Built a quick fun experiment for interacting with LLMs like you're on a talkshow.
It's sort of like NotebookLM in the sense that it makes it very easy to generate a podcast, but in this setup you're the podcast guest.
To use it, just let talkshow know what you want to talk about and it'll ask you a bunch of questions just like you're on a talkshow. It's all via voice and oddly fun to do.
Try creating your own interview - would love any thoughts as you're playing with it.
Using Realtime for enabling the interrupts and stock browser items for most other things. Everything over websockets with some light fast processing before piping to OpenAI. Nothing too fancy.
It's sort of like NotebookLM in the sense that it makes it very easy to generate a podcast, but in this setup you're the podcast guest.
To use it, just let talkshow know what you want to talk about and it'll ask you a bunch of questions just like you're on a talkshow. It's all via voice and oddly fun to do.
Try creating your own interview - would love any thoughts as you're playing with it.