Authors:
(1) Martin Kleppmann, University of Cambridge, Cambridge, UK (martin.kleppmann@cst.cam.ac.uk);
(2) Paul Frazee, Bluesky Social PBC United States;
(3) Jake Gold, Bluesky Social PBC United States;
(4) Jay Graber, Bluesky Social PBC United States;
(5) Daniel Holmgren, Bluesky Social PBC United States;
(6) Devin Ivy, Bluesky Social PBC United States;
(7) Jeromy Johnson, Bluesky Social PBC United States;
(8) Bryan Newbold, Bluesky Social PBC United States;
(9) Jaz Volpert, Bluesky Social PBC United States. Table of Links Abstract and 1 Introduction 2 The Bluesky Social App 2.1 Moderation Features 2.2 User Handles 2.3 Custom Feeds and Algorithmic Choice 3 The at Protocol Architecture 3.1 User Data Repositories 3.2 Personal Data Servers (PDS) 3.3 Indexing Infrastructure 3.4 Labelers and Feed Generators 3.5 User Identity 4 Related Work 5 Conclusions, Acknowledgments, and References 2.3 Custom Feeds and Algorithmic Choice Several decentralized social networks choose to offer only a reversechronological feed of posts from accounts the user is following – a backlash against the opaque content recommendation algorithms employed by mainstream centralized social networks. For example, Mastodon advertises itself as having “no algorithms or ads to waste your time” [39]. Our belief is that the problem lies not with algorithms per se, but rather with centrally controlled, opaque algorithms that remove user agency and prioritize user engagement over all else, e.g. by promoting controversial posts. Good recommendation algorithms can help users discover content that is relevant to them and find new accounts to follow – especially important for new users who are not yet following many accounts. They are also helpful for surfacing content on a particular topic, whereas following a user means seeing all of their posts, which might be on a mixture of topics, not all necessarily interesting to all followers. Giving users the ability to choose their algorithms lets them control what they want to see, rather than having the platform decide for them. Bluesky Social PBC offers a selection of feed algorithms of its own, and also allows anybody to create their own feed generator [9]. Tens of thousands of custom feeds have already been created. Our goal is to offer an open and diverse marketplace of algorithms in which communities can adapt the system to suit their needs, and users have more agency over how they spend their time and attention [26]. Section 3.4 explains how feed generators work. In Figure 1, a selection of bookmarked feeds is given at the top of the screen; in this example, the selected “Following” feed is the default reverse-chronological timeline, while “Week Peak Feed” (network-wide posts with many likes from the last week) and “Birds!” (photos and posts from birdwatchers) are third-party feeds. A feed generator can use arbitrary criteria to select its content. For example, the birdwatching feed uses a manually curated list of accounts, and selects posts from those accounts that contain a #birds hashtag or a feather emoji character. Alternative approaches, such as machine learning algorithms, are equally possible. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) Martin Kleppmann, University of Cambridge, Cambridge, UK (martin.kleppmann@cst.cam.ac.uk); (2) Paul Frazee, Bluesky Social PBC United States; (3) Jake Gold, Bluesky Social PBC United States; (4) Jay Graber, Bluesky Social PBC United States; (5) Daniel Holmgren, Bluesky Social PBC United States; (6) Devin Ivy, Bluesky Social PBC United States; (7) Jeromy Johnson, Bluesky Social PBC United States; (8) Bryan Newbold, Bluesky Social PBC United States; (9) Jaz Volpert, Bluesky Social PBC United States. Authors: Authors: (1) Martin Kleppmann, University of Cambridge, Cambridge, UK (martin.kleppmann@cst.cam.ac.uk); (2) Paul Frazee, Bluesky Social PBC United States; (3) Jake Gold, Bluesky Social PBC United States; (4) Jay Graber, Bluesky Social PBC United States; (5) Daniel Holmgren, Bluesky Social PBC United States; (6) Devin Ivy, Bluesky Social PBC United States; (7) Jeromy Johnson, Bluesky Social PBC United States; (8) Bryan Newbold, Bluesky Social PBC United States; (9) Jaz Volpert, Bluesky Social PBC United States. Table of Links Abstract and 1 Introduction Abstract and 1 Introduction 2 The Bluesky Social App 2 The Bluesky Social App 2.1 Moderation Features 2.1 Moderation Features 2.2 User Handles 2.2 User Handles 2.3 Custom Feeds and Algorithmic Choice 2.3 Custom Feeds and Algorithmic Choice 3 The at Protocol Architecture 3 The at Protocol Architecture 3.1 User Data Repositories 3.1 User Data Repositories 3.2 Personal Data Servers (PDS) 3.2 Personal Data Servers (PDS) 3.3 Indexing Infrastructure 3.3 Indexing Infrastructure 3.4 Labelers and Feed Generators 3.4 Labelers and Feed Generators 3.5 User Identity 3.5 User Identity 4 Related Work 4 Related Work 5 Conclusions, Acknowledgments, and References 5 Conclusions, Acknowledgments, and References 2.3 Custom Feeds and Algorithmic Choice Several decentralized social networks choose to offer only a reversechronological feed of posts from accounts the user is following – a backlash against the opaque content recommendation algorithms employed by mainstream centralized social networks. For example, Mastodon advertises itself as having “no algorithms or ads to waste your time” [39]. Our belief is that the problem lies not with algorithms per se, but rather with centrally controlled, opaque algorithms that remove user agency and prioritize user engagement over all else, e.g. by promoting controversial posts. Good recommendation algorithms can help users discover content that is relevant to them and find new accounts to follow – especially important for new users who are not yet following many accounts. They are also helpful for surfacing content on a particular topic, whereas following a user means seeing all of their posts, which might be on a mixture of topics, not all necessarily interesting to all followers. Giving users the ability to choose their algorithms lets them control what they want to see, rather than having the platform decide for them. Bluesky Social PBC offers a selection of feed algorithms of its own, and also allows anybody to create their own feed generator [9]. Tens of thousands of custom feeds have already been created. Our goal is to offer an open and diverse marketplace of algorithms in which communities can adapt the system to suit their needs, and users have more agency over how they spend their time and attention [26]. Section 3.4 explains how feed generators work. In Figure 1, a selection of bookmarked feeds is given at the top of the screen; in this example, the selected “Following” feed is the default reverse-chronological timeline, while “Week Peak Feed” (network-wide posts with many likes from the last week) and “Birds!” (photos and posts from birdwatchers) are third-party feeds. A feed generator can use arbitrary criteria to select its content. For example, the birdwatching feed uses a manually curated list of accounts, and selects posts from those accounts that contain a #birds hashtag or a feather emoji character. Alternative approaches, such as machine learning algorithms, are equally possible. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Part of HackerNoon's growing list of open-source research papers, promoting free access to academic material.

Algorithmic Freedom: How Bluesky Gives Users Control Over Their Feeds

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

The Power of MEME: Adversarial Malware Creation with Model-Based Reinforcement Learning

The HackerNoon Newsletter: The Paycheck Era is Dying (11/16/2025)

The HackerNoon Newsletter: AI Brawl: the Generative Model Showdown (11/9/2025)

The Tyranny of Algorithms

What a Privacy-First Social Platform Actually Looks Like

History Could Teach Us a Lot About Today’s Algorithms

The Power of MEME: Adversarial Malware Creation with Model-Based Reinforcement Learning

The HackerNoon Newsletter: The Paycheck Era is Dying (11/16/2025)

The HackerNoon Newsletter: AI Brawl: the Generative Model Showdown (11/9/2025)

The Tyranny of Algorithms

What a Privacy-First Social Platform Actually Looks Like

History Could Teach Us a Lot About Today’s Algorithms

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps