AI's unforeseen upset: my journey of defeat at the Oxford Union

5 minute read

In a landmark debate at the famed Oxford Union, postgraduate students from the esteemed Diploma in AI for Business cohort took the stage, armed with an unconventional ally - ChatGPT, a powerful AI language model. The topic of contention was, ‘The House believes employees should be able to work from wherever they feel most productive'. In December 2021, this same cohort successfully debated for the first time in Oxford Union with the use of the Generative AI Megatron Transformer. So, I thought it appropriate to harness the capabilities of generative AI to craft my debate outline and prepare for potential counterarguments from the opposition.

Despite my best efforts, I found myself on the losing side of the debate, demonstrating that even with AI assistance, the complexities of ethical discussions remain far from black and white and cannot contend for human flaws.

It's important to say at this point that the debate topic and positioning (proposition or opposition) was chosen for each of the participants. Thus, the House argument for/against the topic is not necessarily the personal belief of the debater.

The AI-powered preparation began only 18 hours before the event, when all students were provided the assignment for the big debate. I recognized the potential of leveraging ChatGPT's vast knowledge base, and together we meticulously constructed a persuasive argument. ChatGPT, drawing from a vast repository of information, helped me build a compelling case for the positive impact, reinforcing the proposition's core message: by embracing a flexible work model, businesses can empower their employees, improve work-life balance, attract diverse talent, foster innovation, and adapt to the modern world while addressing potential concerns through thoughtful approaches and technological advancements. ChatGPT and I argued that this will result in a more efficient, inclusive and motivated workforce, benefiting both employees and organizations alike. From revolutionizing healthcare with personalized treatments to optimizing logistics and transportation networks for a greener future, my prepared speech showcased AI as a transformative force.

Additionally, ChatGPT provided real-world examples of how AI could enhance decision-making, streamline processes and augment human capabilities. Saïd Business School Lecturer in Organisation Studies, Chris Moos, taught a crash course of ‘the power of influence’ to all Oxford Union debate participants. One of his suggestions was to carry an analogy or theme consistently across your presentation. As such, I wrote a prompt asking ChatGPT to make my debate position memorable for the audience and suggest to me ‘a powerful analogy that captures the essence of my argument.’ ChatGPT created a powerful analogy indeed: ‘Just as a tree thrives when planted in its preferred environment with ample sunlight and nourishment, so do employees flourish when given the freedom to work wherever they feel most productive. Embrace flexibility, and watch your workforce bloom with creativity, productivity, and happiness.’

My team from the programme cohort felt a surge of confidence as we delivered our opening statement, eloquently articulating the promise of the future workforce, the need to allow employees to choose where they work, for a better workforce of tomorrow. However, as the debate progressed, we encountered unforeseen challenges. The opposition adeptly presented ethical dilemmas, fears of the inadequacy of the human making the best decision based on one key word from the House position, ‘feel’. The opposition leveraged the word ‘feel’ to counter that corporations cannot be at the mercy of some pathetic, immature human and their ‘feelings’. While ChatGPT had prepared me with more than ten powerful forecasted counterarguments, the dynamic nature of the debate demanded a level of adaptability and critical thinking that proved difficult to fully grasp with the assistance of an AI. ChatGPT did not prepare me to ‘feel'.

The primary difficulty faced was the subtle nuances of human emotions and empathy, traits that ChatGPT, despite its impressive abilities, could not replicate or even forecast. The opposing team appealed to the audience's emotions, highlighting the fear of a future where employees have more control of employers, and fault of human judgment. Moreover, ChatGPT's responses, while comprehensive and well-informed, were too technical and lacked the emotional intelligence necessary to truly connect with the audience. This reflected a concept shared by Chris Moos in his lecture; the art of persuasion requires not only data-driven reasoning but also the ability to forge an emotional connection, a skill that remains uniquely human.

I was the team’s closer. I was nervous when I stood in the grandeur of the Oxford Union for the first time. So nervous, I forgot to include my ChatGPT-prepared remarks and statistics and I ended up sharing an emotional, personal example of why working from home as a mother of an infant son with a rare disease was mission-critical to keeping me in the workforce during the pandemic when we had a mass exodus of diverse employees. The US Chamber of Commerce estimates one million women left the workforce during that time, and a lack of flexibility was a culprit. Emotional, and unanimously supported with applause from the audience (even the opposition) for my inclusion of a more diverse workforce, I relinquished the final 60 seconds and bailed on the remaining AI-generated talking points.

Despite the loss, the debate served as a poignant reminder that AI, while invaluable for information retrieval and analysis, cannot replace the ingenuity of human intellect and emotional understanding. Ethical debates are multi-faceted, requiring empathy, compassion, and moral reasoning - qualities that arise from lived experiences and deep contemplation. In the aftermath of the debate, I reflected on the limitations of relying solely on AI to prepare for such nuanced discussions. While Generative AI had undoubtedly enriched my knowledge and expanded my understanding of the debate topic, it also underscored the irreplaceable value of human intuition and emotional intelligence.

As AI continues to advance, debates on its ethical implications will remain a pivotal point of discussion in various forums, including the prestigious Oxford Union. The experience of collaborating with Generative AI highlighted the need for a harmonious relationship between human intellect and AI capabilities, reinforcing the notion that the future lies in collaboration, rather than competition, between humans and machines. Walking out of the Oxford Union halls, I embraced the valuable lesson - true debate encompasses not only the pursuit of knowledge, but also the cultivation of empathy and wisdom - attributes that will forever set humans apart from the machines we create.

Oxford Executive Diploma in AI for Business