Mozilla turned its attention to the YouTube recommendation algorithm yet again with TheirTube, in which it created recommendation bubbles for six different personas, based on interviews with actual YouTube users.
Last October, the nonprofit created the YouTube Regrets website to share 28 instances where the Google-owned video site’s recommendations engine led people “down bizarre and dangerous paths,” it said at the time.
Mozilla said in a blog post Wednesday, “YouTube’s recommendation algorithm pays close attention to what you watch, and then recommends other videos that you might enjoy. This algorithm is incredibly powerful: It accounts for 70% of all videos viewed on YouTube. Recommendations are useful, but they can also trap you in an information bubble, reinforcing the same points of view over and over again.”
TheirTube presents the YouTube landing page from the viewpoint of six different personas: climate denier, conservative, conspiracist, fruitarian, liberal and prepper.
Mozilla said each persona was informed by interviews with actual YouTube users who experienced similar recommendation bubbles, and it created six YouTube accounts to simulate those users’ subscriptions and viewing habits.
Visitors to TheirTube can see videos recommended to those six accounts, as well as their complete watch histories, with Mozilla providing the example that a user who was skeptical about climate change was fed more videos debunking scientific evidence about global warming, confirming the bias they already had.
Tomo Kihara, an Amsterdam-based creative developer and Mozilla Creative Media Awardee, created TheirTube, and he said in Mozilla’s blog post, “This project raises questions like: What if platforms were more open about the recommendation bubbles they created? And: By experiencing other users’ bubbles, is it possible to gain a better perspective of your own recommendation environment?”
Mozilla vice president of global programs J. Bob Alotta added, “AI (artificial intelligence) plays a central role in consumer technology today—it curates our news, it recommends who we date and it targets us with ads. Such a powerful technology should be demonstrably worthy of trust, but often it is not. Mozilla’s Creative Media Awards draw attention to this, and also advocate for more privacy, transparency and human well-being in AI.”