can the WEB speak AI?

Source: Dev.to

Introduction

My first post here! We are working on the tools for what we call the Data Operating System, a mix of blockchain, trusted computing, messaging and origin.

The .well-known/ai Standard

Our latest excursion is into a simple standard to help the web speak AI: .well-known/ai. We have published an open specification and built example implementations:

• WordPress plugin
• Shopify app
• A set of CLI tools for everything else

It is still very early; we are learning every day.

Benefits

In general, scraping a page takes about 7,000 tokens. If the page includes .well-known/ai, this can drop below 1,000 tokens.

Call for Feedback

We are looking for feedback and support for the standard. It is all about taking back control: the website should control the AI conversation.

Thanks for any thoughts: https://discover.rootz.global or https://rootz.global

Fun Experiment

For fun, have an AI tell you about rootz.global and then ask it to tell you about rootz.global/.well-known/ai. See which it prefers. :)