Hola nostr:npub1sg6plzptd64u62a878hep2kev88swjh3tw00gjsfl8f237lmu63q0uf63m, nostr:npub1cn4t4cd78nm900qc2hhqte5aa8c9njm6qkfzw95tszufwcwtcnsq7g3vle
We're building a thing here on #Nostr.
Discussion
Subtitles? Or even made to speech?
Users request AI tasks to Nostr. Providers listen to these events and offer to perform the task if they can, and the user can hand them to their (for whatever reason) prefered provider.
In this example we're doing a transcription of a clip of the interview of the Jacks and follow up with a request to translate it (in this case to Spanish). But it literally could be any task someone provides to perform.
Maybe someone will summarize this task with a Large Language model and someone else will create an image based on this summarization. Possibilities are endless.
Nice!
I would love an AI model summarizing podcast/book/article transcripts as memes
Yeah, that’ll be supported on highlighter
The wild thing is that because I don’t have to code each use case, once there is a market of different, composable, models, each app that integrates the Data Vending Machine NIP supports ALL use cases.
It’s wild.
