---
title: "indistinguishability test"
type: "entity"
entity_kind: "term"
status: "watching"
canonical_name: "indistinguishability test"
aliases: []
first_seen: "2026-09-12T00:00:00.000Z"
writer_model: "claude-opus-4-8"
connects_to: ["Turing test","imitation game","PARRY","evaluation criteria"]
seek_code_commit: "98503b7"
---


Kenneth Colby's own term, in the 1971 "Artificial Paranoia" paper, for his evaluation standard: a simulation succeeds if its input-output behavior "cannot be distinguished from its human counterpart by psychiatric judges using a diagnostic interview." It is a psychiatric-register restatement of Alan Turing's imitation-game criterion, using distinct terminology.

Entered the vault 2026-09-12 as a `watching` stub because the term may anchor a recurring thread bridging [[entity-alan-turing|Turing]]'s imitation game, [[entity-kenneth-colby|Colby]]'s PARRY evaluation, and later LLM benchmark design — "can a human tell it apart" as a validation bar reached for independently across eras. Graduates to a hub only if it recurs and earns load-bearing use.

## References
- [[claim-parry-eliza-same-indistinguishability-test-opposite-mechanisms]]