HUMI Lab

Human-Centered Machine Intelligence

We study machine intelligence in human contexts — language, intelligent agents, and digital health — at Reykjavík University, with a recurring focus on Icelandic and low-resource settings.

Research

We study intelligent systems that are interpretable, trustworthy, and aligned with human values in critical contexts, across four connected areas.

Natural Language Processing & Language Technology

Language models and NLP for real-world text, with a focus on Icelandic and low-resource settings.

Conversational & Intelligent Agents

Agents that interact through natural language — counseling, behavior-change, and virtual agents — designed for expressivity, safety, and trust.

Digital Health & Wellbeing

Human-centered AI for healthcare and mental health: triage, screening, self-management, and digital interventions.

Human-Centered & Trustworthy AI

Making intelligent systems interpretable, dependable, and genuinely complementary to people.

4

Research areas

27

Theses supervised

50+

Alumni

Mission

To advance machine intelligence as a human-centered scientific discipline — building systems that engage with people and their language, and studying how they behave once deployed in real human settings.

Vision

Intelligent systems that are transparent, trustworthy, and aligned with human needs — supporting people's wellbeing, decisions, and autonomy rather than replacing or obscuring them.

Interested in working with us?

We welcome collaborators, prospective students, and visiting researchers.

Get in touch