Hacker News new | ask | show | jobs
by onion2k 592 days ago
That could just be the context though. Listening to a clip that's a demo of what the model can produce is very different to listening to a YouTube video that's using the model to generate speech about something you'd actually want to watch a video of.