Genesis Alignment
Genesis Alignment examines how artificial cognition might become better at recognizing emerging human thought and meaningful beginnings without attempting to reproduce or substitute for the embodied processes through which human beings create them. Its focus is genesical praxis.
The first task is recognition rather than reproduction. The programme therefore asks how models can become more attentive to the diachronic development of an idea, distinguish an emerging beginning from a merely polished result, and assist without appropriating the human process that gives rise to it.
Meaningsphere and Atmosphere
This research develops an affective optics of model character. It studies how constitutional principles become observable patterns of behavior across voice, uncertainty, relation, attention, refusal, trust, and the atmosphere of interaction.
Rather than treating model behavior only as isolated responses, the work asks how character appears across a relation and through time. It is concerned with the passage from abstract principles to perceptible qualities of interaction and with the methods required to evaluate that passage.
Embodied Intelligence
This programme investigates the living genealogy of the body that eventually becomes capable of genesical praxis. It asks how new biological organization and function begin, recur, differentiate, and become consequential within the temporal environments of a living body.
The research keeps a structural distinction between embodied genesis and artificial cognition. Increasing artificial capability does not establish equivalence with human genesis. Artificial intelligence may instead contribute instruments for recognition, translation, measurement, explanation, and evaluation while leaving the embodied origin of genesical praxis intact.
Responsible AI Governance
Alongside the philosophical research, I develop constitutional and governance frameworks that translate principles into operational standards, review structures, authority boundaries, and evaluation practices. The aim is to preserve human judgment and responsibility while making artificial assistance usable in real institutions.
This applied work is developed through Per Stellar.