The event highlighted the critical importance of evaluating AI outputs and examining the concept of bias. Justin discussed the human tendency to confirm hypotheses, referencing Wason's rule and Karl Popper's philosophy that every hypothesis has an observation that can refute it—such as the famous example of swans.

A particularly interactive part of the morning involved a game to "uncloak" a GPT, where parents tried to distinguish between AI-generated and human-generated content. This led to lively discussions around safeguards and ethical considerations, with questions ranging from whether AI has a soul to technical inquiries about its prompting architecture.

Justin also delved into the crucial issue of AI regulation, emphasising the importance of a bottom-up approach in schools, especially concerning policy and tasks like report writing. This sparked a debate on whether AI should be allowed in report writing, with arguments presented for both sides. Justin provided valuable insights into the report framework used at CATALYST, which leverages AI to generate the final report.

Parents contributed significantly to the conversation, sharing how AI is utilised in their workplaces and discussing the value of pupil reports. They noted how reports are read differently depending on their context and purpose. The session concluded with thoughts on the long-term impact of AI, drawing parallels to past technological innovations like the microwave, and its effects on skills such as remembering phone numbers and social dynamics.

Overall, it was an engaging morning filled with thoughtful discussion and a strong sense of community interest. If you would like to receive a PDF of the slide highlights from the event, please get in touch with Jonathan Lloyd. Thank you to everyone who attended and contributed to making this event a success.

Jonathan Lloyd
Head of Digital Learning and AI