I explored using ChatGPT’s new voice functionality to interact with students during my education futures class on Friday. I was surprised at how effectively it was able to converse with the class, particularly given it was placed on my iPhone at the front of the room. The students took turns asking it questions and it responded coherently and helpfully to their prompts. It was easy to imagine a slightly more advanced version of the system, running on the room’s AV setup rather than my phone, effectively coordinating a session. At present it wouldn’t be able to deal with multiple overlapping voices, requiring a discipline of turn-taking easy to enforce when presented as an experiment but probably harder to ensure in a real world teaching situation.
This presents a bleak scenario in which the academic staff are reduced to facilitators of interactions with the conversational agent e.g ensuring student interaction proceeds in a way that remains within its capacity to process. In fact in what sense would this residual human be ‘academic’? I find it hard to imagine the complete removal of human experts, at least in research-intensive universities, but it’s easy to conceive of how this might creep in around the edges of a precaratised academy. Will we have academic-led lectures, PGR-led seminars and AI-led tutorials for example? I’m torn because I can see the genuine possibility this could be an amazing tool for facilitating self-led learning in groups, but the possibility of how this might be institutionalised fills me with horror.
Mark Carrigan 