The accelerated adoption of generative artificial intelligence (GenAI) in teaching has provided novel avenues of assisting preservice teachers to develop Technological Pedagogical Content Knowledge (TPACK), but there is still a dearth of empirical data on the TPACK performance of AI. What teachers need to know is how effective AI systems can be used as effective cognitive partners, thus it is essential to teacher education. This research article made a comparison between the objective TPACK performance of ChatGPT and human preservice teachers in a single case-control design. As a normative comparison group, ChatGPT, as a candidate GenAI peer tutor, and 93 preservice teachers in Australia were used. The performance was measured using an adapted objective TPACK scale. The analysis of the Bayesian indicated that ChatGPT was more effective when compared to the human control group with a significant effect. The qualitative analysis of the ChatGPT responses showed rather consistent and logically relating interpretations of TPACK dimensions, especially in design, exertion, and ethics. The results imply that ChatGPT can serve as a more informed other in teacher education because it can offer scalable, personalized, and conceptualized assistance to preservice teachers in TPACK development. It means that the GenAI tools may supplement the traditional mentoring and peer tutoring patterns, improve self-directed learning, and help to create more adaptable teacher-training settings. Moreover, the research points to the necessity of pedagogically motivated systems and institutional provisions to implement AI in teacher education in a responsible way. All in all, the findings are indicative of the empirical evidence pointing to the fact that GenAI systems can play a real role in the development of TPACK, which is why additional studies on the pedagogical role of such systems, their ethical aspects, and application in teacher training are so critical.
Çelik et al. (Sun,) studied this question.