Building a fake computer voice right now is a decent amount of work, if I wanted to build one that sounded like you, for example, I'd need to get a ton of samples of you saying specific phonemes, and train a computer model on that.