Educators Forum: Think Beyond AI

[17:05] Referencing yesterday's top IB performer panel: what actually set the winning students apart was not just intense effort - it was finding balance. They still had time to study, keep a social life, and explore their own interests, rather than sacrificing everything else for exam prep.

[17:04] Ellen Lau's closing takeaway for parents, three things: 1) Take your time - do not panic. 2) Be present with them - work through judgment calls together, including questions neither of you has the answer to. 3) Give them exposure - let them see and try many different things.

[17:03] A practical example for parents: pick an age-appropriate AI news story and explore it together - for instance, scientists using AI to engineer a virus that kills bacteria, and asking "what do you think, does that scare you?" Discussing AI topics and ethics together, alongside AI literacy outside school, is how you pass on the right values. Being present alongside them matters more than having the answers.

[17:03] Samuel Chan, completing the point: with solid foundational knowledge, students should pursue questions that do not have answers yet, with no subject boundaries - chasing what they genuinely want to understand more of, not chasing what the job market currently wants.

[17:02] On "take your time": Hong Kong's average lifespan is now in the 80s, and over half will live even longer than that - there is a lot of time to explore different things. That is a luxury worth holding onto, not giving up for the sake of rushing.

[17:02] Samuel Chan, continuing: genuine interest is what sparks curiosity, which drives students to explore more and stay interested in learning. AI can already do a lot, so what matters most going forward is doing and imagining things AI has not discovered yet.

[17:02] Samuel Chan: genuine interest is what drives effort - a student who actually cares about something will grind at it tirelessly without complaint. Do not let students pick a subject just because "the job market requires it."

[17:01] Ricky Lam, closing: AI can still play an important role in helping students make judgment calls, even as they stay in the driver's seat of their own decisions.

[17:01] Advice to parents: encourage diversity and exploration. The older generation's parents thought carefully about steering kids toward one fixed path, but today every path can lead somewhere - what matters most now is encouraging students to expose themselves to different people and different activities.

[17:00] On parents's anxiety about kids and AI: the advice was three words - "take your time." Do not rush it. Let kids explore whatever they are drawn to on their own; the most important job is nurturing their curiosity so they can develop in that direction themselves.

[16:59] John Tsang, continuing: what university actually gives you is social, experiential learning - but does that still require the traditional 4-5 year model? A lot of the content itself could move online, without everything having to become fully online.

[16:58] John Tsang: over a decade ago MIT made its courses free for anyone, anywhere in the world to learn from - as long as you have the curiosity to actually do it - and Harvard has recently done the same. That raises a question worth sitting with: do we still need universities at all?

[16:58] Ellen Lau asked John Tsang: does this approach (keeping curiosity alive, redesigning assessment, teacher-student alignment on AI use) actually get us to where we want students to be - curious and genuinely learning well?

[16:58] Ricky Lam noted that this level of degree-focused education is still fairly limited, and flagged potential future collaboration with Preface specifically in this space.

[16:56] Ricky Lam: the lab is built so students get real projects out of it, not just theory - good learning is not sitting through lectures and tutorials, it is workshops that let students graduate with an actual portfolio of projects they can point to and say "I worked on this."

[16:56] Ricky Lam: the business school's framing is not "how do we deploy AI" - it is "our students solve real business problems." That is why they built a business co-creation / AI co-creation lab connecting students, teachers, alumni and industry, where real industry problems are brought into the lab alongside a range of technical partners.

[16:55] Ellen Lau asked Ricky Lam directly: on the university side, what has actually been done in this space?

[16:54] Samuel Chan: parents often do not understand what their kids are actually doing on the computer for hours - "why are you always in front of a screen, what are you even doing" - and then compare them anxiously to "someone else's kid who is great at AI, why can't you learn anything." That confusion and comparison anxiety is exactly why parents need their own sessions, not just staff and students.

[16:54] Samuel Chan: schools have already run staff development sessions on AI and students have plenty of opportunities to apply it - the group actually being left behind in this conversation is parents. More seminars aimed specifically at parents were flagged as a real gap worth closing.

[16:53] Dr Lam Chun Kit: guidelines also need to cover how teachers should respond when a student misuses AI - because it is new, many staff will not instinctively know how to handle it. Framing helps: AI-enabled bullying is fundamentally the same problem as social-media bullying, and schools need to communicate that clearly rather than leaving teachers to figure it out alone. A clear, shared AI usage policy is what gets every stakeholder aligned on how to actually follow through.

[16:53] Dr Lam Chun Kit: clear rules are needed on both sides - students need explicit guidelines on what AI should not be used for (like manipulating someone else's photo), and teachers need their own guidelines plus proper training on how to bring AI into teaching, including clear guidance on how to handle homework that students produced with AI help.

[16:52] Dr Lam Chun Kit on how schools are actually adapting: it works on two levels, top-down and bottom-up. From the top, leadership first needs to genuinely understand what AI can and cannot do, then set a shared definition and agreed guiding principles for how staff and students use it - including clear standards, like: if you use AI to produce something, you are still responsible for that output.

[16:51] John Tsang: what subject a student studies barely matters, because every subject is a valid starting point - nothing in the world today isn't interdisciplinary, and any genuine interest can branch out into other fields. It all starts with curiosity.

[16:50] John Tsang: AI is the same kind of tool as the slide rule, then the HP calculator, then the graphing calculator he used decades ago - each felt like a leap, but you still had to understand what it was doing underneath. What matters is not the tool, it is how you use it, and today that means schools should shift focus toward soft skills: curiosity about life and learning, and the ability to pursue what genuinely interests you.

[16:49] Moderator Ellen Lau opened the discussion by asking John Tsang, who has witnessed many education reforms in Hong Kong: how should he see this AI-generation wave of change, and where should schools actually go from here?

[16:48] John Tsang: the real goal at the philosophy and goal level of a school should be teaching students that learning never stops - keeping their curiosity alive and building the ability to go beyond whatever stage AI is currently at, toward whatever comes next. How to actually do that in practice is the harder question worth sitting with.

[16:47] John Tsang: Hong Kong education still runs on "19th-century curriculum, 20th-century teachers, 21st-century students" - AI is not the destination, just the current chapter of a process that keeps moving (AI itself looked nothing like this 6 months ago, and will look different again in 6 months). Chasing the current version of AI misses the point.

[16:47] Ricky Lam: many students already admit to using AI to write their essays for them - the real challenge may be tightening how we assess students, not banning the tool: require a presentation where students have to defend and explain their own work, rather than just accepting a submitted document at face value.

[16:46] Ricky Lam: the current worry is students are already using AI but cannot yet judge whether its output is right or wrong, so basic technical grounding still has to be maintained. If a university's role were just transmitting knowledge, AI already does that better than we do - the real job left for education is teaching the next generation how to make decisions.

[16:44] Ricky Lam: if teachers now have time freed up from marking, that time is better spent teaching students AI's limits and risks, and how to be a responsible decision-maker when using AI - because using AI is still fundamentally about being human, and that "AI literacy" matters more than the tool itself.

[16:43] Dr Lam Chun Kit: some schools initially discouraged students from using AI at all, worried it would just do their homework for them with nothing sinking in - the real fix is not banning AI, it is redesigning homework: is the task something you can solve with one button press, or does it still demand thinking and revision before it is submittable?

[16:43] Ricky Lam (Hang Seng University): being a private university gives more room to experiment, but the real question is not how fast to adopt AI - it is what education is fundamentally for. The university's own focus is less "use AI" and more "use AI well," including the ethics of it.

[16:42] The real challenge is balance: teachers need to learn to use AI in a way that boosts efficiency without losing their professional judgement, and students need to use AI to amplify their learning without losing the essence of learning itself - getting that balance right is genuinely difficult, especially for teachers who trained before AI existed and now have to adapt on the job.

[16:40] If a student skips idea refinement - the back-and-forth of weighing different arguments - and hands writing straight to AI, that iterative step is exactly what is missing from their brain development. The same applies to teachers: if AI takes over entirely, do we still truly know our students, or are we just reading a report instead of learning their needs through real contact with them?

[16:40] Ellen Lau: a recent workshop with educators surfaced the real friction points of bringing AI into schools - is AI in education really as bright and easy to roll out as the efficiency and personalisation talk suggests? The core worry is offloading, handing part of the work of learning to AI, which raises a deeper question: what does learning actually require from the body and brain in the first place?

[16:39] Samuel Chan: freeing teachers from repetitive grading is not just about speed - the real value is giving them back time to rest or to genuinely reflect on how to teach a topic better next time, rather than mechanically ticking through papers and walking away with only a vague impression.

[16:37] Samuel Chan: AI-assisted marking now surfaces data, not just impression - instead of a vague sense that a question "went okay," teachers can see exactly what percentage of students got a specific question wrong and focus teaching time there, the same efficiency gain some government departments are seeing (two weeks of work down to two days).

[16:37] Dr Lam Chun Kit: AI is emerging as a learning partner for students because it solves a real pain point in traditional teaching - it is timely (instant feedback, unlike a teacher marking 30 students' work days later) and personalised to each student's needs and current level, giving direction to keep learning momentum going.

[16:36] John Tsang: every new technology carries both risk and opportunity - the car replaced the horse-drawn carriage and wiped out jobs like coachmen and stable hands, but it also created new ones (mechanics, road building), and enabled shifts like suburban living and roadside advertising. AI's disruption likely follows the same double-edged pattern.