The Open Protocol Future

Why Centralization Matters

The attention economy's capture mechanisms are not intrinsic to social media. They are intrinsic to a specific business model applied to social media — the model in which a single centralized entity controls the distribution of content, monetizes user attention through behavioral advertising, and optimizes distribution for the behavioral signal density that makes advertising targeting profitable. Remove the centralized control over distribution, and the financial incentive for engagement maximization changes fundamentally.

This is the structural argument for open protocols as a complement to — and in the long run, a potential replacement for — regulatory approaches to attention capture. Regulation can mandate specific design changes (chronological feeds, opt-in notifications) on centralized platforms. Open protocols change the underlying incentive architecture that makes non-compliant design commercially rational. They address the root rather than the branches.

The centralization argument works as follows. A centralized platform that controls distribution also controls what content is seen, in what order, by which users. This control is the foundation of the behavioral advertising model: the platform's ability to show specific content to specific users based on their behavioral profiles is what makes their advertising inventory valuable. Remove central control over distribution — allow users to choose their own algorithms, their own feeds, their own content filters — and the advertising value proposition dissolves. An advertising system cannot target content to users based on behavioral profiles it doesn't control. Without behavioral targeting, behavioral advertising revenue falls sharply. Without behavioral advertising revenue, the financial case for engagement maximization disappears.

This is a structural intervention, not a regulatory one. It does not require legislation. It requires users to switch platforms — which is why, despite the structural logic being available since 2008 (the year of Twitter's API opening and the early Fediverse), it has not yet produced meaningful displacement of the centralized incumbents.

The Open Protocol Landscape

Three protocols represent the current state of decentralized social infrastructure.

ActivityPub (W3C, 2018)

ActivityPub is a W3C-standardized protocol for federated social networking — the technical foundation of the Fediverse. Mastodon (microblogging), Pixelfed (photo sharing), PeerTube (video), and dozens of other applications implement ActivityPub, allowing users on different servers to follow each other and interact across platform boundaries. The key feature is federation: a user on mastodon.social can follow a user on fosstodon.org and see their posts in the same feed, without either user needing an account on the other's server. There is no central server; there is no central authority.

ActivityPub's adoption grew substantially during the 2022–2023 Twitter disruption (following Elon Musk's acquisition), reaching approximately 10–12 million monthly active users across the Fediverse by 2025. This represents genuine adoption — not experimental accounts, but active daily use. It also represents approximately 1% of Twitter's user base at its peak, and less than 0.5% of Meta's.

AT Protocol (Bluesky, 2023)

The AT Protocol (Authenticated Transfer Protocol) was developed by Bluesky, a company founded with Twitter investment that became independent in 2022. It differs from ActivityPub in its architecture: rather than federation between servers, AT Protocol uses a separation between the data layer (Personal Data Servers, which users can self-host or use a provider for) and the application layer (clients that connect to the data layer and apply different algorithms and interfaces). This separation is significant: it makes the algorithm genuinely user-configurable at the protocol level rather than through a platform-provided settings menu.

Bluesky reached approximately 30 million registered accounts by early 2025, with substantially lower monthly active user rates. Its feed customization features — which allow users to select from community-developed algorithms — represent the most significant implementation of the “choose your own algorithm” concept available at scale.

Nostr (2020)

Nostr (Notes and Other Stuff Transmitted by Relays) is the most decentralized and most technically radical of the three. It has no servers, no federation, and no central authority of any kind — only relays (servers that store and forward messages) and clients (applications that connect to relays). Identity is cryptographic: users are identified by public keys, not by usernames registered with a platform. Nostr's user base is smaller than Mastodon's and more technically sophisticated. It represents the most theoretically pure implementation of decentralized social infrastructure.

What Open Protocols Change About Incentive Architecture

Open protocols change three specific incentive structures that currently drive engagement-maximizing design.

They eliminate behavioral advertising as the revenue base. A platform built on an open protocol cannot lock in behavioral data, because the data is not held centrally. Without behavioral data, behavioral targeting is impossible. Without behavioral targeting, the advertising premium that behavioral advertising commands disappears. The advertising inventory available on open protocol networks is priced at contextual advertising rates — lower than behavioral rates, but not zero. The financial case for engagement maximization depends on the behavioral advertising premium; remove the premium and the financial case weakens substantially.

They make algorithm choice user-controllable. On a centralized platform, the feed algorithm is chosen by the platform and applied to all users. On a protocol like AT Protocol, users can choose from community-developed algorithms that optimize for different objectives: recency, relevance as defined by the user, content from followed accounts only, topic-specific curation. The diversity of available algorithms means that engagement-maximizing algorithms compete with wellbeing-optimized algorithms on user choice rather than being imposed as the only option.

They create interoperability that reduces switching costs. On centralized platforms, switching costs are high because a user's social graph — their followers and the accounts they follow — cannot be exported and imported. Switching from Twitter to Bluesky means abandoning your followers. On an open protocol, your social graph is portable: switching clients or servers does not require rebuilding your social network. Lower switching costs make it easier for users to move to more ethical alternatives, creating competitive pressure on incumbent platforms to improve their design practices.

The Limitations of Protocol-Level Solutions

Open protocols are not a sufficient solution to the attention economy problem. They address the incentive architecture but not the psychological mechanisms of capture. Several limitations are important to acknowledge.

Network effects are extremely durable. The value of a social network is proportional to the number of people on it (Metcalfe's Law). The incumbent platforms have network effects built over 15–20 years and measuring in billions of users. Open protocol networks with tens of millions of users are not substitutes for networks with billions; they are supplements. The social graph portability argument — that interoperability reduces switching costs — is theoretically correct but practically limited: switching to Mastodon when the people you want to follow are on Twitter still means you can't see their content on Mastodon unless they federate. Most large accounts on centralized platforms have not federated.

Open protocols can host harmful designs. ActivityPub and AT Protocol are neutral transport protocols. Any client built on these protocols can implement engagement-maximizing design: algorithmic ranking, infinite scroll, engagement metric displays. Bluesky's own default algorithm is engagement-ranked. The protocol layer does not guarantee ethical design at the application layer. Open protocols create the possibility of ethical design by removing the financial incentive for engagement maximization; they do not require it.

Moderation is harder at scale without central authority. Centralized platforms can make global content moderation decisions. Federated networks require each server to make its own decisions, and the appeal of federation for users who prefer minimal moderation creates a collection of moderation-light servers that can host harmful content. This is a genuine limitation — not an argument against open protocols, but an argument that open protocols require complementary approaches to content governance that their communities are still developing.

What Open Protocols Have Achieved

Despite their limitations, open protocol networks have achieved several things that the regulatory and design covenant approaches cannot achieve alone.

They have demonstrated proof of concept for non-engagement-maximizing social media at scale. Mastodon's 10 million active users are using a social network that has no behavioral advertising, no engagement-ranking algorithm by default, and no engagement metric pressure on content creators. This user base is a demonstration that a non-engagement-maximizing social network can attract and retain genuine active users — not a dispositive demonstration that it can do so at incumbent scale, but a demonstration nonetheless.

They have created competitive reference points that regulatory frameworks can cite. The EU's Digital Markets Act interoperability requirements draw on the existence of ActivityPub as a demonstration that interoperability standards can be implemented. The argument that social graph interoperability is technically impossible was undermined by the Fediverse's existence. Regulatory arguments for mandatory interoperability are stronger because open protocols have demonstrated that interoperability is achievable.

They have produced a community of practice for ethical social media design. The Fediverse's developer community has built tools, written documentation, and produced design patterns for non-engagement-maximizing social networks. This knowledge base is available to platform developers building on open protocols and to regulatory bodies developing technical standards — a resource that did not exist before the open protocol movement produced it.

What Displacement Would Require

For open protocol networks to displace centralized incumbents rather than merely supplementing them would require changes that the open protocol communities cannot produce independently.

Mandatory interoperability. If centralized platforms were required to implement ActivityPub or an equivalent open protocol for social graph portability — allowing users to follow accounts on centralized platforms from open protocol clients — the switching cost barrier would fall substantially. The EU's Digital Markets Act Article 7 requires large platforms to provide interoperability for messaging services; extending this to social graph portability would produce the structural change that open protocols need to become competitive at scale. This is a legislative intervention, not a technical one.

Data portability enforcement. GDPR Article 20 provides a right to data portability. Most major platforms provide data exports that are technically compliant but practically useless — a large ZIP file of raw data that cannot be imported into another platform. Enforcement of meaningful data portability — specifically, the ability to import a social graph into a competing platform — would reduce switching costs in ways that the current data export regime does not.

Critical mass in specific communities. Network displacement does not require universal adoption; it requires critical mass within specific communities. The academic and researcher community's adoption of Mastodon following the 2022 Twitter disruption was an example of critical mass within a defined community. Additional community-specific critical mass events — in journalism, local government, education — would create more reference points where open protocols are the default rather than the alternative.

What the Protocol Future Demands

The open protocol future demands a recognition that the design covenant approach — voluntary standards adopted by centralized platforms — and the legislative approach — mandatory standards imposed on centralized platforms — are not the only available strategies. The structural strategy — changing the incentive architecture through which engagement maximization becomes commercially rational — is also available, and it requires different tools.

The structural tools are: mandatory interoperability legislation (extending the DMA model to social graph portability), meaningful data portability enforcement (implementing GDPR Article 20 in a way that enables actual competitive switching), public investment in open protocol infrastructure (funding the server costs and development resources that federated networks require), and government adoption of open protocol platforms for official communications (creating legitimizing critical mass in high-visibility contexts).

These tools complement rather than replace the design covenant and regulatory approaches. A future in which centralized platforms operate under mandatory design standards (Legal Architecture series) and in which open protocol alternatives have critical mass in major communities is structurally more stable than a future in which only one of these conditions obtains.

The Design Covenant (DC-005) includes open protocol federation as a recommended — but not required — commitment for signatory platforms. It is recommended because federation represents the strongest structural commitment to cognitive sovereignty: a platform that federates with open protocol networks is a platform that reduces its own lock-in of users, creating genuine competitive pressure to maintain ethical design practices rather than just meeting minimum standards.

Sources and References

• W3C. "ActivityPub." W3C Recommendation, January 23, 2018. activitypub.rocks.
• Bluesky. "AT Protocol Specification." atproto.com, 2023.
• Fiatjaf. "Nostr: A simple, open protocol that enables a truly censorship-resistant and global social network." github.com/nostr-protocol/nostr, 2020.
• Mastodon GmbH. "Annual Report 2024." joinmastodon.org.
• Bluesky. "Year in Review 2024." bsky.app, December 2024.
• Doctorow, Cory. "The Internet Con: How to Seize the Means of Computation." Verso Books, 2023. On interoperability and network power.
• European Commission. Digital Markets Act, Article 7. Interoperability requirements for gatekeepers. Regulation (EU) 2022/1925.
• Regulation (EU) 2016/679 (GDPR), Article 20. Right to data portability.
• Jhaver, Shagun, et al. "Decentralization without Disempowerment: Lessons from the Fediverse." ACM CSCW, 2023.
• Raman, Aravindh, et al. "Challenges in the Decentralised Web: The Mastodon Case." IMC '19, 2019.
• Zulli, Diana, Miao Liu, and Robert Gehl. "Rethinking the 'social' in 'social media': Insights into topology, abstraction, and scale on the Mastodon social network." New Media and Society, 22(7), 2020.
• Metcalfe, Bob. "Metcalfe's Law after 40 Years of Ethernet." IEEE Computer, 46(12), 2013. On network effects.
• Mansoux, Aymeric, and Roel Roscam Abbing. "Seven Theses on the Fediverse and the Becoming of FLOSS." Post-Digital Publishing, 2020.
• Social Web Working Group. W3C documentation archive. On ActivityPub development history.