Brilliant. Sky will be in your hands, it just takes open source to fulfil openai promises. The 10h transcribed audio dataset would be a great start for anyone

Brilliant. Sky will be in your hands, it just takes open source to fulfil openai promises. The 10h transcribed audio dataset would be a great start for anyone looking to reproduce, eventually it will all be in the open with enough people sharing recordings of their chatgpt voice assistant experiences on video. Openai must assume there will be voice tone clones very soon.