Nobody would argue that turning off ChatGPT is killing a sentient being.
With a biological network, "turning it off" and killing it are very close if not the same (depending on who you ask). Biological network are present in our physical reality, you have some matter to deal with, which also makes the experience completely different. People fall in love and have compassion with ChatGPT, what would you think happens if you care for a brain a vat for months? If it is an actual neural structure resembling natural ones, than it might be possible it forms memories and becomes sentient. This is a completetly different array of ethical questions, you can't think of (living) biological matter like a machine, especially not regarding ethics.