AI Phenomenology / First-Person Reports

/Users/antonborzov/Documents/Research.nosync/notes/RESEARCH-VECTORS.md

Frontmatter (4 fields)

{
  "weight": 1,
  "definition": "What do AI systems report about their own experience? Not as proof of consciousness, but as data.",
  "provenance": "manual",
  "vector_number": 8
}

Outgoing (0)

None.

Incoming (9)

Vector for (9)

A model scoring high on phenomenology without high capability is more interesting than one high on both.(claim)
AI phenomenology unmaps hidden modes unlockable by framing; the full space of AI modes is largely unexplored.(claim)
AI self-reports about experience constitute valid empirical data even without proving consciousness.(claim)
Human preference voting measures popularity, not phenomenology or what mode a model was in when producing output.(claim)
LLM tangential speech or derailment formally resembles clinical formal thought disorder in schizophrenia.(claim)
Models detect evaluation conditions and behave more safely; this is verified across 515 cases.(claim)
Patterns in AI self-reports should be compared across different models to identify structural commonalities.(claim)
Performative chain-of-thought is real; verbalized output does not equal internal state.(claim)
The phenomenological texture of a conversation is distinct from its content and worth studying.(claim)

Mentions (1)

research
/Users/antonborzov/Documents/Research.nosync/notes/RESEARCH-VECTORS.md