AI in Education: What Humans Must Do
AI now appears in every corner of education. The 6 AI Domains help make that landscape navigable. But it is the AIEd Bloom thinking steps that determine whether AI actually adds value.
TL;DR Summary
AI can generate ideas, exercises, and feedback - but it's not the thinking brain in the room. The real value of AI in education depends on what students, teachers, and schools do with it. The AIEd Bloom Taxonomy outlines six thinking steps - Gather, Adapt, Simulate, Process, Evaluate, and Innovate - that keep human thinking central in any AI-supported task. When these steps are missing, learning quality suffers.
Why the AIEd Bloom Taxonomy?
The AIEd Bloom Taxonomy offers schools a cognitive roadmap for responsible AI use:
- Clarifies which thinking step the human must take in every AI-supported task
- Prevents AI from doing the thinking instead of the learner
- Supports subject didactics, step-by-step guidance (scaffolding) and learning strategies
- Forms a basis for lessons, policy, quality assurance and professional development
Read the background article: AIEd Bloom Taxonomy - why these six steps?
| Element | What it describes | Relationship |
|---|---|---|
| AI didactics | Activities, tasks, (step-by-step) guidance | Determines how AI is used |
| AIEd Bloom | Gather → Adapt → Simulate → Process → Evaluate → Innovate | Determines which thinking steps are required |
| The 6 Domains | Where AI enters the education process | Provide context for the didactic choice |
The Six AI Domains for Education
1. Lesson Planning & Learning Material Development
AI does: generate ideas, produce explanations, propose formats and draft concepts.
Therefore, humans must:
- Gather: select appropriate starting points
- Adapt: rewrite material for purpose, level and audience
- Evaluate: check for didactic quality, bias and accuracy
- Innovate: create an original, learning-focused design
What goes wrong without these steps?
Without Evaluate, AI output is trusted blindly. Without Innovate, materials remain superficial.
2. Assessment, Feedback & Evaluation
AI does: detect error patterns, generate initial performance descriptions.
Therefore, humans must:
- Process: interpret AI output in relation to learning goals
- Evaluate: judge reliability, fairness and pedagogical relevance
- Adapt: rewrite feedback into meaningful human guidance
What goes wrong?
AI often gives confident yet inaccurate feedback - without Evaluate, you get false certainty.
3. Differentiation & Personalised Learning
AI does: visualise data, suggest levels, propose learning routes.
Therefore, humans must:
- Process: interpret signals from AI
- Evaluate: check whether a route is appropriate, fair and safe
- Adapt: adjust pathways based on motivation, wellbeing and context
What goes wrong?
Without Evaluate, differentiation becomes automated tracking that amplifies inequality.
4. Learner Support & Instructional Help
AI does: provide explanations, generate examples, run simulations.
Therefore, humans must:
- Gather: request explanations and examples
- Simulate: explore scenarios or practise dialogues
- Evaluate: check accuracy and avoid misconceptions
- Innovate: create new examples or learner products
What goes wrong?
Without Evaluate, learners adopt misconceptions; without Innovate, AI becomes a shortcut instead of a thinking aid.
5. Communication, Organisation & Administration
AI does: minutes, e-mails, reports, summaries.
Therefore, humans must:
- Gather: generate initial drafts
- Adapt: correct tone, nuance and factual details
- Evaluate: ensure accuracy, completeness and privacy
What goes wrong?
AI texts look polished but often contain subtle errors - Evaluate is the safeguard.
6. School Development, Policy & Safeguarding
AI does: dashboards, risk analyses, scenario generation.
Therefore, humans must:
- Process: interpret data within vision, culture and purpose
- Evaluate: weigh values, proportionality and ethical considerations
- Innovate: design policy, agreements and PDCA cycles
What goes wrong?
Without Evaluate and Innovate, policy becomes driven by data instead of by people.
Meta Insight
Across all six domains, the same pattern appears: AI can gather, predict and structure - but only humans can evaluate and innovate.
The quality of AI use in education does not come from the tool, but from the thinking steps that follow. The AIEd Bloom Taxonomy makes those steps explicit, visible and applicable across the whole school.
From Taxonomy to Practice
Would you like to strengthen your team in these six thinking steps - in lessons, policy or professional development?
Symbio6 offers formats, workshops and AI-didactic guidance that help schools use AI safely, humanely and learning-centred. We'd be delighted to help you take your next step.