# Facebooks dark pattern design, public relations and internal work culture > ## Abstract ## Abstract - Facebook is criticized for various issues including privacy, dark patterns, false information, and user polarization. - According to leaked documents, there is a gap between official PR and internal discussions and research. - Analyzing these problems by analyzing leaked documents and published news articles ## Introduction - Facebook's active users are 2.9B, using an average of 38 minutes per day based on <2020> - According to materials submitted by employee to the Securities and Exchange Commission and The Wall Street Journal in , Facebook was experiencing many internal problems. RQs: - What type of [dark pattern](https://wiki.g15e.com/en/pages/Deceptive%20patterns.txt) designs Facebook inc. has applied to their site's recommendation algorithm and News Feed? ## Related Work Dark patterns (omitted) Facebook, News Feed and engagement - In the early days, Facebook was a dictionary-like site where people posted their hobbies, education, interests, relationships, etc. on their pages. *Before that, it was "Facemesh," a site that posted the faces of female college students without their consent and voted on their popularity (<2003>). After that, the name was changed to TheFacebook (<2004>) and then changed back to facebook (<2005>). --ak* - A "feed" first appeared in <2007>. A space to collect status updates from acquaintances - In <2011>, "Feed" changed to "News Feed" and became the core experience of Facebook. Collects information from various pages and profiles and displays it dynamically - Starting from <2009> the feed was not in chronological order and added a "Like" button. - Other reactions (Love, Haha, Wow, Sad, Angry) were added in <2016>. This information is used to expose more 'emotion heavy' content - This series of changes increased engagement, but also caused various problems. Hate speech, polarization, and information bubbles - Facebook invested heavily in research to automatically remove (13B). A state of arms race between hate speechers and algorithms. For now, the humans posting hate speech have the upper hand - The manual work of human moderators is also important. The human problem is moderator bias. *Facebook Korea is particularly famous for having serious gender bias issues. Some users call Facebook Blue Ilbe --ak*