What's the point of having a human teacher anymore?

When it comes to feedback, GenAI seems able to do everything a human can do. Maybe it can even do it better than we can. Perhaps GenAI isn’t perfectly accurate yet, sure it makes mistakes sometimes, but it’s improving constantly – and we all know that our fellow humans can also give us a bum steer from time to time!

Beyond accuracy, GenAI excels in areas we know to be valuable for feedback. Students can seek help at any hour, ask unlimited questions, and even pose queries they might feel too embarrassed to ask a human teacher. Given these advantages, what's the point of keeping a human in the feedback loop at all?

In our recently published paper, which we presented at the HERDSA conference and the ACEN/HERDSA Vic snapshots in 2024, we argue that there are actually extremely important reasons to ensure that it is a human providing students their feedback.

Feedback is more than just information transfer

You might have noticed that above we presented feedback just in terms of what AI can do well, which is essentially generating informational content. But effective feedback isn't just about transmitting facts from teacher to student (even though, in the recent past, feedback was all about “knowledge of results” with occasional advice on how to do better).

Scholarship has recognised that effective feedback is fundamentally relational, it happens not merely through the transfer of information but within complex dynamics of human interaction. Effective feedback involves trust, respect, vulnerability, and mutual recognition. Feedback is not merely transactional but deeply interpersonal; it requires recognising and being recognised.

Recognition is essential to effective feedback

Recognition is a foundational condition for meaningful interaction. Drawing on philosophical insights from Axel Honneth and Robert Brandom, recognition comprises the mutual acknowledgment of each other's agency, vulnerability, and shared humanity. This mutual recognition shapes students' academic identities and self-esteem as they see their efforts and capacities validated by teachers, while simultaneously affirming teachers' professional identities through student engagement. Without this mutually recognitive relationship, feedback risks becoming a mere exchange of information, losing its transformative potential.

This relational dimension of feedback highlights a critical limitation of GenAI. While Generative AI systems excel at producing useful and scalable informational feedback, they fundamentally lack the capacity for genuine recognition. GenAI cannot genuinely acknowledge the student's effort or vulnerability because it cannot expend effort or experience vulnerability itself. It operates outside the relational framework that defines meaningful feedback interactions, and this has profound implications for the pedagogical value it can deliver.

GenAI: a source of Extra-Recognitive Feedback

We have therefore proposed a new conceptual framework designed specifically to clarify this crucial distinction. We distinguish between 'recognitive' feedback, where mutual recognition and relational trust between human interlocutors is possible, and 'extra-recognitive' feedback, which GenAI can offer effectively but without genuine recognition. Recognitive feedback strengthens a student's academic identity, contributing significantly to their personal and scholarly development. Extra-recognitive feedback, by contrast, serves well in contexts where repeated practice, immediate corrections, or technical assistance is needed, offering a valuable complementary role to traditional human feedback. We believe that understanding feedback through this recognition-based lens allows us to thoughtfully integrate GenAI into educational contexts, preserving the critical role of human interaction while also leveraging the strengths of AI.

So, what is the point of having a teacher in the age of GenAI? Our recognition-based framework provides one answer. That is, the uniquely human ability to engage in genuine recognitive relationships. Teachers do more than just deliver information. Teachers affirm students' developing identities, create trusting and respectful environments, and navigate the mutual vulnerabilities inherent in learning processes. While GenAI can support many practical aspects of feedback, only human teachers can fully embody the deeper relational aspects that transform educational experiences.

Blog contributors:

- Dr Thomas Corbin

- Dr Gene Flenady

- Associate Professor Joanna Tai

Banner image generated by DALL-E