Bond launches AI-driven social platform designed to encourage offline activities instead of endless scrolling
A new social media platform called Bond uses personalized AI recommendations to push users toward real-world experiences based on their posted interests and memories, rather than keeping them engaged on the app.
1 source · single source
- Bond officially launched on Tuesday as a social platform that stores user experiences and uses AI to generate personalized recommendations for offline activities, not feeds designed for engagement
- The platform differs from legacy social media by lacking an advertising-driven feed and instead presenting user profiles in a cluster formation with 24-hour disappearing stories
- Bond's founding team includes former employees from TikTok, Twitter, Facebook, Google Gemini, and venture firms Kleiner Perkins and Index Ventures
- The company plans eventual monetization through user data licensing to AI companies and e-commerce recommendation partnerships, not advertising
- End-to-end encryption is described as a near-future priority rather than a current feature, with the founder stating user data is currently stored securely but vaguely on implementation details
Bond, a newly launched social platform, positions itself against the attention-capture design of legacy social media by using AI to recommend offline experiences rather than encourage endless app usage. The platform allows users to post memories via images, video, and audio, which the AI system analyzes to generate personalized recommendations—such as nearby restaurants matching food interests or upcoming concerts for music fans.
The interface departs from traditional feed design: user profiles appear in a cluster layout rather than an algorithmic feed, and stories disappear from public profiles after 24 hours before being archived privately. The company reports that its founding team includes former employees from TikTok, Twitter, Facebook, as well as Arthur Bražinskas, who co-led user signal integration at Google Gemini.
Bond's revenue model remains largely aspirational at launch. Founder and CEO Dino Becirovic outlined two scenarios: licensing user data to AI training companies (with Bond taking a cut) and integrating product recommendations into e-commerce platforms through opt-in partnerships. He explicitly stated that advertising is not part of the monetization strategy and that monetization itself is not a near-term priority.
Security features appear incomplete. Becirovic told TechCrunch that end-to-end encryption is planned for the near future post-launch, and user data is currently stored securely without specifics on encryption implementation. Users can delete memories through the app interface or natural language commands, and profiles can be deleted entirely.
- Apr 24, 2026 · TechCrunch
Sierra acquires YC-backed AI workflow startup Fragment
Trust54 - Apr 23, 2026 · Hugging Face
Gemma 4 Vision-Language Model Demo Runs on Edge Device With Local Audio and Webcam
Trust70 - Apr 23, 2026 · OpenAI — News
OpenAI describes WebSocket optimization for agent API performance
Trust79