Recherche
Research for adaptive intelligence.
Biology constrains the starting state. Culture carries accumulated knowledge. Individual experience changes what happens next. Embreier studies how to represent those changes rigorously enough for a model to learn from them.
Biology · Culture · Individual experience · Learning
The direction
People change. So does the system around them.
Kalyr represents the state each decision needs, under explicit permission. The research looks further. Biological, cognitive, behavioural, social and cultural factors may shape how a person acts and develops over time.
07 returns as the new 03Next action starts from a changed person
The research aim is to represent those interactions while every person stays a whole person and every inference stays marked as inference.
A governed research programme
What the picture gains
when people are part of it.
Human-development science supplies candidate variables and priors. Each must earn its place by improving a real decision.
Decompose how people change.
Break human development into states, processes and changes over time, so each candidate variable has a place, a unit and a timescale.
Hold what the fields share.
Neuroscience, biology, developmental and applied psychology, pedagogy and behavioural science each describe how people change. Where they converge, we hold a candidate prior. Where they diverge, we hold a question.
Culture and universals.
Most of what a person knows was worked out with others. The research separates what holds across people from what a team or a culture carries.
Every inference, marked.
A person is partly observable. Any estimate of task-relevant state would carry its uncertainty and an expiry, stay apart from the evidence it came from, and stay open to that person’s correction.
Support that fits the moment.
How what a person has seen and tried on the task could inform which support helps them most: a cue, a question, an explanation or room to work.
Learning in both directions.
How a system could learn which support works for a team, and how people build capability through the work. Both directions are studied.
Immediate and longer-term outcomes.
Where a person is materially involved and governance permits it, the same grammar could distinguish immediate system outcomes from longer-term human outcomes. Performance today and capability later stay separate measures.
Every candidate variable has a simpler proxy to beat and a stated result that would retire it.
The testbed
Somewhere a prediction can be scored.
A developmental claim becomes testable where state is observable, action is consequential and the outcome arrives while it still matters. Real, recurring processes could give all three. The research runs as its own governed programme, apart from Kalyr the product, under its own permissions.
Evaluation
Measure what actually improves.
Same problem. Same data. Same input budget. Measure the difference.
The same process evidence, the same tools and the same memory budget.
A static role, task or team profile added to A.
A time-bound estimate of the person or team state, used where governance permits it, with support chosen to fit.
The test: does C add predictive value over B, net of the burden on people and the cost of governing it? The primary measure is fixed before any data is seen. Calibration, retention and transfer are measured separately, on future time and new sites.
Longer-term outcomes
Measure what lasts.
Four longer-term outcomes, each to be measured apart from the immediate result, where governance permits. Each could move while immediate performance holds steady.
0 / 7
Research areas
The work behind adaptive intelligence.
Human development is the foundation. State, timing and improvement build outward from it.
Foundation
How people learn and adapt
The research framework for representing human learning and adaptation, with observation, inference and judgement each kept in its own place.
Systems
What belongs in the picture, what counts as evidence, and how uncertainty stays visible.
Which move helps, when to ask instead, and when the right move is to watch.
Retention, judgement, transfer and adaptive expertise, each measured on its own.
Research with us
Study how people and systems change together.
Every person stays a whole person. Every inference stays marked as inference. Permission and uncertainty stay explicit, from the first question to the last result.
We work with researchers across AI, neuroscience, biology, human development, behavioural science, human factors, philosophy, ethics and HCI.
Research with us →