How people learn — in rooms, on screens, and around a table
Designed structure shapes possibility.
Rooms, interfaces, rules, tools, representations, and systems influence what people notice, what they can do, and how they think.
I work across learning, technology, design, and systems to understand — and sometimes redesign — those conditions.
shapes possibility
My work has moved from classrooms to schools, curriculum, digital platforms, learning environments, and system-level questions.
The scale changes. The underlying interest does not: how structures shape learning, thinking, behaviour, and possibility.
- Learner
- Classroom
- School
- Curriculum
- Platform
- System
Orientation
If you came here for…
Most people arrive in the middle, from a search or a link. These are the fastest ways in.
- How people learnLearning science, cognition, feedback, task design, and the conditions that support durable learning.Explore learning
- How educational products are designedResearch, product thinking, interaction design, accessibility, experimentation, analytics, and implementation.Explore the craft guides
- How spaces shape learningLearning environments considered as systems of participation, visibility, movement, interaction, and control.Explore learning spaces
- How systems changeComplexity, futures thinking, systems reasoning, implementation, and working under uncertainty.Explore the toolkit
- Something useful on MondayOperational guides, protocols, frameworks, and practical methods.Browse guides
- Something to argue withArguments, propositions, questions, and positions that are still being worked through.Read the notes
- Sources worth keepingBooks, papers, frameworks, definitions, and references that have shaped the work.Browse the library
- The person behind the workBackground, pursuits, making, games, and other influences.About Gerald
New here? How this site works explains the kinds of page and how they connect.
Featured idea
The designed conditions of learning
A learner never encounters curriculum, technology, space, assessment, and social interaction separately.
These conditions operate together.
I am interested in what becomes possible when educational design begins with that larger unit.
Explore the frameworkThe learnermeets them at once
- Cognitive
- Digital
- Spatial
- Social
- Institutional
- Representational
- Temporal
- Affective
Still open
Questions I keep returning to
I have not settled these. Pick one and argue with it.
What should educational technology optimise for if engagement is not learning?
Argued in Evidence of learningWhat can be standardised without suppressing professional judgement?
Argued in A design system is not a component libraryWhich decisions should AI make, which should it support, and which should remain deliberately human?
Argued in AI in learning products (opens in a new tab)
Guides
Educational technology as a craft
Educational technology is often treated as the intersection of two areas: education and technology. In practice the work needs a much broader set of literacies. The craft map lays them out, and the pathways give four ways through the guides.
§04 — Pursuits
Things I do for their own sake
Not hobbies in the sense of filling time. They are where I get to be a beginner, and where the feedback is immediate and unsentimental. Most of what I believe about designed experience I first learned at a table, losing.
Mahjong
Singapore mahjong is a scoring system pretending to be a tile game. A game held together almost entirely by undocumented convention is a small, complete study in tacit knowledge — and a decent argument for why documentation is design work.
mahjong.geraldajam.comContract bridge
The best-designed information problem I know of that you can finish in an evening. Bidding is a constrained language — a small set of legal utterances, all public, all ambiguous — and both partners have to build a shared picture using only those, while two opponents listen in.
Cartoons
A single panel has to carry setup, character, and turn simultaneously, which makes it the most unforgiving compression problem in visual communication — and the closest thing I know to a diagram with a sense of humour.
Trying new tech
I poke at new tools by building something small and slightly too ambitious with them, on the theory that you cannot evaluate a tool from its documentation. Most attempts do not survive, which is the point — the fast, cheap failure is the finding.
Making espresso
Espresso is a process-control problem you can run twice before breakfast. Everything that matters happens inside an opaque basket in twenty-five seconds, so you are never observing the process — only its output, and only after it is too late to change anything.
§05 — Reference
From the library
Where the ideas came from, and a way in when you know the problem but not the word for it.
Learning science
Every other shelf here assumes an answer to how people learn. This one makes that answer explicit, so the rest can be argued with.
Start from the problem
The glossary begins with a term. The concept index begins with a problem. Use it when you know what is happening but do not yet know which concepts might help explain it.
The craft glossary
The vocabulary of making a learning product, from abandonment to zero-downtime release. Each term links to the place in its guide where it is referred to.
§01 — About
Atlas & Loom
Atlas
Read the terrain. The scale changes; the question does not.
System. Policy, implementation, capability, complexity, and long-term change.
The designed conditions of learningDomainLoom
Cross two threads and something gets made.
C1 · Learning × Making.
Maker educationLibrary shelf