Artificial intelligence and other novel technologies in physics education

We are living in a time of significant technological advances and breakthroughs. Increasing computational power facilitates extensive numerical simulations of material physics, astronomical objects such as black holes, weather and climate phenomena, and the dynamics of a global pandemic that we have been going through for a couple of years. Artificial intelligence based chatbots such as ChatGPT by OpenAI appear to have a disruptive impact on science, technology, engineering and mathematics (STEM). Data science and big data are more established concepts that have already found their way into academic research as well as industrial applications. Quantum computing and brain-inspired computer chips are further examples of potentially disruptive technologies. Virtual reality and the so-called metaverse are expected to become increasingly sophisticated and relevant, too. 

All of the aforementioned ideas and concepts are highly relevant not only for research institutes and tech companies, but even more so in education. In particular, physics education research (PER) can also benefit from these developments, for instance by applying methods of the data science and artificial intelligence domains (e.g., artificial neural networks) to data sets and survey results that may then help to make physics education more effective, equitable and enjoyable. Furthermore, it has already become evident that methods based on artificial intelligence will revolutionize the way how physics is taught and learned. For example, students will certainly use clever chat prompts to unleash the potential of systems like ChatGPT and solve homework problems, programming tasks and technical questions. While many educators have recently begun sharing their concerns, we should also start the conversation about the benefits, transformations and potential that come with such technologies. Lastly, PER scientists also need to think about how university curricula have to be updated and even drastically reformed such that students are able to acquire the skills and expertise that they need to tackle the major challenges in the remainder of the 21st century and beyond. 

Certainly, as pioneering studies have indicated (see for example a preprint by Yeadon et al.), some well-established formats such as physics essays may be threatened by the artificial intelligence revolution. It has been shown that essays by ChatGPT achieved similar grades as those written by students, while maintaining a low plagiarism score. Therefore, the study by Yeadon et al. rightfully claims that short physics essays may not be suitable as an assessment method in the near future. However, instead of fearing the considerable changes that are coming, we should try to embrace the new technologies as we had done in the past, too. For instance, calculators, computers, smartphones and the internet were also regarded as dubious technologies that would prevent students from learning the subject matter in a thorough and effective way. Instead, it turned out that these technologies serve as tools for both students and educators in order to improve learning outcome. I am certain that this will also hold true for artificial intelligence and virtual reality based technology.