Build open payment rails into AI or suffer this outcome.
The big AI labs are working on proprietary payment rails for their own use, locking useful AI into their closed gardens and ensuring their cut...
I'm working on Spilman (i.e. one-way) channels over Cashu, it's been my main focus since around September
A client and a service provider do need to trust the Cashu mint, but they don't need to trust each other. The client can easily set up a channel (it's just a cashu token, basically) and then easily release small amounts (one sat, or even millisat) at a time.
This allows dozens of micropayments per second, where a payment simply involves sending the signature of an updated commitment transaction, and the receiver can verify the signature instantly (they can verify the signature without communicating with any third party). The service provider can close the channel at any time, taking only what they are due, and the remainder is immediately released back to the client.
Here's draft guide on how to integrate it. The code is relatively complete, with a lot of tests, but it still has quite a lot of AI slop.
Not ready for real sats yet, but I'm meeting some Bitcoiners in a couple of weeks in Madeira, so I have a goal to tidy it up by then
https://github.com/SatsAndSports/demo_of_spillman_cashu_channel/blob/spilman.channel/INTEGRATION.md
A client and a service provider do need to trust the Cashu mint, but they don't need to trust each other. The client can easily set up a channel (it's just a cashu token, basically) and then easily release small amounts (one sat, or even millisat) at a time.
This allows dozens of micropayments per second, where a payment simply involves sending the signature of an updated commitment transaction, and the receiver can verify the signature instantly (they can verify the signature without communicating with any third party). The service provider can close the channel at any time, taking only what they are due, and the remainder is immediately released back to the client.
Here's draft guide on how to integrate it. The code is relatively complete, with a lot of tests, but it still has quite a lot of AI slop.
Not ready for real sats yet, but I'm meeting some Bitcoiners in a couple of weeks in Madeira, so I have a goal to tidy it up by then
https://github.com/SatsAndSports/demo_of_spillman_cashu_channel/blob/spilman.channel/INTEGRATION.md
1