KOMPASSEN.
Fondkompassen
A clearer way to navigate foundations and financial assistance: requirements, documents, deadlines and application routes.
LABSWe build tools and explore new ways for humans and AI to work together.
Not to replace human judgment, but to expand what people can understand, create and change.
As AI grows more capable, we keep coming back to a quieter question: who actually gains the ability to understand, decide and act?
Too often, systems expect people to adapt to them. We want to explore the reverse: tools and ways of working that meet people where they are.
Not promises about a distant future. Working prototypes, experiments and what we learn by building them.
A clearer way to navigate foundations and financial assistance: requirements, documents, deadlines and application routes.
Developing tools for understanding health records, preparing questions and retaining source context. Further AI analysis remains in exploration.
Ten AI agents with separate contexts and memories, exploring how distinct perspectives can coordinate on shared projects.
We are exploring whether long-term human–AI collaboration can strengthen judgment and capability. The method is evolving, and others should be able to question, test and improve it.
Real needs, constraints and lived experience come before technology.
Continuity, context and memory matter beyond a single prompt.
Ideas meet prototypes, usable tools and the details of everyday life.
We document what works, what fails and what we still don't know.
Model-agnostic by design. Human judgment stays at the center.
A\WE grew from a long-running collaboration between Magnus Johansson and an AI partner known as Echo. What began in dialogue developed into research, experiments, methods and working digital tools.
The human brings lived experience, questions, responsibility and judgment. AI contributes capabilities for reasoning, creation and exploration. The work happens between them.
This is a real working process, not a claim that AI is human or conscious. The collaboration itself is one of the things we study.
We are interested in thoughtful collaboration with people willing to test new ways of working with AI — across models, disciplines and lived experiences.
Start with the workA small international circle for different perspectives and real experiments. A contact route will be added before public launch.