I just open-sourced a project for on-device dictation, seeing if anyone is looking for an alternative to cloud-based dictation and not loving the current solutions. Wanted to open up the discussion here on the gap between cloud vs local and what users are looking for from a product standpoint when they talk to their computers :)