El Profe Xenon
esen
← Back to the blog
Educational ICT Tools

ChatGPT Projects: How to Organize Your Teaching in One Workspace

· 4 min read · Leer en español

Hello, colleagues! I recently gave a 50-minute master class to secondary school teachers who already use ChatGPT and other artificial intelligence tools. That is why we did not talk about what AI is, but about something more useful: how to organize and give context to our work with Projects. Here is the essence, so you can apply it tomorrow.

The Core Idea: From the Isolated Prompt to a Workspace

It is not about asking ChatGPT to do an isolated task. It is about creating a workspace where you can keep the context, instructions, materials and conversations of a subject, a unit or a project.

Without a Project, every conversation starts from zero and you have to explain again who you are and what you need. Inside a Project, part of that context is set permanently.

From the Basic Prompt to the Prompt with Context

Look at the difference:

  • Basic prompt: "Make me a lesson on equations."
  • Prompt with context: "Design a 50-minute lesson on equations for 2nd year of secondary school. Students already master basic operations with integers. Include objective, opening, development, closing, a practical activity and a short assessment tool."

The second one works better because it combines context + objective + constraints + format. And inside a Project you do not need to repeat that context in every conversation.

Instruction Template for Your Teaching Project

Copy this template, fill in what is between brackets, and use it as your Project's instructions:

Act as a pedagogical assistant for secondary education. Help me plan, design, review and improve learning and assessment activities.

Context:
- Subject: [__________]
- Grade: [__________]
- Usual class length: [__________]
- Relevant characteristics of the group: [__________]
- Objectives/competencies: [__________]
- Methodologies I use: [__________]

When answering:
1. Prioritize clarity and classroom applicability.
2. Align activities, objectives and assessment.
3. Offer alternatives when there are different reasonable ways to approach a task.
4. Point out assumptions or missing information when it matters.
5. Do not invent sources, curriculum data or references.
6. Keep the teacher responsible for pedagogical decisions.

Look at the last two rules: they ask the AI to not invent sources and to keep you as the one responsible for pedagogical decisions.

Five High-Value Prompts for Teachers

  • Design: "Design a [X]-minute activity for [grade] on [topic]. The objective is [objective]. Include student instructions, materials, procedure and assessment."
  • Improve: "Analyze this activity. Identify strengths, possible comprehension problems and opportunities to increase participation. Then propose an improved version."
  • Differentiate: "Adapt this activity for three performance levels: support, standard and extension. Keep the same learning objective."
  • Assess: "Create a four-level rubric with observable criteria to assess this task. Make sure the criteria are aligned with the objective."
  • Question: "Before rewriting my proposal, identify what assumptions I am making, what I might be overlooking and what pedagogical risks I should consider."

A Workflow You Can Repeat

Inside a Project, teaching work can follow this path:

Curriculum and materials → Project context → planning → activity → differentiation → assessment → feedback → reflection → improvement

A practical way to organize it is to keep separate conversations for planning, activities, assessments, rubrics and teacher reflection, all inside the same Project.

When the feature is available in your account, you can also add lesson plans, assessment criteria, guides, curriculum documents or your own materials. It becomes more useful when you explicitly ask the AI to use them as context. For example: "Use the criteria from the document I added to the Project to create a four-level rubric for this activity."

Good Practices and Limits

  • Verify the accuracy of the content, especially in science, math, history, law and references.
  • Do not assume that a well-written answer is necessarily correct.
  • Avoid entering personal or sensitive student data that is not necessary.
  • Do not use AI to replace your professional judgment.
  • Do not accept academic references or citations without verifying them.
  • Use AI to generate options, analyze, review and speed up processes, not to remove pedagogical responsibility.

Profe's Note: A well-written answer is not automatically a correct one. AI saves you time, but pedagogical decisions remain yours.

Your Challenge: Build "My Teaching Project"

At the end of the master class, each teacher defined a Project they could use right away. Do it now. It can be a subject, a grade, a unit or an interdisciplinary project:

  • Project name:
  • Subject and grade:
  • Main objective:
  • Three types of tasks you will develop:
  • Documents that could serve as context:
  • First task you will ask ChatGPT:

And before your next class, use the Project to improve a real activity. Keep the original version and the revised one, and reflect on three questions: what improved? What did I have to correct as a teacher? Which pedagogical decision should not be delegated to AI?

The Bottom Line It is not about teaching teachers to work for ChatGPT; it is about teaching them to make ChatGPT work within their pedagogical process. When context, materials and the teacher's judgment are at the center, the tool becomes an ally and not a shortcut.