AffectiveAI Research

Research for a more governable relationship between people and intelligence.

We develop hypotheses, specifications and evaluation frameworks for the operational state of human–AI interaction: context, constraints, oversight and evidence.

Deploying intelligent systems requires more than capability: it requires explicit conditions of responsibility, coherence and control.

Research programme

We study the state of interaction, not the person.

01

Context

How an interaction is situated in a moment, a domain, a set of rules and a shared history.

02

Governance

How systems may operate with declared constraints, oversight and mechanisms for intervention.

03

Evidence

How a consequential decision may retain a reviewable trace without turning people into scoring objects.

Research is under development and validation. We do not infer private mental states, diagnose or carry out social scoring.

The failure does not exist in the isolated model. It exists only in the interaction.
A proposition for discussion, not a claim of completed validation.

Institutional relationship

The science that makes more governable infrastructure possible.

AffectiveAI Research is the scientific, philosophical and human foundation of tuucke.

While this programme investigates how intelligence should relate to people, tuucke builds infrastructure that governs how intelligence is allowed to act.

Meet tuucke →