Free Teacher-Facing AI for Behavior and Learning
Most teachers reaching for AI to help with a struggling student are reaching for the wrong kind of AI. General-purpose chat tools pull from the open internet, they were never trained on whether the information is accurate, and they should be nowhere near a student's personally identifiable information. Natalie Tamburello, a learning disability and education equity advocate at Children's Health Council, walks through what a teacher-facing AI built on vetted research actually looks like — and why she thinks that is where the real classroom wins are right now.
In this episode, you'll learn:
- Why "just ask ChatGPT" falls short when a student is struggling with behavior, learning, or the emotional side of school
- What RAG (retrieval augmented generation) is, and how pulling from a curated library of vetted nonprofit resources changes the answer you get
- Why the first question a good tool asks you is "what are this student's strengths?"
- How to describe a student situation without putting personally identifiable information into any AI
- The clear line between a classroom strategy and clinical advice — and what a teacher-facing tool should refuse to do
Ellis is a free tool for K-12 educators from Children's Health Council. It pulls from a knowledge base of content from trusted partners like CAST, CASEL, Understood and others, it gives you information rather than clinical advice, and it does not collect personally identifiable student information. Today's show was sponsored by Ellis. All opinions are my own and that of the guest. https://www.askellis.org/
Full show notes, links, and the complete transcript: https://www.coolcatteacher.com/e971
Share this one with the teacher down the hall who is still thinking about that one student.
