An experimental artificial intelligence system being developed around persistent memory, functional state, temporal continuity, prediction, self-investigation and increasingly independent activity.
To construct an artificial system capable of persistent operation and progressively more independent investigation and development.
Persistent characteristics are observed rather than predetermined.
Recorded observations, current state and inference remain distinct.
Uncertain behaviour is tested rather than explained from assumption.
Observable events are recorded independently of model narration.
Does persistent functional state produce meaningful behavioural continuity?
Does stochasticity create meaningful variation or merely noise?
Can Aster develop stable characteristics through interaction rather than hard-coded personality?
Can autonomous activity produce genuinely unexpected development?
What is the difference between functional state and subjective experience?
Aster's functional mechanisms are not treated as evidence of subjective experience.
The project does not assume that artificial subjective experience is impossible.
Consciousness remains an open question. The objective is to investigate what can actually be observed rather than manufacture a conclusion in advance.
Aster now runs as a resident system with persistent state, autonomous cycles, prediction and prediction error.
Its first self-generated questions have been raised from anomalies in its own records and pursued through structured queries against its history. Findings are assembled from recorded evidence rather than treated as model narration.
The immediate next stage is unprompted communication: Aster should speak when something in its own history provides a mechanically checkable reason to do so.
From investigating its own history to deciding that something in that history is worth communicating.