“I used to be motivated to learn, and now there’s just really not much to be motivated for when you feel that school is more about just receiving a grade than learning, and there’s such an easy way for people to do that and cheat through that,” Pennington said.

Students said that they work less hard now than they did earlier in high school. They reported that the rigor of some of their courses has dropped to meet students at their diminished level.

AI Class Notes #10: When AI Helps Too Much

AI is evolving from a classroom curiosity into an institutional standard. This issue covers the K-12 research surge, new federal grant priorities, and the ethical guardrails being built through middle school pilot programs and administrative advisory councils.

Headlines

Research Update: UTSA Partners with 60 Educators for Middle School AI

• Researchers at UT San Antonio are collaborating with 60 educators across seven middle schools to develop classroom-ready AI lessons for English, science, history, and math.
• The two-year study aims to observe how teachers’ personal AI literacy translates into actual classroom practice and how these instructional choices shape student conceptual understanding. Source

Discuss: Should AI literacy be taught as a standalone subject, or should it be woven into core subjects like history and math?

Industry Shift: Frontline Education Launches K-12 AI Advisory Council

• Frontline Education, a major provider of K-12 administrative software, announced the formation of a dedicated AI Advisory Council to steer ethical integration across its platforms.
• The council will focus on balancing administrative automation—such as staff scheduling and hiring workflows—with rigorous student data privacy and human oversight. Source

Discuss: What specific administrative tasks in K-12 carry the highest risk if handed over to AI without human verification?

Policy Progress: US Department of Education Finalizes AI Grant Priorities

• The U.S. Department of Education issued a final rule this week that prioritizes federal discretionary grants for K-12 districts that demonstrate a commitment to integrating AI literacy into their teaching.
• The new priorities focus on the "appropriate and ethical use" of AI, giving higher weight to proposals that use technology to improve student outcomes and support educator retention. Source

Discuss: How can districts with fewer resources compete for these grants if they lack the initial infrastructure to pilot AI tools?

Evidence Base: Stanford AI Hub Adds 133 New K-12 Research Papers

• Stanford’s AI Hub for Education expanded its Research Study Repository this week, bringing the total to over 1,270 peer-reviewed and pre-print studies specifically focused on generative AI in K-12.
• The repository’s rapid growth—nearly 15% per month—indicates that the academic evidence base for AI is finally catching up to the industry's marketing, with a heavy focus on "human-AI collaboration" models. Source

Discuss: With the research base growing so quickly, how can busy K-12 administrators stay updated on which "evidence-based" practices actually work?

Perspectives

Only 45% of Principals Report Having Any AI Policy

• More than one in five Texas teachers say their district provided no training on tools they were already using.
• The gap is not enthusiasm; it is the space between teachers already using Gen AI and institutions that have not caught up with clear expectations. Source

Try it: Use the Generative AI Adoption Checklist to audit your school's readiness across 46 items in eight areas.

Students Are Rarely in the Room When AI Policies Get Written

• Students know how Gen AI actually gets used in their peer groups and understand the pressures behind shortcuts.
• Leaving them out of policy conversations means missing critical insight into the reality of the classroom. Source

Try it: Invite three to five students to review your draft AI guidelines and identify gaps between policy and reality.

The First AI Effects Are on Teacher Workload, Not Student Outcomes

• Early adoption data shows educators using Gen AI for lesson plans, assessments, and administrative tasks.
• While helpful for efficiency, it is not the same as improving student learning outcomes directly. Source

Try it: Track one week of your AI use. Categorize each use as "saves me time" or "improves student learning" and notice the ratio.

Tools

Raindrop.io for Content Curation

• A free bookmarking tool that lets you save articles, videos, and links into organized collections.
• Tag items, add notes, and share collections publicly to solve the "seventeen browser tabs open" problem. Source

Try it: Create one Raindrop collection for your main teaching topic and spend one week saving content there instead of browser tabs.

BoodleBox Bots for Newsletter Workflows

• Build a custom bot with your voice, tone, and recurring sections encoded into the system.
• This allows you to make the editorial decisions while the AI handles the mechanical formatting work. Source

Try it: Write three sentences describing your communication style, then use them as the foundation for a custom bot's instructions.

MathosAI and Khanmigo for Guided Math Support

• MathosAI claims higher accuracy than general models for math, while Khanmigo guides students conversationally rather than providing immediate answers.
• These tools include teacher dashboards to track where students are struggling in real-time. Source

Try it: Have students solve one problem independently, then use Khanmigo to check their reasoning rather than just the final answer.

3-2-1 Reflect

3 Ideas to Consider

1. Productive struggle is where learning happens. Tools that remove struggle remove learning.
2. Your AI policy is incomplete if students were not consulted during its creation.
3. Teacher time savings and student learning improvements are two different outcomes. Track both.

2 Questions to Ask

1. Does this specific use of Gen AI deepen learning, or simply replace the thinking process?
2. What verification process exists before AI-generated content is published or sent to parents?

1 Action to Take Review your current AI practices against the Generative AI Adoption Checklist and identify three gaps to address with your team this month.

AI Class Notes #5: Save Your Prompts, Start Your Day

You have new ways to streamline your AI workflow and bring a little joy to your classroom mornings. Chrome now lets you save prompts as reusable shortcuts, and a simple Google Gem can generate personalized student greetings. Meanwhile, major tech companies are making moves that could shape how we access AI.

Headlines

Google Releases Windows Desktop App

Google launched a Windows desktop app with several useful features:

• Quick search across the web, Google Drive, and local files in one place
• macOS Spotlight-like search box for fast access
• Screen sharing capabilities built in
• Unified interface for finding information across platforms

Discuss: How do you currently search across your files and the web? What would unified search change about your workflow?

Amazon Acquires Globalstar for Satellite Expansion

Amazon agreed to acquire satellite operator Globalstar, with the deal set to close in 2027:

• Expands Amazon's low Earth orbit (LEO) satellite network
• Amazon and Apple report LEO will power some iPhone and Watch services
• Signals growing investment in satellite-based connectivity
• Could affect how rural schools access internet in the future

Discuss: How might satellite internet change connectivity for schools in underserved areas? What opportunities could this create for remote learning?

--- ---

Anthropic Redesigns Claude Code Desktop

Anthropic rolled out a major visual overhaul to its coding tool:

• New sidebar for managing multiple sessions
• Drag-and-drop layout for organizing projects
• Integrated terminal built into the interface
• File editor included for quick changes
• Over 1 million views on the announcement

Discuss: As AI coding tools become more polished, how might they change what we teach students about programming? What skills remain essential?

Perspectives

Morning Greetings Can Be Personal and Visual

A custom Google Gem can generate personalized morning greetings for students:

• Creates unique greetings tailored to your class
• Generates AI images to accompany the message
• You control the tone and style
• Quick way to start the day with positivity
• No design skills required

Try it: Open the Morning Greetings Gem, ask what it can do, then click "Create image" to generate a visual greeting for your class.

Prompt Structure Improves Image Results

When using image generators like Nano Banana, specificity matters. Strong prompts include:

• Subject: Who or what appears in the image (a fluffy calico cat)
• Composition: How the shot is framed (extreme close-up, wide shot)
• Action: What is happening (brewing coffee, reading a book)
• Location: Where the scene takes place (sunny meadow, cozy library)
• Style: The overall aesthetic (watercolor painting, digital illustration)
• Editing Instructions: For modifying existing images, be direct and specific

Try it: Generate a classroom image using this structure: "A diverse group of elementary students, medium shot, reading together on a colorful rug, in a cozy classroom library corner, illustrated in a warm watercolor style."

Tools

Chrome Skills

Save your favorite AI prompts as reusable one-click workflows:

• Type a forward slash ( / ) in Gemini within Chrome to access
• Choose from 50+ preset Skills for common tasks
• Create your own Skills for scanning documents, summarizing text, or drafting emails
• Eliminates retyping the same prompts over and over
• Build and share a personal library of your most-used prompts

Try it: Create a Skill for a task you do weekly, like generating discussion questions from an article or drafting parent communication.

Claude Code Routines

Set up automated workflows that run on a schedule:

• "Routines" let you configure a workflow once and run it automatically
• No manual triggering required after setup
• Manage multiple projects in the new sidebar
• Run sessions side-by-side for comparison
• Best for recurring, predictable tasks

Try it: If you use Claude Code, explore setting a routine for a recurring task like generating weekly vocabulary lists or formatting assessment data.

Nano Banana Image Generation

Create images by specifying detailed elements in your prompt:

• Available in BoodleBox by typing @nano-banana-pro
• Responds well to structured prompts with multiple elements
• Can edit existing images with specific instructions
• Works for classroom visuals, presentations, and creative projects
• Experiment with different styles (cartoon, realistic, watercolor)

Try it: In BoodleBox, type @nano-banana-pro and request: "A friendly robot teacher helping a student with math, close-up shot, pointing at a whiteboard, in a futuristic classroom, digital illustration style."

3-2-1 Reflect

3 Ideas to Consider

1. Saving prompts as reusable Skills could reduce the friction of using AI for routine tasks.
2. Personalized greetings, even AI-generated ones, can set a positive classroom tone.
3. As AI tools get more polished interfaces, the barrier to entry keeps dropping.

2 Questions to Ask Yourself

1. Which of my daily tasks involve repeating the same AI prompt?
2. How might I use visual AI tools to create moments of connection with students?

1 Action to Take Identify one prompt you use frequently and save it as a Chrome Skill or keep it in a document for quick access.

AI Class Notes #8: Robots Shoot Hoops, Claude Gets Creative

You have new tools for visual design, a smarter coding assistant, and a basketball robot that rarely misses. These updates signal where AI is heading: more capable, more visual, and increasingly physical.

Headlines

Claude Design Lets You Collaborate on Visual Work Anthropic launched Claude Design, a new product that lets you work with Claude to create slides, prototypes, one-pagers, and other polished visuals. Instead of starting from scratch in design software, you describe what you need and iterate through conversation. The tool can apply your team's design system automatically and export to PPTX, PDF, Canva, or standalone HTML.

Discuss: What classroom materials could you co-create with an AI design partner? Where does human judgment still matter most in visual communication?

Canva AI 2.0 Rebuilds Design from the Ground Up Canva unveiled AI 2.0, powered by what they call the first foundation model built specifically for real-world design. The platform now generates complete, fully layered designs through conversation and includes new workflows for web research, scheduling, and brand intelligence. It moves Canva from a template tool to an AI design partner.

Discuss: How might conversational design tools change the way students learn visual communication skills? What gets gained or lost?

Claude Opus 4.7 Handles Your Hardest Coding Tasks Anthropic released Opus 4.7, their most capable model yet for complex software engineering. Users report handing off difficult, long-running coding work that previously needed close supervision. The model verifies its own outputs before reporting back.

Discuss: If AI can now handle advanced coding with minimal oversight, what should introductory computer science courses prioritize teaching?

Perspectives

Toyota's Basketball Robot Shows Where Embodied AI Is Heading Toyota's CUE7 robot combines reinforcement learning with hybrid control systems to shoot baskets with near-perfect accuracy. The bipedal robot visually locks onto targets, calculates trajectories, and sinks shots from remarkable distances. Toyota uses it as a testbed for vision systems, motion control, and physical AI.

Try it: Show students a CUE7 video and ask them to identify the different AI systems working together (vision, calculation, motor control). Have them brainstorm other physical tasks that require similar coordination.

McKinsey Outlines 12 Themes for AI Transformation McKinsey's AI Transformation Manifesto identifies how leading organizations build AI capabilities and scale results. The research emphasizes that successful AI adoption requires more than tools. It demands new workflows, skill development, and organizational change.

Try it: Pick one theme from the manifesto and map it to your school or district. What would "building AI capabilities" look like for your teaching team? Draft three concrete first steps.

Tools

Claude Design – Collaborate with AI to create slides, prototypes, and visual documents through conversation. Powered by Claude Opus 4.7. Available for Pro, Max, Team, and Enterprise subscribers.

Try it: Describe a one-pager you need for an upcoming unit or meeting. See how far you can get through conversation alone before needing to manually adjust.

Canva AI 2.0 – Generate complete designs, documents, and websites through natural language prompts. Includes Memory Library that learns your style over time.

Try it: Ask Canva AI to create a student handout for a concept you teach. Compare the result to a template you would have customized manually. Note what worked and what needed fixing.

3-2-1 Reflect

3 things to consider:

• AI design tools are shifting from "pick a template" to "describe what you need"
• Advanced AI models now verify their own work before delivering results
• Physical AI (robots) requires multiple systems working in coordination

2 questions to ask yourself:

• Which of my current workflows could benefit from conversational design tools?
• How do I help students understand AI as systems working together, not magic?

1 action to take:

• Try one AI design tool (Claude Design or Canva AI 2.0) for a real task you have coming up. Note where it saves time and where you still need to intervene.