How Platform Recommenders Work – Center for Human-Compatible Artificial Intelligence

Letting Users Choose Recommender Algorithms: An Experimental Study

Recommender systems are not one-size-fits-all; different algorithms and data sources have different strengths, making them a better or worse fit for different users and use cases. As one way of taking advantage of the relative merits of different algorithms, we gave users the ability to change the algorithm providing their movie recommendations and studied how they make use of this power. We conducted our study with the launch of a new version of the MovieLens movie recommender that supports multiple recommender algorithms and allows users to choose the algorithm they want to provide their recommendations. We examine log data from user interactions with this new feature to understand whether and how users switch among recommender algorithms, and select a final algorithm to use. We also look at the properties of the algorithms as they were experienced by users and examine their relationships to user behavior.

Understanding Decentralized Social Feed Curation on Mastodon

Amid rising concerns over moderation, algorithmic control, and platform governance on centralized social media, people are increasingly turning to decentralized alternatives like Mastodon to regain control over their feeds. This shift offers new opportunities to understand how people perceive and curate their feeds. We conducted a two-part study with 21 Mastodon users: first, interviews exploring how they perceive and manage their feeds; and second, a design probe study using BRAIDS.SOCIAL, a web-based feed curation prototype informed by the first part of our initial findings. We learned how seamful design can increase people's trust in algorithmic curation, and surfaced trade-offs people navigate between machine learning-based and rule-based filtering approaches. We also identify a core design tension in decentralized platforms: whether to support personalization through new applications or extensions layered atop existing ones.

Bridging-Based Ranking

There is significant concern about the engagement-based ranking systems used by TikTok, Facebook, YouTube, etc. to recommend content. Bridging-based ranking systems can address one of the most dangerous aspects of such algorithmic recommendations—the push toward polarization and divisiveness that is tearing nations apart—and do so without reducing anonymity or increasing censorship. This report explores what bridging-based ranking is, how it helps (overcoming downsides of chronological feeds and middleware), addresses common objections, and provides early examples of its use and benefits in the wild. The report concludes by providing next steps for platforms, governments, funders, and researchers in order to accelerate the deployment of bridging.

Paper Skygest: Personalized Academic Recommendations on Bluesky

We build, deploy, and evaluate Paper Skygest, a custom personalized social feed for scientific content posted by a user's network on Bluesky and the AT Protocol. We leverage a new capability on emerging decentralized social media platforms: the ability for anyone to build and deploy feeds for other users, to use just as they would a native platform-built feed. To our knowledge, Paper Skygest is the first and largest such continuously deployed personalized social media feed by academics, with over 50,000 weekly uses by over 1,000 daily active users, all organically acquired. First, we quantitatively and qualitatively evaluate Paper Skygest usage, showing that it has sustained usage and satisfies users; we further show adoption of Paper Skygest increases a user's interactions with posts about research, and how interaction rates change as a function of post order. Second, we share our full code and describe our system architecture, to support other academics in building and deploying such feeds sustainably. Third, we overview the potential of custom feeds such as Paper Skygest for studying algorithm designs, building for user agency, and running recommender system experiments with organic users without partnering with a centralized platform.

A Literature Review of Ethical Considerations in Recommender Systems for User-Generated Content in Human-Computer Interaction

The design of user-generated content (UGC) platforms poses challenges in comprehensively addressing the ethical dimensions of recommendation algorithms and applying human-centered methods for their evaluation. This article presents a literature review of 97 studies on UGC algorithms (UGCAlgos) that incorporate human factors and user experience considerations to investigate the ethical issues explored in human-computer interaction (HCI) research. Our review identifies key themes in the ethical considerations surrounding UGCAlgos and the user modeling methods employed. We examine how common ethical concerns in recommender systems, such as content appropriateness, privacy, user engagement, transparency, fairness, and diversity, are studied and contextualized within UGC platforms. Furthermore, we summarize how these concerns are addressed through user modeling approaches, including data characterization, user context, user outsmarting, interface design, user beliefs, and community and societal impacts.

Do You Like These AI Images of Dying, Mutilated Children? Facebook Algorithm Wonders

Facebook's recommendation engine, which is core to the company's business model, is injecting grotesque images of AI-generated child gore into people's feeds.

Reranking partisan animosity in algorithmic social media feeds alters affective polarization

Today, social media platforms hold the sole power to study the effects of feed-ranking algorithms. We developed a platform-independent method that reranks participants’ feeds in real time and used this method to conduct a preregistered 10-day field ...

Rational Silence and False Polarization: How Viewpoint Organizations and Recommender Systems Distort the Expression of Public Opinion

Social media platforms are one of the most important domains in which artificial intelligence (AI) has already transformed the nature of economic and social interaction. AI enables the massive scale and highly personalized nature of online information sharing that we now take for granted. Extensive attention has been devoted to the polarization that social media platforms appear to facilitate. However, a key implication of the transformation we are experiencing due to these AI-powered platforms has received much less attention: how platforms impact what observers of online discourse come to believe about community views. These observers include policymakers and legislators, who look to social media to gauge the prospects for policy and legislative change, as well as developers of AI models trained on large-scale internet data, whose outputs may similarly reflect a distorted view of public opinion. In this paper, we present a nested game-theoretic model to show how observed online opinion is produced by the interaction of the decisions made by users about whether and with what rhetorical intensity to share their opinions on a platform, the efforts of viewpoint organizations (such as traditional media and advocacy organizations) that seek to encourage or discourage opinion-sharing online, and the operation of AI-powered recommender systems controlled by social media platforms. We show that signals from ideological viewpoint organizations encourage an increase in rhetorical intensity, leading to the rational silence of moderate users. This, in turn, creates a polarized impression of where average opinions lie. We also show that this observed polarization can also be amplified by recommender systems that, pursuant to a platform’s incentive to maximize engagement, encourage the formation of viewpoint communities online that end up seeing a skewed sample of opinion. Unlike existing models, these well-known online phenomena are not here attributed to distortion in the formation of opinions nor to the seeking out of like-minded others, but rather to the interaction of the incentives of users, viewpoint organizations, and platforms implementing recommender systems. In addition to showing how these interactions can play out in simulations, we also identify practical strategies platforms can implement, such as reducing exposure to signals from ideological viewpoint organizations and a tailored approach to content moderation.

Bonsai: Intentional and Personalized Social Media Feeds

Social media feeds use predictive models to maximize engagement, often misaligning how people consume content with how they wish to. We introduce Bonsai, a system that enables people to build personalized and intentional feeds. Bonsai implements a platform-agnostic framework comprising Planning, Sourcing, Curating, and Ranking modules. This framework allows users to express their intent in natural language and exert fine-grained control over a procedurally transparent feed creation process. We evaluated the system with 15 Bluesky users in a two-phase, multi-week study. We find that participants successfully used our system to discover new content, filter out irrelevant or toxic posts, and disentangle engagement from intent, but curating intentional feeds required more effort than they are used to. Simultaneously, users sought system transparency mechanisms to effectively use (and trust) intentional, personalized feeds. Overall, our work highlights intentional feedbuilding as a viable path beyond engagement-based optimization.

Understanding Social Media Recommendation Algorithms

Access the PDF version of this essay by clicking the icon to the right.

The political effects of X’s feed algorithm

Feed algorithms are widely suspected to influence political attitudes. However, previous evidence from switching off the algorithm on Meta platforms found no political effects1. Here we present results from a 2023 field experiment on Elon Musk’s platform X shedding light on this puzzle. We assigned active US-based users randomly to either an algorithmic or a chronological feed for 7 weeks, measuring political attitudes and online behaviour. Switching from a chronological to an algorithmic feed increased engagement and shifted political opinion towards more conservative positions, particularly regarding policy priorities, perceptions of criminal investigations into Donald Trump and views on the war in Ukraine. In contrast, switching from the algorithmic to the chronological feed had no comparable effects. Neither switching the algorithm on nor switching it off significantly affected affective polarization or self-reported partisanship. To investigate the mechanism, we analysed users’ feed content and behaviour. We found that the algorithm promotes conservative content and demotes posts by traditional media. Exposure to algorithmic content leads users to follow conservative political activist accounts, which they continue to follow even after switching off the algorithm, helping explain the asymmetry in effects. These results suggest that initial exposure to X’s algorithm has persistent effects on users’ current political attitudes and account-following behaviour, even in the absence of a detectable effect on partisanship.

The political effects of X’s feed algorithm

Feed algorithms are widely suspected to influence political attitudes. However, previous evidence from switching off the algorithm on Meta platforms found no political effects1. Here we present results from a 2023 field experiment on Elon Musk’s platform X shedding light on this puzzle. We assigned active US-based users randomly to either an algorithmic or a chronological feed for 7 weeks, measuring political attitudes and online behaviour. Switching from a chronological to an algorithmic feed increased engagement and shifted political opinion towards more conservative positions, particularly regarding policy priorities, perceptions of criminal investigations into Donald Trump and views on the war in Ukraine. In contrast, switching from the algorithmic to the chronological feed had no comparable effects. Neither switching the algorithm on nor switching it off significantly affected affective polarization or self-reported partisanship. To investigate the mechanism, we analysed users’ feed content and behaviour. We found that the algorithm promotes conservative content and demotes posts by traditional media. Exposure to algorithmic content leads users to follow conservative political activist accounts, which they continue to follow even after switching off the algorithm, helping explain the asymmetry in effects. These results suggest that initial exposure to X’s algorithm has persistent effects on users’ current political attitudes and account-following behaviour, even in the absence of a detectable effect on partisanship.

Towards a Post-Social Media Studies

For two decades, "social media" has been the master lens through which scholars have understood digital communication—an era defined by user-generated content, networked publics, and participatory culture. That era is drawing to a close. This paper argues that three interrelated dynamics are dissolving the social media paradigm: an algorithmic shift from social-graph-based to interest-based recommendation, which is remaking the active "user" into a passive "viewer"; the generative AI revolution, which is replacing user-generated content with synthetic media and decoupling platforms from any dependence on human participation; and an exodus from public platforms toward private, closed spaces. Together, these dynamics are giving rise to three distinct post-social formations: algorithmically governed broadcasting platforms, semi-private spheres and micro-communities, and AI-mediated communication as a new media form in its own right. Understanding this transformation demands a fundamental reorientation of the field—new conceptual tools that move beyond networked publics and participatory culture, new methods suited to synthetic and ephemeral media environments, and renewed attention to the political stakes of communication systems that no longer require human participation.

Towards a Post-Social Media Studies

For two decades, "social media" has been the master lens through which scholars have understood digital communication—an era defined by user-generated content, networked publics, and participatory culture. That era is drawing to a close. This paper argues that three interrelated dynamics are dissolving the social media paradigm: an algorithmic shift from social-graph-based to interest-based recommendation, which is remaking the active "user" into a passive "viewer"; the generative AI revolution, which is replacing user-generated content with synthetic media and decoupling platforms from any dependence on human participation; and an exodus from public platforms toward private, closed spaces. Together, these dynamics are giving rise to three distinct post-social formations: algorithmically governed broadcasting platforms, semi-private spheres and micro-communities, and AI-mediated communication as a new media form in its own right. Understanding this transformation demands a fundamental reorientation of the field—new conceptual tools that move beyond networked publics and participatory culture, new methods suited to synthetic and ephemeral media environments, and renewed attention to the political stakes of communication systems that no longer require human participation.

Agency Among Agents: Designing with Hypertextual Friction in the Algorithmic Web

Today's algorithm-driven interfaces, from recommendation feeds to GenAI tools, often prioritize engagement and efficiency at the expense of user agency. As systems take on more decision-making, users have less control over what they see and how meaning or relationships between content are constructed. This paper introduces "Hypertextual Friction," a conceptual design stance that repositions classical hypertext principles--friction, traceability, and structure--as actionable values for reclaiming agency in algorithmically mediated environments. Through a comparative analysis of real-world interfaces--Wikipedia vs. Instagram Explore, and Are.na vs. GenAI image tools--we examine how different systems structure user experience, navigation, and authorship. We show that hypertext systems emphasize provenance, associative thinking, and user-driven meaning-making, while algorithmic systems tend to obscure process and flatten participation. We contribute: (1) a comparative analysis of how interface structures shape agency in user-driven versus agent-driven systems, and (2) a conceptual stance that offers hypertextual values as design commitments for reclaiming agency in an increasingly algorithmic web.