In this news and research-packed episode, Ray and Dan dive deep into the AI highlights from EduTech 2025 in Sydney - reflecting on the vibe, standout presentations, and the surprisingly light AI presence on the expo floor. They unpack major news from the UK’s Department for Education, OpenAI’s model pricing shake-up, and raise serious red flags over Meta AI’s privacy approach.
The duo also tackles the big questions educators face: is AI destroying the planet? Can we trust AI with student data? And what do students themselves think? Featuring insights from two key research pieces - Australia’s "From Gen Z to Gen AI" and Jisc’s UK-based "Student Perceptions of AI 2025" - this episode reveals how students are using AI, what they’re worried about, and why institutions need to catch up.
Links:
UK Government's Policy Paper "Generative AI in Education"
OpenAI releases o3-pro (Model Release Notes) and slashes the cost of o3-pro for developers, compared to o1-pro
OpenAI forced to retain all conversation logs by the New York Times lawyers, and details of OpenAI's response
Meta.ai privacy approach - so far, only covered by Crikey
Energy/water usage of ChatGPT
In Sam Altman's blog he said "the average query uses about 0.34 watt-hours, about what an oven would use in a little over one second, or a high-efficiency lightbulb would use in a couple of minutes." (about the same as a Google search)
Dr Karen Boyd's excellent blog post we discussed: "Ethics & LLMs: Sustainability" which also contains the great comparison charts for electricity and water use for common activities.
Details on Karen's upcoming book, Amplify Good Work
Research on students' AI attitudes and use
JISC's research "Student Perceptions of AI 2025", from Sue Attewell