SPEECH BY MR TAN KIAT HOW, SENIOR MINISTER OF STATE, MINISTRY OF DIGITAL DEVELOPMENT AND INFORMATION & MINISTRY OF HEALTH, AT THE SINGHEALTH DUKE-NUS EDUCATION CONFERENCE, 25 SEPTEMBER 2026

Mr Cheng Wai Keung, Chairman, SingHealth

Professor Fong Kok Yong, Deputy Group Chief Executive Officer (Medical and Clinical Services), SingHealth

Associate Professor Shiva Sarraf-Yazdi, Vice Dean of Education, Duke-NUS Medical School

Professor Fernando Bello and Clinical Assistant Professor Moy Wai Lun, Co-organising Chairpersons of the Conference

Ladies and gentlemen,

1.       A very good morning. This year’s theme, “Learning Reimagined: Education for Tomorrow’s World”, is especially timely.

2.       Healthcare is changing rapidly. The professionals we train today will work in a world very different from the one many of us entered. They will need strong clinical foundations. But they will also need to work across disciplines, adapt to new models of care, use technology intelligently, and keep learning throughout their careers. We cannot simply prepare people for the healthcare system we have today. We have to prepare them to keep learning as the system changes.

Learning beyond professional silos

3.       Let me begin with something that will remain important whatever happens with technology. There is a lot of excitement around artificial intelligence, AI. There is a lot of excitement about what it can do, but there is a lot of concerns what it might bring as well. So let me talk about something that will remain important, whatever happens with AI and the developments around it. Patients do not experience healthcare as a collection of separate professions. Take an older patient managing several chronic conditions - perhaps diabetes, heart disease and declining mobility. The patient may see a GP, a specialist, a nurse, a pharmacist, a therapist, a medical social worker and community care providers. To the patient, this is one journey. No single profession can provide all the care that this patient needs. So, we need professionals who are excellent within their own disciplines, but who can also work effectively across them.

4.       Interprofessional education is not simply about learning alongside one another. It is about understanding one another’s roles, communicating clearly, coordinating decisions and working together around the patient. I am encouraged by the partnership between this conference and the Asia Pacific Interprofessional Education and Collaboration Network, or APIPEC. The challenges facing healthcare education do not stop at our borders, and neither should our solutions.

5.       Closer to home, SingHealth is strengthening this work through the Centre for InterProfessional Healthcare Education (CIPHE), with programmes such as Educating Health Professionals in Interprofessional Care (ehpic™) and Virtual Interprofessional Teaching and Learning Programme (VITAL) helping professionals learn from one another and from colleagues in Singapore and abroad.

Learning without boundaries

6.       Technology is also changing where and how learning happens. The SingHealth Marketplace is one example. It builds on the SingHealth Learning Management System and will allow healthcare professionals to discover and access programmes developed by SingHealth educators, with the potential to extend learning and knowledge exchange across Singapore and the region. I think this points to an important principle: Learning should follow the professional, not end when the professional leaves medical school. Whether someone is a junior doctor, an experienced nurse, an allied health professional or a senior clinical leader, the need to learn does not disappear.

Reimagining learning in the AI era

7.       But there is another development that will challenge our assumptions about healthcare education even more profoundly. I spoke aboutartificial intelligence earlier.

8.       AI is already changing healthcare practice. It can analyse information, support clinical decisions, summarise records and reduce administrative work. Increasingly, AI will do more than answer our questions. It will help coordinate workflows, synthesise information across systems and perform multi-step tasks. So, the question for educators is not simply, how do we teach healthcare professionals to use AI? That is basic. We need to ask a deeper question: How should we teach healthcare professionals when AI changes the way healthcare itself is practised? Because AI changes not only how people work. It changes how they learn to work.

9.       A doctor does not become a good doctor simply by memorising facts. Competence develops through doing - taking a history, examining a patient, forming a differential diagnosis, discussing a case with a senior, making a judgement, sometimes getting it wrong, learning why, and doing better the next time, and I think that is the heart of what Dr. Aliaga talked about just now on error-based learning. Some of this effort is necessary, and I like what the earlier speaker shared about struggling. Sometimes struggles are taught. Some of this friction is actually how expertise and judgement are built.

Wasteful friction and meaningful friction

10.      Let me make a distinction. I think we need to distinguish between two kinds of friction. I call it frictions, but Dr Aliaga called it struggle. In every industry, every workplace, and certainly in healthcare as well, there is what I call wasteful friction. It is the work that consumes time and attention without creating much value: repeated data entry, duplicative documentation, unnecessary administrative work, searching for information that should already be available. We should remove this friction wherever we can, increasingly with more capable AI.

11.      But there is also meaningful friction. The effort involved in working through a difficult case. Forming an initial judgement. Explaining one’s reasoning. Encountering something unexpected. Receiving feedback. Making a mistake safely and learning from it. That effort may be slower. But sometimes the effort is the learning. So the principle should be: Remove wasteful friction and preserve meaningful friction. Or, I put another way: Automate the task, not the learning.

12.      This matters because there are different risks. You can become deskilled because you no longer practise a capability. You can become mis-skilled because you internalise something incorrect from AI. And you can become never-skilled because technology arrives before you have properly developed the underlying competence. These are not arguments against AI. They are arguments for better education.

Learning without AI, with AI, and through AI

13.      I am not a medical professional, and I am not clinically trained. But my background is technology, and I spend a lot of time working with engineers in technology, even frontier engineers from frontier technology companies, and I see this happening already. Software engineers who use AI for coding, for debugging, for problem solving, very good engineers, young people, fresh out of top schools, but they did not develop the competence, experience, and judgment needed to solve complex engineering problems when they become more senior. Because they took too much AI to automate too many things, and you lose the precious friction that is formative. So I think we should use AI thoughtfully, and it has to start in education, and has to start in learning.

14.      Perhaps we can think about it in three stages. First, learning without AI. Learners need enough foundational knowledge and experience to understand the ground truth for themselves. Second, learning with AI. AI can become a validator and also challenger - helping learners test their thinking, identify blind spots and recognise when the AI may be wrong. Third, learning through AI. Once the foundation is secure, AI can help learners synthesise information, work through more complex problems and take on new forms of work.

15.      The goal is therefore not to keep AI out of education. Nor is it to let AI take over education. The goal is to use AI to make people better learners and better professionals.

Simulation becomes even more important

16.      And that is where SingHealth’s work in simulation becomes particularly interesting. Simulation already helps healthcare professionals develop clinical, communication and teamwork skills in realistic and safe environments. It can also help test workflows before they reach the real world.

17.      The SingHealth Duke-NUS Institute of Medical Simulation, for example, worked with Singapore General Hospital to simulate stroke activations, trauma workflows and emergency transfers when planning its new Emergency Department.

18.      As AI becomes more capable - and more agentic - simulation can become even more important. Imagine an AI agent coordinating several steps in a clinical workflow. Before deploying it widely, we should be able to ask: Where does it help? Where does it fail? Where should it escalate? What happens when the data are incomplete? What happens when a clinician disagrees with it? And most importantly, how does the human-AI team perform together? Simulation can therefore become more than a place to train people. It can become a place to test the systems people will work with.

From education to capability

19.      So when we talk about reimagining learning, I think we should think beyond courses and classrooms. We want professionals who can work across disciplines, use AI critically, recognise its limitations, exercise judgement under uncertainty and adapt as technology and models of care evolve.

20.      The measure of good education is not simply how much information we can put into a learner. It is how capable do we make that learner of learning, adapting and exercising judgement. And this is a shared responsibility. Educators have to rethink curricula. Clinical leaders have to create environments where people can learn safely. Technology teams have to build tools that support learning rather than undermine it. And for institutions, you have to create the time and space for professionals to keep developing.

Recognising our educators

21.      As technology and systems advance, it is still our people who give them purpose. Healthcare education remains a deeply human endeavour. This morning, I am honoured to congratulate the 74 recipients of the SingHealth Duke-NUS Academic Medicine Education Institute Golden Apple Awards 2026.

22.      Each recipient has contributed in different ways, but all share a commitment to helping our healthcare professionals learn, grow and provide better care. To every awardee, my heartfelt congratulations. Your work reminds us that even in a world of increasingly capable technology, people remain at the centre of healthcare, and educators remain central to developing the people who care for us.

Closing

23.      So in conclusion, let me end where we began. “Learning Reimagined” is not simply about using new technologies in education. It is about preparing healthcare professionals for a world in which the technology around them will continue to change. We want professionals who are confident enough to use AI, critical enough to challenge it, disciplined enough to know when not to rely on it, and grounded enough to remain accountable to their patients. We want learning that crosses professional boundaries. Learning that follows people throughout their careers. Learning that is supported by technology, but not determined by it. And learning that helps our healthcare system become better at learning itself.

24.      Because the healthcare system of tomorrow will need not just better technology. It will need people who are better equipped to work, learn and adapt with it. That, ladies and gentlemen, to me, is what learning reimagined should mean for healthcare. Thank you very much.