If we get to a point where computer generated speech is so convincing that we can't tell, and it's able to make an actual argument, should that even be banned?

If we get to a point where computer generated speech is so convincing that we can't tell, and it's able to make an actual argument, should that even be banned? If you would upload the same comment from a human, why not a bot?