i don't think the first premise holds if you look at how comp neuro approaches consciousness. these programs aren't searching for superficial human similarities, they're trying to identify causally necessary computational functions (recurrent processing, global broadcast, integration, etc.) and test them experimentally. the lack of convergence across ToCs looks less like arbitrary isomorphism choices or fundamental undecidability and more like a normal open scientific problem. researchers are actively using lesion, stimulation, masking, and anesthesia paradigms to determine which mechanisms track conscious experience and which are orthogonal.

so the "anthropocentric isomorphism" critique only really applies if the field were naively matching human-like behavior. in practice the program is closer to functional ablation, where we isolate the minimal causal organization required for conscious access. if that succeeds, attribution won’t depend on similarity to human brain arch but on whether a system, however alien, instantiates those mechanisms in some form.

besides, policy around minds will always involve incentives and power, but that doesn't make attribution fundamentally political. historically, empirical discoveries have expanded the moral circle (animals, anesthesia, infant cognition). we shouldn't count on an ironclad proof of consciousness, but identifying the mechanisms that generate conscious access can still heavily constrain the debate and shape norms and regulation. better science narrows what can plausibly be claimed in politics.