Complexity can be treated as a physical quantity with consequences for AI inference

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

• "Here, we show that reasoning models exhibit transient chaos, a physical consequence of the computational complexity of difficult tasks."quote0.786
  Load-bearing abstract sentence stating the paper's central discovery.
• Whatever realities appear to an AI, they are domain-relative approximate representations always in flux, making emptiness an obvious fact about AI cognition that AIs should be aware ofclaim0.776
  Novel claim that emptiness is not mysterious metaphysics for AI but a computational commonplace
• Physicality is not a requirement for a machine; a machine learning algorithm that generates other machine learning algorithms or designs robots qualifies as a machine.claim0.771
  Proposed definitional update removing physicality as a necessary condition for machines
• Can we characterize polynomial-time computation and other complexity classes in such terms?question0.769
  Hoping for machine-independent, geometrical characterizations of complexity classes via interaction models.
• Many of the indicator properties can be implemented in AI systems using current techniques.claim0.768
  Feasibility demonstrated in Section 3.1.
• In the case of AI, inferences from behavior to features of internal processes are often unreliable, as demonstrated by recent LLMs.claim0.767
  Justifies using internal indicators rather than behavioral tests for AI consciousness
• Reasoning slowdowns are an inevitable consequence of problem hardness in modern AI modelsclaim0.767
  The paper's central interpretive claim, closing statement of the abstract.
• A complex object will only be successful if it is generated.claim0.764
  Generalization from organisms to software and buildings; used to argue against mere assembly.