Posit Academy in the Age of Generative AI - Lessons from the Frontlines
summary
Posit Academy runs a ten-week curriculum with nearly 200 learners (mix of R and Python, many PhDs) and finds AI coding assistants rated poorly — over half of learners give them two out of five stars. Generative tools excel at explanations, documentation and refactoring and work well inside IDEs or GitHub Copilot, but they fail to teach threshold concepts (tidy data, modular code, visualization as exploration) and learning dispositions. Recommendation: introduce assistants gradually (start around week three for code explanations, escalate to IDE integration and paid copilots near project showcase).
AI-written summary · about
chapters (8)
- 0:00Generative AI and Posit Academy intro
- 2:39Why AI coding assistants failed
- 4:51Threshold concepts in data science
- 7:22Beyond threshold concepts
- 8:32AI tools analogies for coding
- 10:20AI assistants in curriculum
- 13:40Threshold concepts for generative AI
- 14:45Q&A on AI tools
related talks
43mBecome a 10x developer by harnessing the power of Generative AI
33mAI assisted software delivery
25mDemocratising learning of AI - experiences from the University of Oxford
45mPanel Discussion on AI for Code Generation
40mBuilding the Future Developer: Trends in Open Source, Education...
22mAI, Make It So - A Dive into the Universe of Prompt Engineering
01h 02mPANEL - Unleashing the Future: The Impact of Gen AI on Development Productivity
29mImpact of #GenerativeAI in Development and #SoftwareTesting