Human capability is the outcome
The measure of learning technology is not how much it can do. It is what the learner becomes more capable of doing because of it.
Our origin
Teaching and learning shaped our research direction.
Experience as a teacher and as an online and offline learner revealed gaps in feedback, assessment, and opportunities to ask questions. Exploring those gaps through a personal project became the starting point for SILT.
Before SILT
Learning needs a conversation.
My experience as a teacher and as a student in both offline and online settings shaped the questions behind SILT. Seeing learning from both sides brought the conditions that support learners into focus: useful feedback, meaningful assessment, and opportunities to ask questions.
Across those settings, I saw a gap between having learning materials and having responsive support. Feedback could help someone decide what to practise next. Assessment could guide the development of a skill. A question could open a conversation that a course or resource alone could not provide.
I began exploring these possibilities by developing and refining a personal project around the feedback, assessment, and questioning I wanted in a learning environment. That exploration connected practical design choices to wider questions about how people learn.
How can learning environments make useful feedback, meaningful assessment, and room for questions more available?
The role of AI
AI changed the question.
As generative AI became widely available, learning changed almost overnight. Students could generate essays, explanations, code, answers, summaries, and solutions in seconds.
That capability was extraordinary. It also raised a different concern: if technology can increasingly do the work for us, how do we make sure people are still becoming more capable themselves?
How do we know whether someone understands?How do we preserve the effort that makes learning durable?When should AI help, question, or step away?What should remain human?
A teaching perspective
Feedback is part of the learning environment.
Teaching had shown me the importance of responding to each learner, and the practical demands of doing so. Learners do not arrive with the same prior knowledge, confidence, pace, interests, or ways of making sense of new ideas.
A teacher can use questions, conversation, and assessment to shape the next step in learning, but sustaining that attention for every learner takes time and support. Young people were also entering an AI-shaped world without necessarily being given safe, structured ways to learn how to use these systems well.
The concern was not simply that students would use AI. They inevitably would. The concern was whether they would learn to use it in ways that strengthened understanding, judgment, independence, communication, curiosity, creativity, and responsibility.
From thesis question to research direction
A thesis began the exploration.
SILT gave it a wider purpose.
My final-year thesis explored adaptive learning, feedback, assessment, Socratic questioning, and active recall. It opened a wider inquiry into how a learning system could respond to a learner, guide practice, and support the development of understanding and skill.
The thesis surfaced questions that technical possibility and product intuition could not responsibly answer alone. What kind of feedback improves understanding? How should AI adapt without reducing learners to fixed profiles? How much context should a learning system remember? What should remain outside its authority?
Pursuing those questions brought learning science and engineering together. SILT grew from that direction as an interdisciplinary research team.
Experience in teaching and learning revealed gaps in support.
A thesis gave us a way to explore them.
The questions it opened became a research direction in learning science and engineering.
Why SILT exists
As technology becomes more capable, how do we ensure humans become more capable alongside it?
SILT studies how learning should evolve alongside technology. We are not starting from the assumption that every new technology belongs in education. We are interested in when technology improves learning, how it should be designed, what boundaries it requires, and what human capabilities it should strengthen.
Explore the researchWhat we believe
The principles beneath the work.
Understanding matters more than output
Producing the right answer is not the same as understanding. We care about retrieval, explanation, application, questioning, transfer, and independent reasoning.
Technology should augment people
Teachers, mentors, peers, families, and communities remain part of how people learn. Technology should expand their capacity, not quietly erase their roles.
Personalization should expand potential
Support can adapt to where a learner is without turning that moment into a permanent label. Start where the learner is. Do not leave them there.
Productive effort is worth protecting
Difficulty is not always a design failure. Some effort is necessary for memory, reasoning, confidence, and independence.
Human judgment remains where consequences matter
Technology may observe, summarize, recommend, and adapt. Decisions with lasting consequences require context, responsibility, and accountability.
Privacy is part of learning design
Personalization should not require unlimited memory or surveillance. We are interested in the minimum useful context required to support a learner well.
Learning stays connected to life
Learning happens through conversation, making, teaching, movement, culture, places, failure, and experience. Technology should connect people back to the world.
