This week, we're trying something a bit different and doing a short episode. The gloves come off as Libby and Owen engage in a lively debate about the "hallucinations" in large language models (e.g. unexpected and hard to explain errors) and their impact on building educational products.They spar on the nuances of model hallucinations, discussing the various forms and potential consequences. Owen presents a "spicy take" on the matter, advocating for the value of engagement and interaction even if it means accepting a certain level of inaccuracy. Libby, however, expresses concerns about the accuracy of information in educational settings, particularly in K-12 schools. She emphasizes the importance of the high bar set by traditional educational tools in terms of factual correctness.Who scores an ed-technical knockout?  You, the listeners, will decide!Join us on social media: BOLD (@BOLD_insights), Libby Hills (@Libbylhhills) and Owen Henkel (@owen_henkel) Listen to all episodes of Ed-Technical here: https://bold.expert/ed-technical Subscribe to BOLD’s newsletter: https://bold.expert/newsletter Stay up to date with all the latest research on child development and learning: https://bold.expert Credits: Sarah Myles for production support; Josie Hills for graphic design

In the sixth episode of this series, Libby and Owen talk to Matt Glanville, Director of Assessment at the International Baccalaureate (IB). The IB works with over 5,000 schools worldwide in 160 countries to offer a range of curriculum programmes and qualifications. Early last year they shared their progressive stance on AI – rather than banning AI tools to try to prevent cheating, they recognize that they will become part of everyday life, and therefore students need support to use them ethically, transparently, and safely. Matt (another ex-teacher ❤) talks about IB’s position and how they deal with cheating. We discuss what’s new about cheating in a post-ChatGPT world and what isn’t (spoiler – malicious and determined cheating isn’t a new problem for schools). Matt poses some provocative questions about how we define cheating. Is it cheating to use large language models (LLMs) to give you ideas for your essay? What about having an LLM write an essay using the ideas you give it? We also touch on the implications for formal assessments. Will oral exams make a comeback? Most importantly, we introduce Owen to Mills & Boon. Guests and resourcesThe International Baccalaureate’s AI resources and content The International Baccalaureate’s March 2023 statement about ChatGPT and AIMatt talking about IB’s position on AI at the ACES annual meeting Join us on social media: BOLD (@BOLD_insights), Libby Hills (@Libbylhhills) and Owen Henkel (@owen_henkel) Listen to all episodes of Ed-Technical here: https://bold.expert/ed-technical Subscribe to BOLD’s newsletter: https://bold.expert/newsletter Stay up to date with all the latest research on child development and learning: https://bold.expert Credits: Sarah Myles for production support; Josie Hills for graphic design

Sian Cooke discusses the potential of generative AI in education, particularly in saving teachers' time and workload. She emphasizes the importance of technology access and addressing the digital divide. The chapter also highlights the need for AI literacy among teachers and addresses misconceptions and concerns about AI in education.

In the fourth episode of this series, Libby and Owen talk to John Roberts, co-founder and Director of Product and Engineering at Oak National Academy. Oak was originally created as an online classroom in 2020 as a rapid response to the coronavirus outbreak. They have delivered over 150 million lessons in their online classroom. They have now become a new national body supporting curriculum and providing free resources to teachers of 4- to 16-year-olds in England.Oak National Academy recently launched a couple of AI experiments (a quiz designer and lesson planner) designed to help teachers save time. John talks us through the experience of starting Oak during the pandemic, what potential he sees for AI to help save teacher’s time, his thoughts on AI generating quality content, and how they’re optimising base large language models for education (and find out more about retrieval augmented generation or RAG).There’s also some good bonus content from John (an ex-physics teacher) on what a negative displacement in a longitudinal wave indicates...Guests and resourcesTES profile on John Roberts Information on Oak’s AI experimentsInformation about Oak’s approach to open government licensing Join us on social media: BOLD (@BOLD_insights), Libby Hills (@Libbylhhills) and Owen Henkel (@owen_henkel) Listen to all episodes of Ed-Technical here: https://bold.expert/ed-technical Subscribe to BOLD’s newsletter: https://bold.expert/newsletter Stay up to date with all the latest research on child development and learning: https://bold.expert Credits: Sarah Myles for production support; Josie Hills for graphic design

Kristen DiCerbo, Chief Learning Officer at Khan Academy, discusses the innovative AI tutor, Khanmigo, which uses GPT-4 technology. She shares her journey melding education with tech, highlighting the transformative potential for special education. Kristen explains the principles of mastery learning that guide Khan Academy’s approach, emphasizing personalized learning paths. She also reveals insights from collaborating with OpenAI and explores the future of AI in education, illustrating how technology can enhance student engagement and knowledge retention.

Join two former teachers - Libby Hills from the Jacobs Foundation and AI researcher Owen Henkel - for the Ed-Technical podcast series about AI in education. Each episode, Libby and Owen will ask experts to help educators sift the useful insights from the AI hype. They’ll be asking questions like - how does this actually help students and teachers? What do we actually know about this technology, and what’s just speculation? And (importantly!) when we say AI, what are we actually talking about? In this episode, Libby and Owen talk to Dr. Dora Demszky, Assistant Professor in Education Data Science at Stanford University. Dora’s background is in linguistics and natural language processing. Her work focuses on using natural language processing tools to support educators. Dora has been working with large language models for a whole, before the recent explosion of interest. So she shares what it’s been like to move from having to explain what a large language model is, to having the world obsess about them! She also talks about the potential and limitations of using large language models to support educators with activities like teacher coaching and writing feedback to students. Next episode Libby and Owen speak to Dr. Paul Atherton, founder of Fab Inc and Fab Data. Like Dora, Paul is also interested in building AI driven tools for teachers, but he focuses on doing so for teachers in highly resource constrained places like Sierra Leone. Guests and resourcesDr. Dora DemszkyDora’s paper on using chatGPT for teacher coachingCoverage of Dora’s research showing positive benefits of providing automated feedback to instructorsJoin us on social media: BOLD (@BOLD_insights), Libby Hills (@Libbylhhills) and Owen Henkel (@owen_henkel) Listen to all episodes of Ed-Technical here: https://bold.expert/ed-technical Subscribe to BOLD’s newsletter: https://bold.expert/newsletter Stay up to date with all the latest research on child development and learning: https://bold.expert Credits: Sarah Myles for production support; Josie Hills for graphic design

Join two former teachers - Libby Hills from the Jacobs Foundation and AI researcher Owen Henkel - for the Ed-Technical podcast series about AI in education. Each episode, Libby and Owen will ask experts to help educators sift the useful insights from the AI hype. They’ll be asking questions like - how does this actually help students and teachers? What do we actually know about this technology, and what’s just speculation? And (importantly!) when we say AI, what are we actually talking about? In the first episode of this new series, Libby and Owen talk to assessment expert Daisy Christodoulou MBE. Daisy is Director of Education at No More Marking, an online assessment platform. She’s written three books on education, including one on assessment. No More Marking has been doing a lot of research in ChatGPT and its educational applications. Daisy talks about why she thinks we should be hesitant to use ChatGPT in classrooms and schools. She also highlights that ChatGPT is currently too error prone for many educational use cases. Next episode Libby and Owen speak to Dr. Dora Demszky, Assistant Professor in Education Data Science at Stanford University, about using large language models to support educators. Guests and resourcesDaisy Christodoulou MBEDaisy’s guidance on large language models (LLMs) for schools Transcript from The Commons Science and Technology Committee’s inquiry into the use of artificial intelligence (AI) in education, including oral evidence from Daisy ChristodoulouJoin us on social media: BOLD (@BOLD_insights), Libby Hills (@Libbylhhills) and Owen Henkel (@owen_henkel) Listen to all episodes of Ed-Technical here: https://bold.expert/ed-technical Subscribe to BOLD’s newsletter: https://bold.expert/newsletter Stay up to date with all the latest research on child development and learning: https://bold.expert Credits: Sarah Myles for production support; Josie Hills for graphic design