT suggested to me as a solution to this problem:

Figure 2: ChatGPT-4’s renditions of electric circuits consisting of a lightbulb, a wire and a battery. None of these “realistically looking” renderings make scientific sense.

In my view, in this new era, fundamental knowledge matters more than ever. Students (and teachers) must not only know the formulas or problem solving algorithms but also understand the principles behind them—why they work, when they apply, and what their limitations are. Only then can they use AI wisely, critically, and creatively. I think it is deeply misleading to believe that AI will solve all the problems in education. It will help solve some problems, but it will also create many new problems, including the problem of assessment of student learning.

As physics educators, our mission is clear: to ensure that the next generation builds the intuition and knowledge base that will empower them to question, to innovate, and to imagine beyond what any algorithm can provide.

Here’s to a year of teaching that strengthens both the minds and the scientific spirit of our students.