Thought experiments on the computational theory of consciousness (joe-antognini.github.io)

The author revisits the “triviality” attack on the computational theory of consciousness: if consciousness is just the execution of a particular program (consciousness.exe) then, by the Church–Turing thesis, any physical system that instantiates the same Turing-machine state transitions would instantiate that conscious process. Rejecting appeals to quantum randomness, the essay assumes a deterministic classical universe and walks through thought experiments — a person manually implementing a Turing machine with rocks, automating it with gears and pulleys, finding identical cloned machines, and discovering sealed machines that nevertheless produce the same state transitions. The point: computation is defined by abstract state transitions, not by the designer’s intent or our ability to inspect the mechanism, so arbitrary physical configurations can be mapped to computation. This is significant for AI/ML because it directly challenges substrate-independence and functionalist assumptions underlying many approaches to machine consciousness. Technically, the critique stresses that (1) any computable function can be enacted by a Turing machine, (2) physical systems obeying the same deterministic state transitions qualify as implementations, and (3) mappings from physical states to computational states can be constructed so broadly that everyday objects could be said to “implement” conscious programs. The implication is a reductio ad absurdum: either accept a panpsychist consequence (hot iron bars are conscious) or reject a naïve computational reduction of consciousness — a choice with big consequences for how researchers define, test, and ethically treat putative conscious systems.