Expertise must precede AI reliance.
This ecosystem of six integrated, free tools—built by a classroom teacher using Gemini, Adobe Express, Suno, and Google AI Studio—helps students work WITH artificial intelligence, not outsource TO it. Grounded in learning science and designed for replication, these resources make AI literacy practical, accessible, and focused on building the critical-thinking skills students need to be confident participants in an AI-assisted workforce.
Retrieval Practice Meets Personalized AI Tutoring
Empowering Teachers to Build What Students Need
Transforming Curriculum Into Stories Worth Remembering
Content That Plays on Repeat—Literally
Voice-Powered Learning for Every Student
Reflection as the Bridge to AI Literacy
My 8th-grade U.S. History classroom includes students from more than 40 home language backgrounds, a wide range of learning needs, and adolescents whose brains are in a critical developmental window.
Traditional instruction asks a lot of a 13-year-old brain. During a lecture, students are listening, deciding what matters, and writing it down, all at once. That's three competing tasks drawing from working memory that can only hold 4 to 7 pieces of information at a time. The brain's planning and focus center won't fully mature for another decade. This isn't a discipline problem. It's normal development, and it's what every tool in this ecosystem was designed around.
When generative AI arrived in classrooms, I saw students reaching for AI before they'd built the knowledge to evaluate what it gave them back. The American Historical Association named this exact risk in their 2025 guidelines: when students rely on AI without developing their own expertise, they end up in "an unproductive loop" where minimal engagement leads to uncritical acceptance of flawed outputs.
🧠 Expertise must precede AI reliance. Students build the knowledge to evaluate AI outputs before they rely on AI to produce them.
🔬 Every tool is grounded in learning science. Retrieval practice, cognitive load theory, dual coding, desirable difficulties, and adolescent neuroscience shape every design decision.
Three things every tool has to do
✅ Support all learners. Bilingual accommodation, scaffolding for diverse learning needs, and accessibility are built into the defaults, not added as afterthoughts. My campus serves students from more than 40 home language backgrounds. Accessibility-first design isn't optional.
✅ Create learning that sticks. If a tool produces short-term engagement but not durable learning, it doesn't belong here. Retrieval practice, spaced review, and desirable difficulties are preferred over surface-level interactivity.
✅ Cost nothing. Every tool runs on free platforms available through Google Workspace for Education or public-facing services. No district budgets, no subscriptions, no paywalls between a teacher and replication.
The ecosystem moves students through a learning arc, with tools designed to be modular. Teachers can use one phase, all four, or substitute their own preferred platforms at each step.
Teach — The Writer's Room Transforms curriculum into short episodic narratives with embedded "Pause and Reflect" questions, built on what we know about story, episodic memory, and cognitive load.
Practice — The Study Coach Gem A session-based AI tutor constrained to TEKS-aligned 8th-grade U.S. History. It uses retrieval practice instead of multiple-choice recognition, and provides automatic language support and step-by-step scaffolding when it detects a student is struggling.
Apply & Create — Student-chosen creation platform Students demonstrate what they've learned by creating. I use Adobe Express and Vibe Coding with Gemini, but this is the most flexible phase of the framework. Canva, Book Creator, podcast tools, video platforms, any creation tool that fits your students and your content works here.
Reflect — The AI Work Log When students use generative AI in the Apply & Create phase, they complete an AI Work Log documenting what they did versus what AI did, which AI outputs they rejected and why, and what knowledge they needed to prompt effectively. When creation doesn't involve Generative AI, the Reflect phase isn't required. This is intentional. Reflection on AI use only makes sense when AI was used.
Three additional tools sit outside the student learning arc but extend what's possible in the classroom.
🎵 Song Lab uses Suno to produce curriculum-aligned songs, leveraging dual coding and musical memory to help students retain content.
🎙️ Audio Lab produces short-form instructional audio and routes teachers to the right platform (Suno for rhythmic mnemonics, Google AI Studio for clean spoken dialogue and World Language pronunciation).
🔧 The Study Coach Gem Creator is replication infrastructure. Any teacher in grades 6–12 can generate a fully customized AI tutoring Gem in about three minutes by answering five questions. One Study Coach serves one classroom. One Gem Creator serves unlimited classrooms.
Designed to be replicated
The ecosystem is released under a Creative Commons BY-NC-SA 4.0 license.
These tools were designed to be stolen.
Take them, adapt them for your subject, share what you learn, and teach the next teacher.
AI Disclosure: Portions of this resource were created with assistance from AI tools including ChatGPT, Claude, Google Gemini, Adobe Express, Google AI Studio, and/or Suno. All prompts, instructional design, curation decisions, and pedagogical frameworks are original work by the author.