Low-and middle-income countries (LMICs) spend millions annually on teacher training. Despite these investments, training programs often fail to equip teachers with the content knowledge, pedagogical skills, and preparation needed to appropriately foster students' foundational, socioemotional, and critical thinking skills. These weaknesses in training have led to an unfortunate reality in which many education systems fall short in adequately preparing and supporting teachers (see example from Sub-Saharan Africa), resulting in a teaching force that lacks the content and pedagogical knowledge needed to improve student learning outcomes.
Can AI help address these challenges?
Given these realities and the quickly changing technological environment, policymakers can hardly keep the same teacher training model. AI tools, such as ChatGPT, highlight a growing need for teachers to equip themselves with adequate digital competencies to effectively utilize and instruct with AI (e.g., see an example of a framework for AI literacy or AI-ready educators).
Although this evidence base is limited, much has been produced regarding how students engage with technology and the conditions for effective human-technology interaction (see here, here, and here). With this reality in mind, we sought to understand where, when, and how AI can be used as a tool to help current (and future) teachers based on the evidence currently available:
- First and foremost, teaching – and by extension teacher training – is a profession that relies heavily on human interactions. Available evidence from the pandemic and beyond indicates that the presence of an in-person trainer or coach is essential to building a strong relationship that facilitates behavior change. Given the importance of teacher-student interactions, it’s unlikely AI will replace teachers. However, utilizing AI as a tool to support teachers requires a new relationship between how teachers and students interact with technology. Thus, as these technologies continue to advance and become more complex, the nature of teacher training and how teachers interact with students will also need to evolve.
- Teachers should be trained on how to use AI, and it must be tested extensively before rollout. If AI were to be used to train teachers, it would require a lot of testing and planning, especially in low-resource settings – this type of intervention should not be rolled out at scale from the beginning of the project but should be tested and adjusted based on the teachers' experience. More work is needed to clarify exactly how to train teachers so that they cultivate the digital competencies required to use AI effectively. What we know now is that significant differences exist in teachers’ perception of technology that affect how open they are to learning and adopting new digital competencies.
- Given the quickly evolving technological landscape, policymakers must prepare "future-proof" teachers. With any training offered today, there is a risk that the investments in specific digital skills will be outdated in five years. New interactions between human beings and smart machines raise new fundamental questions that cannot yet be answered comprehensively by the education and professional development of teachers. Therefore, it may be useful to better understand and further investigate the concept of augmented work and augmentation strategies for the teaching profession.
- Currently, AI can enable teachers to spend less time on routine tasks and more on quality interactions. Suggestive evidence from other sectors has shown that access to AI increased worker productivity by 14% and greatly impacted novice and low-skilled workers. These lessons could be extrapolated to the education sector as well. Assuming teachers have received the appropriate training, AI could be used as a tool to help teachers complete repetitive or non-cognitive tasks. Using these tools in this manner could save teachers time on administrative tasks, reduce their workload, improve their information literacy, and ultimately allow them to focus more on high-quality instruction and relationship-building with students.
- AI has the potential to change what and how teachers teach, which will directly affect the training teachers’ receive. Currently, teachers are trained to focus on utilizing technology to enhance students' grasp of traditional content. However, with the integration of AI, the emphasis could shift towards ensuring teachers have the skills to improve content comprehension and, more importantly, acquire knowledge relevant to today's world. Assuming their students have the necessary foundational literacy and numeracy skills, this transformation may allow teachers to allocate more time to focus on higher-order skills such as critical thinking, collaboration, and problem-solving. This new context presents a dilemma for new teachers, who are expected to "cover" vast amounts of content. Hence, teacher training programs should equip teachers to thoughtfully navigate the balance between content coverage and promoting deeper learning.
The bottom line
Many unknowns remain; policymakers must proceed cautiously given the limited evidence base. For instance, there's no guarantee that AI wouldn't automate and replicate bad ways of teaching, as it still cannot decipher bad vs. good teaching practices. Moreover, there are unforeseen issues of equity, ethics, privacy, and safety. Last, and most importantly, just like any other educational technology (a device, a platform, or software), AI is only a tool. It does not guarantee the success or failure of any intervention.
Ultimately, the usefulness of AI depends on the quality of the intervention, the readiness of the technology, the user experience, perceptions, incentives, and the level of human support. Better evidence is needed. We welcome further research in this field, focusing on teacher training in low-resource settings.
Can we trust AI to improve teacher training? Let us know what you think in the comments.
To receive weekly articles, sign-up here
Join the Conversation