Boiler Up for the Future of Education with Purdue Professor of Practice Dr. Kathryn Dilworth
Description
School’s Out Saturdays on the AGI (Podcast) - The most important conversation with the people shaping tomorrow one lesson at a time.
Episode Summary:
Jason sits down with Purdue Professor Dr. Kathryn Dilworth to unpack her groundbreaking QUIZ methodology that's teaching students to control AI instead of being controlled by it. From building custom LLMs with student coders to discovering why female students get better AI outputs than males, this conversation challenges everything you thought you knew about AI in education.
Key Timestamps:
[03:45 ] - Why libraries rejected AI education partnerships and the resistance in humanities
[08:12 ] - China's half marathon with 20+ robots vs American Bigfoot beer videos
[11:30 ] - High stakes vs low stakes AI use and the QUIZ model origin
[15:45 ] - Query: Starting with forced prompts and prewriting without A
I[18:20 ] - Understanding: Reading everything the AI generates
[20:15 ] - Interrogation: Leaving AI to verify all factual claims
[22:40 ] - Zeitgeist: Shaping content for specific audiences and contexts
[25:30 ] - Students asking permission NOT to use AI and ethics instruction
[28:45 ] - Finding your "ChatGPT moment" - what AI frees you to do
[32:10 ] - Gender differences in AI prompting styles and outcomes
[35:50 ] - The Alexa politeness experiment and training AI through behavior
[38:30 ] - Students using AI for mental health support - concerns and context
[42:15 ] - STEM skills need EQ, humanities skills need AI literacy
[45:40 ] - Philosophy and psychology asking the best questions about AI
[48:20 ] - Students' fear and stress about AI's impact on their futures
[51:05 ] - Building a custom LLM with QUIZ guardrails from the ground up
[54:30 ] - Computer science students emerging from philanthropy class
[56:45 ] - Creativity paradox: Students love AI for creativity, fear losing it
[59:10 ] - Domain expertise and staying in your lane with AI
[1:00:45 ] - Alpha School's 2-hour AI learning model and educational reform
[1:02:15 ] - Van Halen's brown M&Ms as AI quality control test
Guest Links
Dr. Kathryn DilworthProfessor of Practice | Purdue University John Martinson Honors CollegePhilanthropy, Academic Writing, and Information Science specialist focused on AI for Education, Trust in Nonprofits and Fundraiser WellbeingLinkedIn: https://www.linkedin.com/in/kathryn-dilworth/Purdue University John Martinson Honors College: https://honors.purdue.edu/Follow UsInstagram @AGISchoolsOutFacebook + LinkedIn: @AGIPodcastStay in the LoopFuture-proof your organization, your team, and your career → work with JasonLinkedIn: https://www.linkedin.com/in/jason-grant-padgett/Podcasts, keynotes, educational workshops, and one-on-one coaching📧 jason@phoenixsolutionsgroup.orgPhoenix Solutions GroupAffordable, pragmatic education, training, and consulting in AI literacy, policy, and projects.🌐 https://www.phoenixsolutionsgroup.org/Tools We RecommendMindStudio → Automate your workflowshttps://get.mindstudio.ai/q7n7u5cttkdqRiverside.fm → Turn one recording into a week's worth of contenthttps://riverside.sjv.io/e1xyBr