⚖️ Address from Mdm Lee Lin Yee

Divisional Director, Educational Technology Division, MOE

2025-AIEd-Sharing-Leading-with-AI

🧠 Key Points:

Mdm Lee outlined the Ministry’s emerging stance on AI in education, acknowledging its potential and challenges.

1. ⚙️ AI Adoption Philosophy

  • Singapore’s education system doesn’t shy away from AI—MOE supports using AI to:
    • Improve teaching productivity.
    • Enhance feedback quality.
    • Support in-class responsive teaching via real-time data.
  • AI should help teachers teach better and students learn better, not replace pedagogical processes.

2. 📊 Data-Informed Learning

  • AI enables data collection during learning, not just after assessments.
  • Captures how students engage with digital tools, like time-on-task, interaction with chatbots, or cognitive behaviors.
  • Can help assess engagement and attention in class (though raises concerns about over-interpretation).

3. 🧠 Sensorial and Cognitive Learning

  • Learning is not purely cognitive; sensorial and emotional experiences make learning “stickier.”
  • Cognitive disequilibrium (moments of confusion or surprise) paired with emotional resonance boosts learning.
  • Experiential learning remains central despite technological advances.

4. 🧠 The Learning Paradox

  • Referenced research showing declining IQ trends post-1975, possibly due to over-reliance on external memory (e.g., Google).
  • Argued for maintaining foundational knowledge in students’ long-term memory to support higher-order thinking.

🧒 AI Use by Age and Maturity

  • P1–P2 (Grades 1–2): Minimal use of tech; priority is on complete learning experiences.
  • Around P4 (Grade 4): Potential start point for AI introduction, under strong supervision.
  • Secondary level: Greater independence in AI use, aligned with students’ cognitive and emotional maturity.
  • Emphasized that maturity varies, so guidelines are flexible, not rigid.

🧑‍🏫 Pedagogical Grounding

  • AI should support principles like:
    • Zone of Proximal Development (ZPD).
    • Productive struggle: learning shouldn’t be too easy or too hard.
  • Teachers must act as curriculum leaders, guiding students on:
    • When to use AI.
    • How to critically evaluate AI outputs.

🛡️ Updated MOE AI-in-Education Ethics Framework (AIDT)

Four Core Principles:

  1. Curriculum Leadership
    • Teachers must understand both benefits and risks.
    • AI cannot determine what counts as “school-feeding” (i.e., doing students’ work).
    • AI must not substitute human interaction in emotionally sensitive contexts.
  2. Inclusivity
    • Ensure all students have equal access to AI tools.
    • Beware of widening achievement gaps between high- and low-ability students.
    • Digital and AI literacy is critical: some students may just aim to finish tasks, while others use AI to learn deeply.
  3. Human-in-the-Loop
    • Teachers should remain involved in high-stakes decisions (e.g., subject selection, personal advice).
    • Avoid AI making sole decisions in such contexts.
  4. Output Verification
    • Teachers must scrutinize AI outputs:
      • Tools may change back-end engines without transparency.
      • Free tools are often experimental and may switch models for cost-saving.
    • Transparent tools are preferable.
    • Educators must test and verify AI content continuously.

🧭 Final Thoughts

  • Schools have freedom to explore AI, but must do so responsibly.
  • AI should enhance—not replace—good teaching.
  • Educators should design usage with clear guardrails to ensure learning integrity.
  • Emphasis on ongoing critical engagement, teacher oversight, and ethical awareness.

⚖️ Address from Mrs Christine Low

Principal, Boon Lay Garden Primary School

AI-in-BLGPS.-Final

🎙️ Opening Remarks: A Humble, Practical Start

  • Christine Low candidly shared that she came as a learner, not as an expert on AI.
  • She humorously mentioned using ChatGPT to generate slides after being asked to present—found 20% useful, but the rest lacked contextual relevance.
  • Acknowledged the value of AI as a starting point but emphasized the need for localised, school-specific input.

🌱 School Vision: “Garden of a Diverse Classroom”

  • Boon Lay Garden Primary serves a diverse student population, making customisation and personalisation of learning essential.
  • AI is viewed as a tool to support this vision by differentiating instruction and promoting inclusive learning.

🧩 Strategic Framework and Foundations

  • AI is integrated into the school’s strategy map, under the goal of creating an ICT and e-Pedagogy-enabled learning environment.
  • Christine stressed the importance of foundational systems before adopting AI:
    • The EPEC framework (e-Pedagogy for Engaged Classrooms).
    • Strong adoption of SLS (Student Learning Space) since 2018.
    • Gradual progression:
      • 2019–2021: Deepening understanding of e-Pedagogy.
      • 2022: Blended learning and SLS implementation.
      • 2023: Shift toward DI (Differentiated Instruction) powered by e-Pedagogy.
      • 2024: Exploration of Key Applications of Technology (KAP) and AI tools.

🚀 AI Implementation via the “ARISE” Taskforce

  • A cross-functional team named ARISE (AI + “rise”) was created to:
    • Lead the AI journey.
    • Coordinate AI experiments across subjects.
    • Explore emerging AI tools.
    • Create school-wide ethical policies and safety considerations.
    • Encourage teacher-led professional learning.

🧪 Classroom Examples of AI Integration

Christine showcased real classroom examples across subjects, reinforcing AI’s pedagogical value rather than its novelty.

📘 English Language

  • A teacher used an open-source chatbot configured to ask only questions, not provide answers.
  • Encouraged content generation for composition writing (not offloading), aligning with AI as a cognitive enhancer.

✍️ P2 English (Grammar & Writing Feedback)

  • Use of SLS Annotated Feedback for composition writing.
  • AI corrected grammar with natural suggestions (e.g., “mega-thirsty” → “very thirsty”).
  • Supported multi-draft writing—a process traditionally done manually by teachers.

➕ Math (P4)

  • Teachers used SLS Feedback Assistant to guide students through problem-solving steps.
  • Data on performance helped target misunderstandings (e.g., knowing formula vs. applying it correctly).

🔬 Science

  • Used a multimodal AI tool for labeling plant parts and voice recording to explain functions.
  • Tool offered instant feedback on explanations, supporting diverse cognitive modalities.

🈶 Chinese

  • Teacher input detailed rubrics into SLS.
  • AI gave specific, automated feedback aligned with language learning standards.

🏯 Social Studies

  • Students engaged in role-play dialogue with Qin Shi Huang, asking questions about unifying China.
  • Fostered historical inquiry and deeper understanding via AI simulations.

📚 Pedagogical Anchoring of AI

  • Teachers use a pedagogical framework to guide AI use:
    • Start with “Why?” – What’s the learning intent?
    • Align AI tools to goals like cognitive challenge, student voice, multimodality, and inquiry.
  • Emphasis: “Pedagogy before technology.”

🛠️ AI for Teacher Workload & Capacity Building

📝 Report Writing

  • Teachers created custom prompts to generate personalized student remarks.
  • Input included student outcomes, conduct, and learning needs.
  • AI reduced remark-writing time for a class to under a minute.
  • However, first drafts were often “cold”—teachers needed to reinsert their voice and context.

🧪 Lesson Design

  • A teacher built an AI-powered lesson design generator using SLS scaffolds.
  • Teachers input details (topic, level, ability) to receive a draft lesson plan—efficient and tailored.

🔎 Mentoring with AI

  • Senior teachers mentor peers using AI:
    • Example: XT Math teacher uses SLS Data Assistant and Terrain Co-Pilot.
    • AI analyzes student answers and identifies specific misconceptions (e.g., misidentifying base/height in area of triangle).
    • AI flags students needing follow-up.

🌐 Professional Culture and Sharing

  • Regular TTT (Time To Talk) sessions foster professional dialogue and knowledge sharing.
  • Teachers share innovations through platforms like SGLTC and TCF.
  • The school encourages a culture of risk-taking and experimentation with AI tools.

🤝 Equity and AI

  • Christine emphasized AI’s potential to close the opportunity gap for students from disadvantaged backgrounds:
    • Many lack tuition or external support.
    • AI can act as a “personal tutor”, making learning more equitable.
  • Referenced CDC initiative using AI for personalised learning instead of traditional tuition.

🧭 Closing Message: AI as GPS, Not the Driver

  • Reiterated the school’s belief: AI is a means, not the end.
  • Must be anchored in pedagogical principles, customised for context, and used with ethical guidance.
  • Boon Lay Garden is starting small, but fully committed and excited for the journey ahead.

⚖️ Address from Mr Inderjit Singh

Principal, Boon Lay Garden Secondary School

AI-@-BLS

🎙️ Opening Remarks: A Humanist Story

  • Mr. Singh clarified that his presentation was a snapshot, not an exhaustive list of tools.
  • He focused on the thinking behind the school’s AI journey, the pedagogical foundations, and the cultural mindset they cultivated.

🏁 Starting Point: From Digital Divide to Tech Integration

  • In 2020, 35% of students lacked any learning device.
  • The pandemic forced the school to scramble for devices, leading to reflection on why technology matters in learning.
  • Initial push (2021) emphasized consensus building among teachers—many were not tech-savvy.
    • Some were overwhelmed (e.g., not knowing where the power button was on an iPad).
    • Others were excited by how engaged students became with technology use.

🧠 Cultural Shift: From Resistance to Engagement

  • No KPI was set for tech use; instead, teacher autonomy and observation of student engagement led to organic adoption.
  • Teachers began to share experiences, noting improvements in classroom interest and participation.

🚀 AI Adoption: Metacognition Over Offloading

  • The school does not promote cognitive offloading (AI doing the thinking).
  • Instead, they emphasize metacognitive use of AI—students must think before using AI for feedback or refinement.

🛡️ Ethical Foundations & Safety First

  • As AI tools emerged in 2023, safety and ethics became top concerns:
    • Data privacy, accuracy, and appropriate use.
    • Teachers tested tools before use, ensuring pedagogical soundness and reliability.
  • Students are taught to critically evaluate AI output, recognizing that AI isn’t always accurate.

🧑‍🏫 Professional Development and Leadership Structures

  • Formed a Tech-Enabled Learning Team (TEL):
    • Includes representatives from every department and senior teachers.
    • Focuses on pedagogy-first, tech-second integration.
  • Monthly EE sessions help share strategies and updates.
  • The school uses SLS (Student Learning Space) extensively, praising recent improvements.

🧪 Pedagogical Use of AI in Classrooms

🟡 Traffic Light System for AI Use

  • Developed an internal framework:
    • Red: No AI allowed (e.g., certain assessments).
    • Yellow: AI allowed for some parts (e.g., research, drafting).
    • Green: AI can be fully used under guidance.
  • Purpose: To prevent over-reliance and encourage productive struggle.

🧠 Examples of Student Use

  • Students write essays, then use AI to generate counterarguments, helping them think from multiple perspectives.
  • Feedback loops allow students to reflect and say, “This doesn’t sound like my voice”, fostering self-awareness.

🧪 Visual AI Literacy Examples

  • Shared image recognition failures (e.g., huskies vs. wolves on white backgrounds) to teach AI bias and model limitations.
  • These help students learn critical thinking without deep subject mastery.

🧰 AI Tools in Use

  • SLS Short Answer Feedback: Used across subjects for grammar correction and concept feedback.
  • Audio Transcription Tools: For English, Tamil, Chinese—highlight pronunciation or grammar issues.
  • FabChat, PadSearch: Facilitate higher-order thinking and research.
  • Data Assistant + IPP: Analyze student responses, identify misconceptions, and guide teacher intervention.
  • QReport: Summarizes qualitative student input and reveals common patterns.
  • AIBots (in beta): Generate modified SNID lesson outlines with accessibility and inclusivity in mind.

🎨 Creative, Playful Use of AI

  • Used ChatGPT to redesign the school canteen, including menus and signage:
    • e.g., “Ayam Legend” (chicken stall) and “Teliti” (tea shop).
  • Served as low-stakes experimentation that familiarized staff with AI use in fun, practical ways.

⚖️ Balancing Innovation and Caution

  • Recognizes the dualities of AI adoption:
    • Innovation vs. overhype.
    • Pedagogical gain vs. trend-chasing.
    • Freedom to explore vs. need for ethical guardrails.
  • Emphasized constructive struggle as core to learning:
    • AI must not short-circuit the learning process by replacing thinking.
    • Students should use AI after trying on their own to enhance understanding.

🧭 Final Reflections

  • The school maintains a “human-in-the-loop” principle.
    • Teachers must understand AI tools deeply before recommending them to students.
    • AI should augment human creativity and decision-making, not replace it.
  • AI should serve the school’s goals and pedagogy, not the other way around: “We must steer the ship; the ship cannot be steering us.”