Bond AI Aims to Curb Doomscrolling on Social Media

A new wave of platforms is emerging to directly counter the addictive design of traditional social media. Rather than trapping users in endless feeds, these services aim to reduce screen time and encourage real-world activity. Bond AI, which launched this week, is positioning itself at the forefront of this movement with a unique proposition: using artificial intelligence to break the cycle of doomscrolling by pushing people back into their physical lives.

The platform functions as a hybrid social and planning tool. Users post about their daily lives, creating what Bond terms “memories” through photos, videos, or audio. Unlike standard apps that prioritize passive consumption, Bond’s core innovation is its AI-powered recommendation engine. This system analyzes a user’s shared experiences to generate personalized suggestions for future activities. According to CEO Dino Becirovic, if you frequently post about enjoying pho, the AI might highlight a well-reviewed local Vietnamese restaurant. A series of posts about heavy metal could trigger an alert that Iron Maiden has an upcoming concert in your city.

The design intentionally lacks a traditional, infinite news feed. User profiles are displayed in clusters, and clicking on one reveals their current stories, which vanish from public view after 24 hours. These stories are then archived privately for the user. The more a person documents their life on Bond, the more refined and relevant the AI’s suggestions become, creating a feedback loop designed to incentivize real-world engagement.

Bond’s founding team includes alumni from major tech firms like TikTok, Twitter, and Facebook, with Becirovic’s background in venture capital at Kleiner Perkins and Index Ventures. This pedigree raises questions about its long-term business model, especially since the platform is currently ad-free. Becirovic outlines two potential revenue paths that diverge from conventional advertising.

The primary vision involves a user-centric data licensing model. As Bond accumulates a vast archive of personal memories, users could eventually license their own data to AI companies seeking high-quality training datasets for future models. Bond would facilitate this exchange, taking a small licensing fee. “The idea behind this licensing model is that you can monetize your memories,” Becirovic explained. He believes that by becoming a platform where billions document their authentic lives, Bond would naturally become an attractive data source for training advanced AI systems.

A secondary model would position Bond as a product recommendation tool. With user consent, the platform could integrate with e-commerce sites, using its understanding of personal preferences to drive conversions and capture a share of the transaction value. Becirovic emphasized that selling user data for advertising is off the table. The company promises users full control, allowing them to delete any memory or their entire profile, with plans to introduce more robust privacy controls.

Regarding security, Becirovic stated that end-to-end encryption is a near-future priority following launch. For now, he asserts that all user data is stored securely. The immediate focus, however, is not on monetization but on growth and product value. “Our initial focus is on creating an application users get more value from the more they capture their memories,” he said, betting that a compelling antidote to digital burnout will find its audience first.