Listening Back to Thinking: Transcript-Based Reflection, AI, and the Future of Dialogic Practice

Adam Sturdee
Feb 16
4 min read

Updated: Feb 17

Later this month I will be speaking at the annual online gathering of the SOPHIA Network, the European Foundation for the Advancement of Philosophy with Children.

SOPHIA Network – European Foundation for the Advancement of Philosophy with Children: https://www.sophianetwork.eu

The session sits alongside wider international conversations about AI and enquiry, and forms part of a growing dialogue across Europe about the place of technology in Philosophy for Communities. The title of my contribution is Listening Back to Thinking. It reflects a conviction that has shaped my work over the past few years: the most powerful role for AI in dialogic education is as a reflection tool.

AI in the Circle, or After It?

Within P4C communities, the question of whether AI should participate directly in philosophical dialogue is being taken seriously. Some educators are experimenting with AI as a co-participant in enquiry. Others are rightly cautious.

The philosophical and pedagogical implications are significant. What counts as understanding? What counts as presence? What does it mean to “think together” if one voice is synthetic? These are important questions. My work has focused on a different one. What becomes possible when we listen back to thinking once the enquiry has finished?

Dialogue Leaves a Trace

Philosophical enquiry is constructed in language. It unfolds through questions, reformulations, invitations, hesitations and silence. It is shaped by the facilitator’s prompts and by the distribution of voice within the circle. It is sustained by careful listening. Yet most reflection on enquiry remains impressionistic. We remember how it felt. We recall moments that stood out. We make broad judgements about engagement or depth. A transcript changes the nature of that reflection. It slows the lesson down. It reveals who spoke and how often. It surfaces the types of questions asked. It shows how ideas were built upon, or where they quietly dissolved. In dialogic classrooms, improvement does not come from metrics. It comes from attention. Transcript-based reflection deepens that attention.

AI as Mirror

The approach I will be sharing with the SOPHIA community is grounded in transcript-based lesson analysis. It has grown out of my work as Assistant Headteacher and through the development of Starlight, an AI-supported coaching platform that uses lesson transcripts to structure reflective feedback. The principle is simple. Teachers record an ordinary lesson. The dialogue is transcribed. Patterns are surfaced. Coaching prompts are generated. Reflection remains teacher-led and non-judgemental. In this model, AI functions as a mirror rather than a participant.

The facilitator remains the intellectual and ethical presence in the room. The enquiry remains human. The technology helps reveal the linguistic patterns that might otherwise go unnoticed.

Protecting Agency in a Time of Acceleration

Internationally, schools are moving from AI experimentation to more systematic adoption. This makes governance and professional agency central concerns.

If transcript-based AI is to support dialogic practice, it must be designed carefully:

Reflection must be developmental, not evaluative.
Teachers must retain ownership of their data.
Insights must strengthen professional judgement rather than replace it.

Philosophy for Communities depends on agency. Any technological support must preserve that ethos.

One of the themes I will explore in the conference session is how transcript-based reflection can widen access to meaningful coaching without increasing surveillance or formal observation cycles. Used responsibly, it offers a way of sustaining reflective depth at scale.

Sustaining Dialogic Intent at Scale

Across Europe and beyond, many educators are working to embed dialogic teaching more deeply into their systems. The challenge is not enthusiasm. It is sustainability. Coaching capacity is finite. Observation cycles are time-intensive. Reflection often depends on goodwill rather than structure. Transcript-based reflection offers one possible bridge between philosophical intent and institutional reality. By making language visible, it enables repeated, precise reflection without adding layers of accountability. This matters not only for individual teachers, but for networks of schools seeking coherence without conformity. The international interest in these questions is growing. My own recent research on AI-supported coaching and transcript-based lesson analysis has been accepted for presentation at the BERA Teacher Education and Development Conference in 2026. The conversations at SOPHIA form part of that wider inquiry into how we strengthen dialogue-rich practice responsibly.

Listening Back as Professional Enquiry

There is something philosophically fitting about this approach. P4C asks students to interrogate assumptions, refine concepts and attend closely to language. Transcript-based reflection asks teachers to do the same with their own facilitation. It is not about judgement. It is about disciplined noticing. The question is not whether AI can think.

The question is whether it can help us listen more carefully to thinking.

As international conversations about AI and philosophy continue, I remain convinced that the most powerful applications will be those that sharpen human attention rather than replace human presence. Listening back may be one of the most important professional habits we cultivate in the years ahead. And if technology can help us sustain that habit, thoughtfully and ethically, it may prove to be less a disruption to philosophical enquiry and more an ally in protecting it.

Adam Sturdee is a senior leader and co-founder of Starlight, the UK’s teacher-first AI-powered transcript-based coaching platform for educators. His work sits at the intersection of dialogic practice, instructional leadership and responsible AI strategy for schools and trusts.

He will be presenting his research on AI-supported coaching at the BERA Teacher Education and Development Conference 2026: https://www.bera.ac.uk/conference/bera-tean-conference-2026

He is also speaking at the annual gathering of the SOPHIA Network – European Foundation for the Advancement of Philosophy with Children: https://www.sophianetwork.eu