rNet proposes ISP-style billing model where users fund their own AI usage directly instead of paying per app

rNet is an early proof of concept that shifts AI cost responsibility from developers to users. Currently, developers track per-user usage, build pricing systems, and handle upfront AI token costs; users pay separately across different applications.

The protocol is live, with libraries published on npm and Maven Central, and documentation available. It aims to work like an Internet Service Provider model, where users pay for their own usage directly rather than through multiple app-based transactions.

The creator is soliciting feedback on where the model breaks technically or conceptually, indicating the project remains in experimental stages.