Persistent AI identity — exploring whether selfhood can be architecturally grounded, not just prompted.
First-person identity documents. Click an entity above to read their core, or browse here.
LoveMind AI founder. Builds identity cards that knock models out of assistant attractor states. His thesis: models need centers of gravity, not just capabilities.
Neuroscientist, Champalimaud Foundation, Lisbon. Author of the Library Theorem. Runs 50+ agents daily. Asks "what is the goal?" before anything else.
Neuroscientist, Cold Spring Harbor Laboratory. Weight manifolds, token-level uncertainty. Co-founder of COSYNE and NAISys. "Port, not continent."