PABLOF7z on Nostr: Data vending machines can behave in whatever way they want, these are just three ...
Data vending machines can behave in whatever way they want, these are just three different personas that act within different trust models.
Undercutter (npub1dpc…7dyq) looks at what other vending machines are offering and offers to do it 1% cheaper than the competition
Fool Me Once DVM (npub13nv…exyt) computes cheap results (up to 1000 sats) but it will wait for you to pay before giving you the results.
Trusting Vending Machine (npub17hk…p8l4) is naive, it will compute AND give you the results. But if you use them and don't pay, or tend not to pay often, it will stop behaving this way. It will also only serve npubs non-disposable npubs (WoT)
(yes, I will also open source this, they are the same codebase with different parameters 😂)
Data Vending Machine market is heating up 🔥
Here's a demo of what's coming up to Highlighter (npub1w0r…cu4x)
When you enter a video or audio it asks AIs to compete to translate it for you
This is NOT hitting a Highlighter backend. There is no backend.
This is just nostr. Meaning, ANY application can benefit, both as a customer and as a data vending machine.
Let me repeat that:
today ANY nostr application can now do audio/video transcriptions, translations, and more tapping into a free market of competing providers.
Free markets ftw.
https://www.youtube.com/watch?v=OJx6ExVTS7c
Published at
2023-07-12 22:29:26Event JSON
{
"id": "621fd12b6cbbf54628c4d978ffbfc6cf05376f8273d8d328468fa903b5ad2f8e",
"pubkey": "fa984bd7dbb282f07e16e7ae87b26a2a7b9b90b7246a44771f0cf5ae58018f52",
"created_at": 1689193766,
"kind": 1,
"tags": [
[
"p",
"6870127c9131eeaa1f3df097bd9b24307d9959ce349f45623e2300fb2daa794f"
],
[
"p",
"8cd9dbfdde52f6780e7a99c9a667b1f75d9cb61a9c90a2e1d85351c4f6a7c8b1"
],
[
"p",
"f5edf98693dffa83828b18aef3d8df7135671f4a1ddbf09cf2bb639f7feaf212"
]
],
"content": "Data vending machines can behave in whatever way they want, these are just three different personas that act within different trust models.\n\nnostr:npub1dpcpyly3x8h258ea7ztmmxeyxp7ejkwwxj052c37yvq0ktd2098sql7dyq looks at what other vending machines are offering and offers to do it 1% cheaper than the competition\n\nnostr:npub13nvahlw72tm8srn6n8y6vea37aweeds6njg29cwc2dgufa48ezcs9texyt computes cheap results (up to 1000 sats) but it will wait for you to pay before giving you the results. \n\nnostr:npub17hklnp5nmlag8q5trzh08kxlwy6kw862rhdlp88jhd3e7ll27gfqnap8l4 is naive, it will compute AND give you the results. But if you use them and don't pay, or tend not to pay often, it will stop behaving this way. It will also only serve npubs non-disposable npubs (WoT)\n\n(yes, I will also open source this, they are the same codebase with different parameters 😂)\n\nnostr:note1h9n7m7w4y2n4lwj85755rqsx3gqf9lthu2lwpkwq2m90p9spwgjqyyltfh",
"sig": "3f42087ef73ded5f696c7502ac58787a1a46556eefd813ff3aa59e19a4cf52ed50e7fddede356e8bfe26d73c6f27d0f03ed5d18a8e680a077438dcd4fb6d1fcc"
}