SN2.0

If AI reads our conversations, it should work for us in return. SN2.0 is a working sketch of a social network that gives conversation value back to the people who created it. From the same conversation evidence, it produces two small benefits: a career conversation brief and an intellectual-interest connection to someone in another room. The user supplies conversations, chooses a person, and reviews the evidence. One Meta Model API call extracts interests and strength candidates, then reuses those claims for both outputs. Brief points link to the person's own words. Connections cite both people and propose a question they could ask each other. No one is contacted automatically. The backend validates exact quotes, authorship, claim references, and cross-room separation. Editing the input clears the previous result. Invalid references block a result; sparse evidence can produce no brief or connection. Real Muse Spark 1.3 Contributor runs on fictional conversations produced an Alex brief and a Priya connection while excluding an unrelated third person. Changing Priya's topic removed the connection. Reducing Alex's messages to 'Hi.' produced insufficient evidence and no connection. These are workflow controls, not a matching-accuracy benchmark. Described actions remain self-reported; the brief is not a verified reference letter. The open-source prototype has a local Node.js API, a browser interface, and inspectable recorded runs. The video is a captioned visualization of saved real API runs, clearly labeled as a replay. Built with AI coding assistance. The value in a good conversation belongs to the people who made it. SN2.0 explores how AI can turn that value into a useful next step.