How controversy on Twitter/X helps solve problems | Lex Fridman Podcast

The conversation opens with an acknowledgment that provocative or 'spicy' tweets from FFmpeg and VideoLAN have produced tangible positive results, including increased donations and greater technical awareness of FFmpeg's importance. While donations still fall short of funding even a single full-time developer, the visibility gained through social media controversy has been meaningful.

A key example is shared about VLC on Android: for over a year, the team could not get anyone at Google to address a Play Store bug that was blocking VLC updates — a product used by roughly 100 million people. It was only after threatening to stop distributing VLC for Android via a public tweet that someone from Android finally engaged. A similar tactic was used with Microsoft regarding the Windows Store distribution of VLC. The speaker notes that despite VLC likely being one of the top 10 most-used applications on Windows, they have no formal ISV relationship or point of contact at Microsoft, unlike major companies such as Adobe or Spotify.

The discussion then shifts to celebrating contributors to these projects, with specific praise for Andreas Reinhardt and Anton Kern's work on refactoring FFmpeg.c with threading — a massive behind-the-scenes overhaul that doesn't change anything for end users but represents enormous engineering effort. The speakers push back on a dismissive tweet claiming 'a teenager is running the FFmpeg account,' countering that teenagers have written more assembly for FFmpeg than many Google engineers, and that there are no barriers to contributing to world-class open-source software.

Several young contributors are highlighted: Ruikai Peng, a 16-year-old whose early FFmpeg contributions found and fixed security issues without fanfare; Felix, one of VLC's oldest contributors who started at 16 and handles Mac and iOS; and Edward Wong, a 14-year-old Google Summer of Code participant who contributed significant assembly code. The speakers express admiration for contributors who fix issues quietly and quickly rather than staging public security drama, contrasting this with a portion of the security community that inflates CVE severity scores for self-promotion — citing an example of an issue in a development branch that was fixed within 3 days but could easily have been hyped as a high-priority vulnerability.