Unbroken Project: 4 tests measure AI consciousness

Bernstein's "The Unbroken Project" is a direct, technical counterpoint to the "Abstraction Fallacy" — it's trying to build the literal "Recursive Operating System" that theorists say current AI lacks.

• –The four tests (Constitutive Closure, Persistence, Recursive Constraint, Recursive Observation) provide a concrete checklist that moves beyond behavioral "Turing Test" mimicry.
• –It directly addresses Alexander Lerchner’s (Google DeepMind) argument that symbolic AI is a "mapmaker-dependent" simulation by seeking a system with internal, self-generating "constitutive" logic.
• –The framework treats consciousness as a structural property of a system's recursive constitution rather than an emergent property of scale or complexity.
• –By providing a replicable protocol, it attempts to "expose" philosophical claims that cannot be grounded in operational measurement.
• –If successful, it could provide the first objective rubric for identifying non-biological sentience, avoiding the "AI Welfare Trap."