Yes, LaMDA is not sentient. But this begs the question: as these models become more sophisticated, and we did end up creating a sentient AI, how would we know?