I'm happy to announce my very first pre alpha data vending machine provider.
But first...
What is a data vending provider ?
The concept of data vending provider has been imagined by
.The NIP is not yet merged but Nostr magic in place, we can already build and use it. github.com/nostr-protocol/nips/tree/vending-machine
This specification let builder, user, AI provider to create a distributed market of the data.
For the purpose, lets say Alice can provide transcription (transform audio to text), image generation, GPT thing. Bob want to transcribe a youtube video but don't want to deal directly with AWS or other thing (Microsoft Azure)...
Bob can publish a specific note on Nostr that say (hey, I want to transcribe this youtube video)? Alice can respond, "hey, I can do that for a small fee". Bob can pay. Alice can now work on it (obviously she can shit or even scam) and publish the result.
How to use DVM GeniusBot
That the fun part. Transcription job can be made on dev.highlighter.com/web or vendata.io.
We will show on a demo on the last one.
Let's go to the "Create Job" tab and click on "text extraction".
Type a youtube URL and optionally a segment.
Click on create. After 10 secondes to 1 minute (depending how the DVM GeniusBot is healthy), a proposition appeared.
Pay the lightning invoice.
... After few minutes (depending the audio duration), DVM GeniusBot provides a transcription.
