For a class project, an Australian graduate student creates an AI program that can transcribe audio to written text. They collected their own data by interviewing 50 students. They use the audio and text transcripts to train the model. 

Scenario 2

A Japanese hospital receives a grant to help train an AI lung-cancer detection program. The hospital is viewed to specialize in lung-cancer, has high-grade imaging capabilities, and images that span many years. 

Scenario 3

A U.S. bank wants to build an AI system that identifies people who may default on a loan. They use a large dataset that includes loan default rates for people all over the country. They do not want to discriminate based on sex or race, so they remove these variables from the training data. 

Scenario 4

A technology firm wants to improve efficiency in hiring and creates an AI system to identify resumes for potential candidates. It trains the algorithm using a plethora of resumes from current employees. 

Link: https://pz.harvard.edu/sites/default/files/Think%20Pair%20Share_2.pdf (Think-pair-share)

Group 2: https://padlet.com/abhasiben0308634/card-breakout-room/r4Q6nBjx7O0XnoGa-page:HablbJNFdk_oM65Vb4bbK

Group 3:

https://padlet.com/abhasiben0308634/card-breakout-room/r4Q6nBjx7O0XnoGa-page:9bHz4QQ0vFiW8VE-Ffxqn

Group 4:

https://padlet.com/abhasiben0308634/card-breakout-room/r4Q6nBjx7O0XnoGa-page:ZNuC2UsfalzaM75F2Wp_0

Group 5:

https://padlet.com/abhasiben0308634/card-breakout-room/r4Q6nBjx7O0XnoGa-page:QxVkGzs2w6oeTKUp1xgrT

(Examples of AI in education)

Resources to refer:

Center for Integrative Research in Computing and Learning Sciences (CIRCLS). (2024, January 4). Glossary

of Artificial Intelligence Terms for Educators. https://circls.org/educatorcircls/ai-glossary

Commonwealth of Australia. (2023). Australian Framework for Generative Artificial Intelligence in Schools.

https://www.education.gov.au/schooling/announcements/australian-framework-generative-artificial

intelligence-ai-schools

UNESCO. (2023). Guidance for Generative AI in Education and Research.

https://unesdoc.unesco.org/ark:/48223/pf0000386693

Bibliography:

IMB. (2023). What is Generative AI? https://research.ibm.com/blog/what-is-generative-AI

https://www.conceptmap.ai/ (Free tool)

- image recognition (activator)

What is AI Literacy? Competencies and Design Considerations

https://dl.acm.org/doi/10.1145/3313831.3376727

https://docs.google.com/presentation/d/1d1ZSd9A-k_kWpbxpSyc2gc8TRqs_3q9m/edit?usp=sharing&ouid=114360871162927790666&rtpof=true&sd=true (Slide deck for participants) 

Team MERL

AI is defined as deep learning based on the machine learning, which utilizes the broad data, to mimic human's thinking process to self-generating something.

I = System that passess the "Turing Test" and can make decision

focus more on science aspect.

unique part:

for the government:

focus more on the function.

unique word: "without any human involvement"

for the research organisations and Uni partners:

includes both the sources and the outcomes

unique part: the only one provides the societal impact.

Ai can efficiently find the right candidate via the following:

• Work Experience

• Educational Background

• Name of University

• Time Efficient for head of recruitment

• Shortlisted talent

Limitations:

The research is taken from current employers.

If the company wanted new skills, the Ai will need updating.

The Ai created can discredit talented individuals who might not fit the requirements identified.

Team Mearl

• Bias Alert! ⚠️ AI can show bias based on its data.

• Your Voice Matters! 🗣️ The commands you give affect the output.

• Think Critically! 🧐 Don’t trust everything blindly!
  How do YOU use AI? Share below! ⬇️
  #AICritique #ThinkSmart #StayInformed #AI

Impact(s): such decision would have a direct impact on the process (most likely a kind of supervised learning from onset)

Also, we input our school's best students performance table of different semester, as well as other school's definition and performances.

Let AI to collect the features of 'best students'

How to avoid it?

• 🗣️ Collect data from all kinds of people—different ages, backgrounds, and accents!

• 🎧 Record in fun settings, like noisy cafes, to make your AI tougher.

• ✅ Double-check transcriptions for mistakes.

• 🔍 Test your AI on different groups to catch weak spots.

Try it! Mix up your data and test it out. Can your AI handle a new accent? Share your results! 🚀

________________________________________

AI systems learn from data, but if that data is biased, the AI can make unfair decisions. For example, if a hiring AI is trained mostly on data from one ethnicity, it might unfairly favor that group and discriminate against others.

Bias in AI can lead to discrimination and unfair treatment. To reduce bias, we should use diverse data, test AI outputs regularly, and promote fairness.

As students, you can help by being critical of AI and supporting fair practices. Together, we can build more inclusive technology!

__________________________________________

Safety of privacy:

In relation to school, our students need to be aware that what they post and leaving a large digital footprint can be used against them. It can lead to bullying and other nefarious actions.

These prohibited applications, we appreciate can work for some countries. However we need to obey the rules of other countries when we visit them and appreciate their need for digital security, eg. Fingerprint analysis when going through immigration.

TEAM KLEARM

Part-2

• With your table group, brainstorm potential unintended consequences of AI in education. Discuss ways of increasing teacher agency, with a goal of ensuring they feel an empowered human-in-the-loop with AI.

2. Rafaela was complacent in her confidence of the self-driving cars ability. However the company had given her an agreement of the safety precautions of the car. She may have changed her behaviour if she knew more about the product.

3. The testing will have been completed and the vehicle will be road legal. The engineers will have set the sensors for human detection. However the company and the road department had not given the public an accurate announcement of the cars faults. In this case, the sensors.

4. The human error in this case points towards the lack of signs in the area that label the road as a "Self-driving car road" The largest human error in this case, the low agency. Would be the cyclist that had made the decision to shortcut, and not cross properly over the road.

5. Self-driven cars are convenient, they are the latest trend and also sell the illusion of wealth. In terms of harmfulness, they have the capacity for errors, some may even be fatal errors.

Thank you for reading! TEAM KLEARM!

• Personalisation: How do you think AI can help you personalise your learning experience? Can you give an example of how it might adapt to your individual needs?

• Decision-Making: In what ways do you feel empowered to make choices about your learning when using AI tools? How does this impact your motivation?

• Critical Thinking: How can you ensure that you critically evaluate the information provided by AI? What strategies might you use to verify its accuracy?

• Collaboration: How do you see AI facilitating collaboration among students? Can you think of a project where AI could enhance teamwork and communication?

• Responsibility: What responsibilities do you think you have when using AI in your learning? How can you use these tools ethically and responsibly?

Q2: If you were asked to develop a personalized learning path for the public assessment. How will you use AI to help?

Q3: How could an AI tutor that always gives you the right answer actually slow down your learning?

Q4: If you use AI to generate all your ideas, what might happen to your own creativity?

Q5: If an AI only shows you things it thinks you'll like, how might your worldview become limited?

This tool offers seamless integration with the daily operations of GC. This is however dependent on the availability of machines connected to Ss' school account. While all learning artefacts are created and recorded on the GC platform, specific tasks should be pre-planned and related to the specific lesson planning rather than ad hoc operations (despite the tool is capable for that purpose as well).

However, I have been eagerly looking forward to making Math presentation slides using Magic School which I cannot as it has limitations. Overall, it is a good and handy platform.

1. Enhanced Creativity and Visual Learning

2. Efficient Collaboration and Organization

3. Streamlined Assessment and Feedback

Magicschool pursues different learning styles in addition to getting help for what is needed most. E.g. Text Leveler & Text Scaffolder scaffold text based on the students’ reading levels. I can use these tools to support students reading case studies and articles.

Interested area: knowing the functions and limitations of AI tools applicable to humanities lessons

Canva ai

Creating poster or infographic in a more efficient and all-rounded way 

e.g. creating different samples [good and bad] to better illustrate to students

It may cater class diversity e.g. a template with basic elements to start, enriched with more concepts to add on by students of different levels

-I think yes. For science teaching, it has Active Learning including: Inquiry-Based Learning,PBL, Collaborative Learning. And hands-on pedagogy.

one of my self-evaluation: which part in the teaching used the most of my time.

As a new teacher, I use the most of my time to plan the courses, which includes lesson plans, activities, and evaluation...

one of the most time-assuming task is activities since I need to consider more than 2 events and examine which is more appropriate in what way, in what class...

I uses the free version padlet and foundit could generate more than 5 plans with my prompt.

and for AI, I think we give some autnamy to get more efficiency.

Using MagicSchool AI in the educational setting has been a thought-provoking experience, offering both benefits and challenges. Here are some reflections on the use of this AI platform:

1. Personalized Learning

2. Efficiency and Time-Saving

3. Engagement and Motivation

Reflecting on the use of MagicSchool AI underscores the importance of leveraging technology to enhance learning experiences while upholding ethical standards, fostering inclusivity, and preserving the essential role of human educators in nurturing well-rounded and empowered learners.

Seattle Times. https://www.seattletimes.com/life/lifestyle/internet-creates-new-opportunities-for

cheating-but-also-learning/

Learning engagement - resources

T-chart document

Blose, A. (2023, April 12). As ChatGPT enters the classroom, teachers weigh pros and cons. NeaToday.

https://www.nea.org/nea-today/all-news-articles/chatgpt-enters-classroom-teachers-weigh-pros

and-cons

Faulstick, B. (2023, September 12). In the age of ChatGPT, what’s it like to be accused of cheating? Drexel

News. https://drexel.edu/news/archive/2023/September/ChatGPT-cheating-accusation-analysis

Sidoti, O. & Gottfried, J. (2023, November 16). About 1 in 5 US teens who’ve heard of ChatGPT have used

it for school work. Pew Research Center. https://www.pewresearch.org/short

reads/2023/11/16/about-1-in-5-us-teens-whove-heard-of-chatgpt-have-used-it-for-schoolwork/

Spector, C. (2023, October 31). What do AI chatbots really mean for students and cheating? Stanford

Graduate School of Education Blog. https://ed.stanford.edu/news/what-do-ai-chatbots-really-mean

students-and-cheatingUNESCO. (2023). Guidance for Generative AI in Education and Research.

https://unesdoc.unesco.org/ark:/48223/pf0000386693

Williamson, B. (2023, October-December). AI must be kept in check at school. The UNESCO Courier, p.6-8.

https://unesdoc.unesco.org/ark:/48223/pf0000387029_eng

Mollick, E. R., & Mollick, L. (2023, September) Assigning AI: Seven Approaches for Students, with Prompts.

http://dx.doi.org/10.2139/ssrn.4475995

AI Spectrum Placement

• Task: Rate yourself on a scale of 1 to 5 from “Completely Pro-AI” to “Completely Against AI in Student Learning.”

• Action: After placing a mark on the spectrum to represent your personal stance and then write 3 bullet points justifying your position.

Purpose: Helps clarify where you stand and why—compare with other group members.

Consider: 

• Are there inaccuracies or misrepresentations? 

• What would you add that’s missing? 

• What points speak to you the most? Why? 

1. Learning Goals Snapshot – Describe the knowledge and skills your students should confidently demonstrate by the end of the course. Think of this as the “dream outcome” for your learners.

2. Current Checkpoints – Explain how you currently measure and track student progress in these areas. (What tools, tasks, or assessments are you using?)

3. Assessment Wins – Reflect on the assessments that are giving you strong, reliable insights into how well students are developing their knowledge and skills. What’s working really well?

Assessment Challenges – Identify the areas where your current assessments need more development, refinement, or better use to truly capture student progress. Where could you invest more time, creativity, or energy?

European Commission: Directorate-General for Education, Youth, Sport and Culture".. (2022). Ethical

Guidelines on the Use of Artificial Intelligence (AI) and Data in Teaching and Learning for

Educators. Publications Office of the European Union. https://op.europa.eu/en/publication-detail/-

/publication/d81a0d54-5348-11ed-92ed-01aa75ed71a1/language-enEuropean Parliament. (2023, September 12). Artificial Intelligence Act: Deal on Comprehensive Rules for

Trustworthy AI. https://www.europarl.europa.eu/news/en/press-room/20231206IPR15699/artificial

intelligence-act-deal-on-comprehensive-rules-for-trustworthy-ai

UNESCO. (2023). Guidance for Generative AI in Education and Research.

https://unesdoc.unesco.org/ark:/48223/pf0000386693

United StatesDepartment of Education. Office of Educational Technology. (2023). Artificial Intelligence and

the Future of Teaching and Learning. Insights and Recommendations. Washington, D.C.

https://tech.ed.gov/ai-future-of-teaching-and-learning/

3 things I have learned:

2 questions:

1 challenge:

• What are the most important things you’ve learned during this workshop? 

• What are you most excited about regarding how AI can help in your classroom? 

• What are you most concerned about? 

• How do you hope to extend your learning now that you’ve completed this course? 

-I know some useful AI, but I have to spend some time to learn how to use them.

-I frequently use AI for preparing the lesson.

-Not much daily tasks were given to AI.

- algebraic concepts

- expansions and transformations

-  problem-solving skills

2. Current Checkpoints

- quick quizzes

- in-class exercise

3. Assessment Wins

- interactive

- provide a change of mode of study (from purely listening to giving outputs)

4. Assessment Challenges

- quizzes that are anonymous would not be able to provide me insights on each of their individual abilities

- I could try and make those quizzes anonymous among themselves, but I can access their responses individually

Students should understand the products of photosynthesis and be able to design a fair test to prove the occurrence of photosynthesis.

Current Point:

By reviewing their lab reports and correctly stating the variables of the experiment they designed.

Assessment Wins:

Engaging in activities like cooking instant noodles to learn how to conduct a fair test by investigating how to cook noodles to achieve a desired chewy texture.

Completely Pro AI in Student Learning

1. AI is everywhere; it’s unavoidable.

2. The key issue is how to use AI critically.

Instead of policing AI use, we should redesign assessments—focusing on process, critical thinking, and real-world skills. The goal isn’t to resist technology but to harness it responsibly. Clear guidelines, teacher-student collaboration, and ethical discussions can turn AI from a threat into an ally.

Ultimately, education’s value lies not in perfect outputs but in cultivating adaptable minds. AI won’t replace learning—but how we integrate it will define the future of fairness and innovation in schools.

i - the local HK edu authority does have some basic policies on the recommended usage of ai in the education sector

ii - we have a number of departments who are taking a forward-thinking approach to teaching the above to our junior secondary students

iii - I have been doing something right with the related policies in my lesson planning

2 questions

i - who need to spearhead the implementation such policies?

ii - how should such policies be monitored and by whom?

1 challenge

i - consistency of such implementation

1) understand the policy, and guide students to refer the policy

2) understand the limitation of AI

3) stress that use in a proper way( citation etc)

2:

1) How to assess whether students follow the policy

2) how to set the rules about AI in class to MYP studnets

1:

How to choose AI for different purpose and how to check the Authenticity

2 - Ethics, Policies

1 - How can school continuously develop AI policies to suit the school, educator and student needs?

-Pay attention to the the policies (International, Regional, School)
-Usage of AI in the future
-Keep learning the AI tech

2 Questions:

I have -How to find the best AI for teaching?
-How to use more AI intensionally?

1 Thing I have challenge

-Design the assessment by using AI

3 things learnt: 

-the analogy of AI support, parent support and tutor support

-the similarities between the emergence of Google search and generative AI

-the framework of AI Literacy

2 questions

-What are the steps to create a school-wide approach in using AI in student learning to cater to personalised learning?

-What education programmes can be provided for parents about using AI in student learning? 

1 challenge

-The school-wide approach to using AI in student learning will not be built in one day.

My key takeaways are:

1. Accessibility is a Foundation, Not a Feature: The question "Is the system accessible by everyone without barriers?" is a powerful litmus test. It immediately shifts the focus from mere functionality to universal design. It forces us to consider the full spectrum of human experience—ability, language, socioeconomic status, and age. True innovation isn't just about what the AI can do; it's about ensuring that everyone can benefit from it.

2. Fairness is a Process, Not a Destination: I’ve moved from seeing bias as a static "flaw" to be fixed, to understanding it as a dynamic risk to be continuously managed. The most important procedures aren't just the technical ones for detecting bias in data, but the human-centric ones: feedback mechanisms, ethical review boards, and transparent documentation. This is not a one-time checklist but a cycle of auditing, learning, and adapting.

3. Respect is Designed, Not Assumed: The question of whether an AI treats learners respectfully by adapting to their individual needs struck a chord. It frames the goal not as creating a smart tool, but as creating a respectful digital partner in learning. This means building systems that are humble enough to acknowledge their limitations and flexible enough to serve the unique human in front of them.

4. My Role is Critical: Whether as a developer, educator, policymaker, or simply an informed user, this material underscores my responsibility. I must be a critical consumer and a vocal advocate. I can ask tougher questions of technology providers, provide clearer feedback on biased outcomes, and always centre the needs of the most vulnerable learners in any discussion about AI.

This study has ultimately reinforced that the goal of educational AI isn't to create the most powerful algorithm, but to foster the most empowering and inclusive learning environment. The real measure of success won't be the model's accuracy, but its equity.

Most excited: some new AI platform

most concern: how should education shift to fit the new era of AI.

The application of AI in subjects area

2) The most excited part about using AI: How AI can enhance teachers' efficiency in lesson planning, teaching and planning; to look for time to explore various AI tools that are relevant to my subjects and year levels

3) Most concerned: academic integrity and how to sustain our critical thinking when using AI

Over the past two days, I have immersed myself in learning about AI in education—from understanding its safety and data protection to exploring its practical uses, benefits, and challenges in the classroom. This workshop has not only deepened my knowledge but also strengthened my connection with the team, creating a supportive community eager to innovate together.

I am genuinely excited about integrating AI tools in my classes this upcoming academic year. The possibilities for enhancing student engagement and personalized learning feel inspiring. At the same time, I remain aware of my students’ concerns about the potential overuse of AI, especially regarding academic honesty. Addressing these worries thoughtfully will be crucial to fostering responsible and ethical AI use.

Moving forward, I want to continue expanding my AI expertise through ongoing professional development, collaboration, and hands-on experimentation. Staying curious and open will help me harness AI’s full potential while navigating its challenges wisely.

Here’s to a year of learning, growth, and meaningful innovation!

Best,
Your future self

peer discussion and exploration of different AI tools

most excited: experience from other colleagues

• the ethical issues of AI

• the implementation of AI in teaching and learning

What are you most excited about regarding how AI can help in your classroom?

• increase students' engagement in classes

• have students in control of their learning

• reduce the differences between students

What are you most concerned about?

• time management - time might be spent on getting use to the technology, instead of the knowledge we intended to teach in that lesson

How do you hope to extend your learning now that you've completed this course?

• mathematics is an abstract subject to teach, master and find fun within. For those who despise maths, AI might be able to change their perspective.