There has recently been widespread discussion of whether GPT-3, LaMDA 2, and related large language models might be sentient. Should we take this idea seriously? I will discuss the underlying issue and will break down the strongest reasons for and against.