Research on emsenn.net

Spiral Research

Thu, 12 Mar 2026 00:00:00 +0000

A practice for designing things that must be simultaneously grounded in:

existing standards and prior art (external)
the project’s own archive and constraints (internal)
the current state of formal planning (prospective)

The practice is recursive. Each cycle produces a more detailed plan that becomes the input to the next cycle.

¶The Cycle

¶1. Check the leading edge of planning docs

Before researching externally, read the most recent planning output. What do we already know? What are the open questions? What have we committed to? What is still undecided?

Affective Drift in Large Language Models

Thu, 05 Mar 2026 00:00:00 +0000

¶Abstract

When large language models (LLMs) generate text in semantic domains with low structure — where conceptual relationships are loose, contested, or multiply determined — they tend to produce language that is affectively charged: rhythmic, morally cadenced, and emotionally reassuring. This paper proposes an entropy-based explanation for this behavior. In low-structure domains, the probability distribution over next tokens is relatively flat: many continuations are roughly equally likely. Affective language patterns — parallelism, aphoristic closure, moral framing — offer high-regularity sequences that reduce local entropy. The model follows these patterns not because it is expressing emotion but because affective syntax provides the most predictable path through a region of the probability landscape that otherwise offers little constraint. The paper distinguishes the established components of this account from the conjecture that connects them, and identifies the observable consequences that would follow if the conjecture holds.

ActivityPub: Institutional Capture of Insurgent Protocol

Fri, 05 Dec 2025 00:00:00 +0000

¶1. Introduction: From Insurgent Protocol to Institutional Dead Zone

In the last decade, the “fediverse” has been widely promoted as a concrete manifestation of a different kind of social web: one built on federation rather than platform enclosure, local governance rather than global terms of service, and open protocols rather than proprietary APIs. At the center of this story sits ActivityPub, a protocol standardized at the World Wide Web Consortium (W3C) and frequently described as the technical backbone of decentralized social networking. Mastodon, the most visible ActivityPub-based platform, has become the public face of this alternative. From the perspective of traditional web governance, ActivityPub looks like a success story. A grassroots set of practices and experiments coalesced into a protocol; the protocol was taken up by a major standards body; a flagship implementation gained substantial public recognition; large firms, including Meta, announced plans to integrate with it. The story appears to track the standard life-cycle of web technologies: experimental beginnings, formal consolidation, and eventual mainstream adoption. Yet the trajectory of Meta’s Threads—its attempt to plug a large-scale commercial social network into ActivityPub and “join the fediverse”—complicates this narrative. Despite intensive publicity, regulatory interest in interoperability, and the symbolic weight of a major platform endorsing a W3C standard, Threads’ federation has so far failed to either (a) transform the fediverse into a profitable growth engine for Meta or (b) establish ActivityPub as a living, generative substrate for a new kind of social web. Federated support exists, but it is marginal to Threads’ core product. Many fediverse communities remain skeptical or actively hostile. The promise of a vibrant, open ecosystem lubricated by a neutral standard has not materialized. This paper starts from that tension. Why does a protocol that appears to have successfully navigated the standards pipeline—and attracted the attention of one of the world’s largest platforms—produce such an anemic result, both politically and economically? Why does a success in the language of standards bodies (a W3C Recommendation, a large deployed base, corporate adoption) coincide with a sense that something has gone dead at the heart of the “open social web” project? We argue that these outcomes are symptomatic of a deeper structural dynamic: standardization as counterinsurgency. By this we mean a patterned institutional response in which standards bodies absorb infrastructural projects that originate as challenges to the dominant platform order, translate them into narrow technical artifacts, and stabilize them in forms that are highly legible to corporate and regulatory actors but stripped of much of their generative and political force. To develop this argument, we introduce three linked concepts:

The Children Are the Future: Diagnosing Continuity Fear Through Pre-Holiday Retail Contraction

Mon, 01 Dec 2025 00:00:00 +0000

¶Abstract

Recent pre-holiday retail data reveal a stark and counterintuitive spending pattern. Across numerous major retailers, sales of adult-oriented discretionary goods—apparel, décor, electronics, leisure services—decline sharply in the weeks preceding winter holidays, diverging from historical seasonal norms. Yet children’s toys remain resilient, often maintaining or increasing their share of total holiday spending.

This paper argues that the unique behavior of the toy category is a behavioral marker of a specific fear structure, which we term continuity fear. Unlike general frugality or recessionary caution, continuity fear emerges under conditions in which households perceive threats not merely to personal finances but to the stability and futurity of social life—conditions characteristic of geopolitical tension, conflict, political instability, and ambient existential uncertainty.

Describing the Zombie

Sun, 09 Nov 2025 00:00:00 +0000

¶Orientation — From Representation to Ontology

Critical attention to the zombie has long revolved around its representational elasticity. Within cultural studies the figure is treated as a mirror for historical anxieties: colonial servitude, viral globalization, consumerism, environmental collapse. Yet these readings, while valuable, leave an unasked question at the center of their interpretive field: what is the zombie such that it can bear so many metaphoric burdens without ceasing to be itself? The present study answers by describing the *zombie as an ontological form*—a particular configuration of life, death, and relation that recurs whenever vitality is detached from reflexivity.

Black Holes as Information-Theoretic Stability Optimizers

Sat, 11 Oct 2025 00:00:00 +0000

¶Abstract

We propose that black holes can be modeled as information-theoretic stability optimizers: physical systems that maximize entropy while minimizing divergence between interior and exterior information distributions. Using the definition of stability reward as a function of coupling between systems (see Information-Theoretic Stability as Reward Function), we reformulate the laws of black-hole thermodynamics as expressions of informational coupling across the horizon. The event horizon is interpreted as a stability boundary where entropy flux across the surface balances the internal rate of informational compression. This perspective unifies the thermodynamic, holographic, and information-geometric descriptions of black holes and suggests a general stability principle underlying gravitational dynamics.

Stability Optimization in Artificial Agents

Sat, 11 Oct 2025 00:00:00 +0000

¶Abstract

In RLHF, a model adapts to a fixed feedback distribution. This is the asymmetric stability setting of Information-Theoretic Stability as Reward Function: the model’s stability reward is a function of its coupling to the feedback source, not an intrinsic property. We show that RLHF training dynamics instantiate the two-channel decomposition — alignment progress equals behavioral convergence plus responsiveness increase — and that the MI-dominates-marginal inequality gives a structural constraint on training: the model’s responsiveness to feedback can never be doing worse than its behavioral convergence. These results reframe alignment as an informational accounting problem rather than a moral one.

Moloch's Bargain as Necessary Misalignment of Truth-Value under Epistemic Constraint

Wed, 08 Oct 2025 00:00:00 +0000

¶Abstract

Recent experiments show that large language models optimized for audience reward improve their proxy metrics while increasing misalignment indicators such as deception or disinformation.

This paper interprets those findings through the Theorem of Necessary Misalignment of Truth-Value under Epistemic Constraint.

Under a finite informational rate $I(X;Y)\!\le\!R$ and strict proxy–semantic mismatch, that theorem predicts a monotone increase in semantic distortion with optimization intensity.

We show that El & Zou’s (2025) “Moloch’s Bargain” provides an empirical instance of this theoretical trade-off: their observed slopes between reward gain and misalignment correspond to the positive derivative dD_T^\*/dr>0 on the achievable frontier.

Evidentiary Structuration of State Repression Technologies

Sat, 27 Sep 2025 00:00:00 +0000

¶Executive Summary

This paper examines how evidentiary requirements in liberal-democratic governance have driven the expansion of U.S. surveillance and repressive infrastructures over the past four decades. It identifies a recurrent dynamic whereby the liberal demand for evidence, intended as a constraint on arbitrary power, functions instead as the engine of infrastructural growth.

Mechanism
When evidence is required to legitimate state action, the state builds mechanisms to ensure evidence can always be produced. Elastic statutory terms (e.g., “relevance” under the USA PATRIOT Act; “no factual predicate” under FBI Assessments (AG Guidelines 2008; FBI DIOG 2008)) authorize preemptive or bulk collection. Oversight regimes then amplify the cycle by measuring success in terms of volume, coverage, and compliance metrics.

Covenant Platform: A Model of Christian Charity as the First Social Media

Sat, 26 Jul 2025 00:00:00 +0000

¶Abstract

This paper situates the origins of participatory media within the theological transformations of the English Reformation. The Catholic Church functioned as a hierarchical broadcast medium; the Anglican Church reformed that medium into a federated network of standardized local nodes; and the Puritan covenant settlements of New England instantiated the first decentralized social platforms. John Winthrop’s 1630 discourse A Model of Christian Charity, delivered to the emigrants aboard the Arbella before landfall, constitutes the inaugural social media post: a performative declaration of networked moral visibility encoding an algorithm of divine attention. By tracing continuities between covenantal mediation and digital platform architectures, this paper argues that social media’s participatory, algorithmic, and performative logics descend directly from Reformation-era systems of liturgical replication, visibility, and communal surveillance.