One of the main challenges in initial teacher education in a low-resource context like North Macedonia is limited access to good quality school-based mentors to support pre-service teachers (PSTs) in their multiple mentor roles: educator, support, acculturator, model and sponsor (Malderez and Bodoczky, 1999). Indeed, overly directive mentoring practices, also known as ‘judgementoring’ (Hobson and Malderez, 2013), are an international phenomenon, too (Hobson, 2016). In our study, we explore the extent to which the Noticing application can complement human mentors in providing good quality mentoring experiences to support PST reflection. We formulated the following research questions: 1. Did using Noticing support the PSTs’ reflective practice? If so, in what ways? 2. How did the two mentoring experiences (facilitated by Noticing and by human mentors) compare? Was there a preference for one type of mentoring over the other? 3. How can, in practical terms, AI tools like Noticing be integrated into initial teacher education programmes? Data was collected over two semesters, using a sequential exploratory approach, based on Dörnyei‘s QUAL > quan design (2007: 171). We first interviewed four PSTs who appeared to have used Noticing in different ways. On the basis of that data, we put together a mixed-method survey to poll the full cohort of 20 PSTs. We also looked at some of the PSTs’ transcripts of their interactions in Noticing. We inductively open-coded all qualitative data. To make sense of the quantitative data, we used descriptive statistics and paired t-tests. The findings suggest agreement that having mentoring conversations in Noticing was a positive learning experience which stimulated the PSTs to critically examine their teaching and plan forward in a safe, judgement-free environment. There was a feeling that through this, Noticing helped them support their learners’ learning. Opinions were more varied, however, when the PSTs compared the mentoring experiences in Noticing and those with their human mentors. This result suggests complementary roles of the two mentoring formats, rather than superiority of one format over the other. The implications for initial teacher education, therefore, are for the two formats to be used in parallel, with busy mentors delegating some of the potentially lengthy and time-consuming conversations to Noticing in order to focus on their other mentor roles, all the while monitoring the PSTs’ development and picking up from where Noticing left. In this way, human intelligence truly interacts with AI, resulting in hybrid intelligence (Akata et al., 2020).

Akata, Z., Balliet, D., De Rijke, M., Dignum, F., Dignum, V., Eiben, G., et al. (2020). A research agenda for hybrid intelligence: Augmenting human intellect with collaborative, adaptive, responsible, and explainable artificial intelligence. Computer, 53(8), 18–28. Dörnyei, Z. (2007). Research Methods in Applied Linguistics. Oxford: Oxford University Press. Hobson, A.J. (2016), “Judgementoring and how to avert it: Introducing ONSIDE Mentoring for beginning teachers” International Journal of Mentoring and Coaching in Education, Vol. 5 No. 2, pp. 87-110. Malderez, A., & Bodoczky, C. (1999). Mentor courses: A resource book for trainers. Cambridge: Cambridge University Press.