A practical learning path · 2026

Use AI to extend thinking—not replace it.

A curated field guide for educators exploring AI tools, agentic systems, classroom practice, assessment, and institutional policy.

Begin the learning path Read the Grade 6 guardrails
For K–12 and higher education Checked July 30, 2026 No tracking or sign-up
Two adult educators in a bright learning commons discussing work on a laptop with books and handwritten notes.

The tool can assist. The educator still decides what matters.

The central question

“Can AI produce an answer?” is no longer the interesting question. The question is how to use it without surrendering judgment, creativity, disciplinary knowledge, or productive struggle.
If you explore only six things

A path from chatbot to agentic thinking.

Do not try to consume everything. Follow the sequence, try one authentic task at each stage, and decide what deserves deeper attention.

Read before student use

Exploration and student access are not the same thing.

Student-facing use must follow district approval, privacy requirements, product terms, and age restrictions—not merely the technical availability of a website.

Teacher exploration An educator privately experiments using non-sensitive materials.
Student-facing use Students access a tool, create accounts, enter prompts, or submit classroom work.
Maple Ridge–Pitt Meadows · SD42

Start with the local direction.

  • SD42’s May 2026 update says the initial 2026–27 student rollout will provide education-version Microsoft Copilot to Grades 9–12 at teacher discretion. Elementary use remains part of a measured discussion about purposeful technology.
  • SD42’s consent and technology guidance says personal or sensitive information must not be entered into unapproved third-party applications, including generative AI tools.

Public products have their own restrictions.

Product availability does not override school or district policy.

ChatGPT
At least 13, or the local minimum age of consent; parent or guardian permission is required under 18.
Claude
At least 18, or the higher minimum age required in the user’s location.
AI Studio
At least 18 and intended for professional or business development—not a student-facing service for minors.
NotebookLM
Available to all ages through Google Workspace for Education when enabled and administered by the school or district. Personal accounts have separate age requirements.

For Grade 6, begin with teacher-led demonstration using district-approved tools and non-identifying material. Keep student reasoning on paper or in discussion, and confirm any direct student use with the school or district.

Core tools and concepts

Explore the landscape in three lanes.

Start with conversation, move into creation, and then study the systems that let AI plan, act, observe, and adjust.

Lane one

Converse & study

ChatGPT fundamentals

Beginner

A concise introduction to effective conversational use before moving from one-off questions to sustained workflows.

Master ChatGPT in One Video

25 min

A practical Futurepedia walkthrough. Use it for workflow ideas rather than memorizing an interface that will keep changing.

NotebookLM

Beginner

A source-grounded research and study environment with citations, study guides, audio overviews, mind maps, and quizzes.

Lane two

Create & prototype

Google AI Studio

Intermediate

An adult developer environment for experimenting with Gemini models, prompts, multimodal inputs, structured output, and prototype applications.

Watch the 28-min guide ↗

Visual Studio Code

All levels

A free, open-source code editor and useful home base for experimenting with files, extensions, code, and coding agents.

Lane three

Work with agents

OpenAI Codex

Agent

Works with projects, inspects and edits files, runs commands, researches, tests its work, and reports results.

22-min tutorial ↗ 10-min Skills tutorial ↗

Claude Code

Agent

Anthropic’s agentic coding tool. The selected tutorial begins in the desktop app rather than the terminal.

22-min beginner guide ↗

What is MCP?

5–10 min

The official introduction to the open standard connecting AI applications with data sources, tools, and workflows.

Building Effective AI Agents

15–20 min

Anthropic’s useful distinction between fixed workflows and agents that choose their own actions, with an emphasis on simple composable systems.

Knowing When Not to Use AI

11 min

Compare AI agents, conventional rules, machine learning, and human judgment before deciding AI is the right tool.

See the tools and ideas in action.

Product interfaces age quickly. Prefer recent videos, then verify important details in the official documentation.

Grade 6 teacher path

AI literacy through mathematical thinking.

The goal is not faster answers. It is stronger explanation, verification, challenge, and responsibility for the mathematics.

02

SD42: AI update and next steps

The current local direction on equity, safety, student well-being, and the distinction between secondary rollout and elementary exploration.

05

ChatGPT Foundations for Teachers

A practical vendor course on prompting, lesson planning, rubrics, communication, critical review, privacy, and human oversight. It supports teacher learning; it is not permission for student use.

University professor & AI committee path

Move from tool adoption to institutional judgment.

Use current provincial and university guidance for policy, then treat vendor research and product case studies as useful—but interested—evidence.

UBC: Teaching with GenAI

Practical faculty guidance on course design, learning activities, feedback, assessment, approved tools, and privacy review.

UBC: AI and Assessments

Assessment redesign centred on critical thinking, knowledge application, academic integrity, and clear student expectations.

What Does AI Mean for Education?

A 42-minute Anthropic discussion of personalized learning and assessment alongside deeper questions about what students should learn.

Vendor perspective

How Educators Use Claude

An empirical snapshot of higher-education use across teaching, research, administration, custom tools, and grading. Read the methodology critically.

Vendor research

Claude for Education

An emerging institutional model: campus access, learning mode, guided questioning, administrative use, and university-wide governance.

Product case study
Keep learning

YouTube channels worth following.

A mix of primary product sources, conceptual explainers, and practical independent tutorials.

OpenAI

Primary-source product announcements, demonstrations, research, and official examples.

Visit channel

Anthropic

Primary-source material on Claude, Claude Code, agents, MCP, safety, and education.

Visit channel

Futurepedia

Accessible walkthroughs of fast-moving tools. Expect promotions and check publication dates.

Visit channel

Leon van Zyl

Practical coding-agent experiments and examples of building working software from ideas.

Visit channel

IBM Technology

Short conceptual explanations of agents, language models, RAG, MCP, security, and system design.

Visit channel

Official channels are strongest for current product behaviour, but naturally present the vendor’s perspective. Independent tutorials are often more practical, but may contain sponsorships, affiliate links, or quickly aging advice. Both are valuable when read critically.

A good way to continue

Choose one authentic task—not another pile of links.

  1. Turn a set of readings into a source-grounded seminar guide.
  2. Compare three interpretations of a literary passage and audit their evidence.
  3. Create a Grade 6 mathematics misconception activity.
  4. Ask an AI agent to build a small interactive learning tool.
  5. Redesign an assessment so students document and defend their process.
  6. Draft an AI-use disclosure rubric for a course or classroom.

Keep the task small. Save the prompts and outputs. Note where human judgment changed the result. That reflective record is more valuable than collecting dozens of tools.