Indicator selection must avoid the minimal implementation problem while remaining open to varied forms of consciousness, including exotic AI forms.

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

• Indicators used to assess AI systems for consciousness should focus on theories' central explanatory posits, abstracting away from implementation details that may differ across systems.claim0.825
  First of four guidelines for deriving indicators; prevents over-restriction to human-specific features
• Many of the indicator properties can be implemented in AI systems using current techniques.claim0.821
  Feasibility demonstrated in Section 3.1.
• AI systems which possess more of the indicator properties are more likely to be conscious.claim0.810
  Graded claim about the rubric.
• Constraints and biases of human intelligence are best avoided by ‘designing for emergence’ rather than implementing a pre-existing theory of intelligence.claim0.803
  Recommendation for creating non-anthropocentric machine intelligence.
• Building AI systems with more indicator properties will increase the likelihood of consciousness.hypothesis0.803
  Guiding hypothesis of the rubric.
• There are no obvious technical barriers to building AI systems which satisfy the indicator properties.claim0.802
  Feasibility claim about near-term conscious AI.
• Superficial behavioral indicators of consciousness are especially vulnerable to the gaming problem in AI contexts because engineers can design non-conscious systems that manifest them.claim0.802
  Further motivation for preferring internal computational indicators over behavioral ones
• The theory-derived indicator method provides a tractable way to reduce uncertainty about AI consciousness by deriving indicators from computational functionalist theories.claim0.799
  Central methodological claim of the paper