Note to the reader: Vincent Chang was a speaker at this debate. This article contains direct quotes from the speakers, and the author has attempted to present the main contents of the night objectively. 

With the recent announcement of free Chat-GPT 5 access for all students of Oxford University, freshers took to the Oxford Union stage on 25 October to debate a claim that is now especially difficult to ignore: 

‘This House Believes That ChatGPT is Better Than Your Tutor.’

From Union President Moosa Harraj’s invitation, the opposing bench featured Professor Ken Goldberg of UC Berkeley, leading roboticist and artist whose work explores the boundaries between human creativity and automation; Dr Matthew Williams, Associate Professor in Politics at Jesus College, leading the charge as one of Oxford’s many formidable tutors; and Kaelyn Grace Apple, known online as The RedHead Academic, an educator and influencer whose commentary on university life has reached audiences far beyond academia.

As a fresher participating, it certainly felt like an intimidating line-up: three scholars who have spent their careers shaping, or warning about, the future that the motion invoked.

The freshers did not relent in their proposition: 

Opening the case for the proposition, Natalie Curtis (MBA, Saïd Business School) reframed the debate through the lens of pedagogy. Drawing on personal experience, she spoke “from a teacher’s heart,” recalling how she had “laboured countless hours preparing and specialising lessons,” imagining “their faces, their eyes, their questions, their struggles, their hearts” in order to deliver the best education possible. She described the profound satisfaction of those moments when understanding clicked, the “nearly visible light that illuminated their faces when finally the concept was understood.”

“I am emphatically imploring you to remember one thing: Using ChatGPT is not void of humanity, because on the other side of the screen, there is a heartbeat of humanity that is you.” (Natalie Curtis, 1st Proposition Speaker)

From this perspective, Natalie argued that ChatGPT should not be feared as a replacement for human engagement, but valued as a tool. She positioned the technology as a supplement to Oxford University’s transformative education: “ChatGPT is not here to replace humanity. It is here to serve it.”

In my own speech, I aimed to bolster Natalie’s case with computational theory. Tutorials, I argued, are performances of intelligence: structured acts of reasoning judged by clarity, precision and argument.I suggested that Oxford’s system prizes “how smart you appear” over “how much you understand”: if you can appear intelligent in the degree-defining finals, then you will be awarded regardless of whether that was true understanding, or simple mimicry. This kind of “imitation” is the very terrain in which artificial intelligence now excels.

“The exams of intelligence we’re assessed upon are really exams of performance.” (Vincent Chang, 2nd Proposition Speaker) 

I then reinforced the Freshers’ case with an argument from practicality: tutors are not, and cannot be, administrators or counsellors, yet their roles increasingly demand both. Students face schedules so densely layered with lectures, events and commitments that tutorials often clash with the wider cultural life of the University. We made the point that ChatGPT offers time flexibility and cognitive precision. By freeing both tutors and students from the friction of scheduling conflicts and human limitations, I argued, AI can allow the University’s educational and cultural life to flourish in ways previously constrained by time.

Closing the proposition’s case, Mia Tipp (Balliol College, English Language & Literature) expanded the Freshers’ case further, and pivoted the question toward psych

ology and the student experience. She cited the Oxford English Dictionary’s definition of “better” as “more desirable,” and argued that for a timid fresher, the non-judgemental chatbot could indeed feel preferable.

“We may take as long as we wish to formulate our answer to AI. Without the unsettling consciousness that we are taking up another’s precious time.” (Mia Tipp, 3rd Opposition Speaker) 

Mia discussed how ChatGPT could bridge the generational void between student and tutor, drawing on terms like “Brat summer, delulu and rizz”.  She concluded the proposition’s case by encouraging the audience to acknowledge a “flip-side to the magic of the tutorial system”, and invited them to agree that ChatGPT could be seen as ‘better’ if they “admit to their fear”.

The opposition bench answered in kind:

Professor Goldberg, combining an engineer’s clarity with a director’s timing, opened with an anecdote of Einstein and his chauffeur; from this, he showed how a model like ChatGPT “still doesn’t understand what it’s saying.” Predictive fluency, he warned, is not the same as comprehension.

“AI stands for Artificial Intelligence, but IA stands for Intelligence Augmentation. Oxford tutors embrace that combination: AI plus IA.” (Dr Ken Goldberg, 1st Proposition speaker) 

Dr Williams, invoking the witty authority of the Oxford tutorial, defended imperfection. “What you learn in a tutorial,” he said, “is not how to get the answer right, but how to survive being wrong.” In a further interactive episode with the audience, Dr Williams requested, “if you could please look into the eyes of the person sitting next to you what you are going to reminisce about are those eyes that you just looked into: the people. Because this is the best university in the world – I believe it is – it’s the best because of you.” (Dr Matthew Willaims, 2nd Opposition Speaker) 

Kaelyn Grace Apple closed the opposition’s case from the angle of innovation. She referred to the fundamental limitations of training models such as ChatGPT: 

novel human data is necessary to train them, but reliance on the same generative AI results in decreasing human novelty overall. 

“By buying into ChatGPT and our schools, we cripple the next generation of innovators, churning out an assembly line of compliant workers who have been taught to rely on technology.” (Kaelyn Grace Apple, 3rd Opposition Speaker) 

When the vote was called, the verdict was decisive: 17 in favour, 52 against. The Union had ruled that ChatGPT is not better than your tutor. 

The debate’s news value lay less in the outcome than in its precise timing. It treated the relationship between AI and pedagogy with a seriousness rare for an opening night, reflecting the quiet shift already underway in Oxford’s academic ecosystem. Tutors are now reading essays that cite directly from, or are subtly informed by machine reasoning; students are navigating the etiquette of using it. The motion’s absurdity concealed its relevance.

The evening closed without laughter and lively discussion. For the freshers, the loss was expected, but gracefully taken. For the Union, it was a reminder that the great Oxford tradition of debate remains capable of adapting to the century it inhabits. What emerged overall was a portrait of an institution in motion: a vision of Oxford defined by its human tutorship, yet already inhabited by the tools that may one day test its limits.