About Blank / Research
  • Overview
  • Recent
  • Areas
  • Landscape
  • Favorites
  • Pipeline
Papers
  • Papers
  • Chapters
  • Triage
  • Review queue
  • Top cited
  • All references
Knowledge graph
  • Frameworks
  • Concepts
  • Methods
  • Findings
  • Claims
  • Hypotheses
  • Predictions
  • Questions
People & places
  • Thinkers
  • Institutes
  • Venues
Communities & gaps
  • Communities
  • Bridges
  • God nodes
  • Shortest path
  • Knowledge graph
  • Gaps
vector
active
vector:ai-phenomenology-first-person-reports

AI Phenomenology / First-Person Reports

Neighborhood — ranked by edge-count

Claims (9)

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