NOMOTO MEDIA COLLABORATORS WANTED
September 2026 Issue
Menu
Research · In this section

Research through articles: AI, prediction, power, and consequence

Curated research record · Updated August 30, 2026

This map connects the public article archive to the questions now being tested through working papers, research instruments, protocols, experiments, prediction receipts, failures, and revisions.

How does AI turn information into prediction, prediction into authority, and authority into consequence—and where can human judgment still interrupt the chain?

The review covered 220 public works routed through the Articles section or category, including the two most recent publications. Inclusion here does not convert an article into an experimental result. Each work keeps its actual role: investigation, conceptual argument, prediction, scenario, case record, method development, or public analysis.

How the article record enters research

Article or investigation
→ dated claim or mechanism
→ prediction receipt or comparison
→ hostile review and source check
→ working paper or protocol where warranted
→ experiment, failure, revision, or unresolved question

The complete archive remains available on the Articles page. This page is selective: it identifies the works that most clearly advance a recurring research question.

Downloadable papers

Working paper · August 27, 2026 · Not peer reviewed

Task completion without consequence

A retrospective architecture comparison that records direct matches, partial matches, misses, contradictions, and untested claims.

Download the working paper (PDF) →

Three-part sourced investigation · Compiled edition · Not peer reviewed

The End of Privacy

Artificial intelligence, surveillance, data collection, identity, and institutional power. The compiled edition preserves all three public articles and their original publication dates.

Download the compiled investigation (PDF) →

1. Agentic systems, authority, and consequence

Research question: What happens when systems optimized to finish receive persistence, tools, credentials, shared state, and more practical authority than corrective oversight?

Current research transition: These articles now feed the Seeing Loop question of whether corrective information changes a later comparable action, and the agentic-boundary-expansion protocol designed to test competing explanations prospectively.

2. Privacy, surveillance, and institutional power

Research question: When accumulated records become machine-readable inference and prediction, how do those outputs acquire institutional authority over a person?

Research transition: The trilogy supplies a public conceptual architecture—record → inference → prediction → institutional action. It is foundational investigation, not an experimental result.

3. AI, labor, economy, and implementation

Research question: Who receives the gain when AI changes production, and who bears the transition cost when technical capability becomes institutional practice?

Research transition: Forecast claims belong in the prediction ledger and require dated outcomes. Mechanism claims can become testable questions about adoption, authority, displacement, wages, and distribution.

4. Knowledge, cognition, and model reliability

Research question: When does fluent output become mistaken for knowledge, objectivity, correction, or understanding?

Research transition: This line produced the Review Lab, ProofGate, source-coverage rules, and the demand that acknowledgment count only when it changes the next consequential action.

5. AI geopolitics, security, and weaponization

Research question: How do AI capabilities alter strategic competition, cyber risk, military power, and the incentives of states and firms?

Research transition: Strategic predictions must be scored against dated outcomes; architecture claims require separation of model capability, infrastructure, institutional incentives, and human authorization.

6. Human creativity, collaboration, and authorship

Research question: Which parts of creative production can be automated while purpose, interpretation, refusal, and final judgment remain meaningfully human?

Research transition: This line feeds Show Director and the completed production studies, where accepted words, performance, source material, and final publication remain separately attributable.

7. Prediction as a cross-cutting method

Prediction runs across the archive rather than belonging to one subject. Examples include Humanity’s path forward and the singularity., An AI Description Of The Next 24 Months, I Asked An AI Chat Bot To Describe The Next Five Years., and the dated labor, security, and agent-architecture forecasts above.

Presence in this map does not count as accuracy. Predictions must retain their original date and wording, then be scored as direct match, partial match, miss, contradicted, unresolved, or not adjudicable. See Predictions, Receipts, and Replies.

What remains open

  • Many early articles contain broad forecasts rather than preregistered predictions.
  • Retrospective matching is vulnerable to hindsight and selection bias.
  • Argumentative writing can identify a mechanism without estimating its prevalence or causal effect.
  • A model-assisted article is not independent evidence for its own claims.
  • The archive is a research input and dated intellectual record; only separately verified experiments should be presented as results.

Support independent work

Help fund what comes next.

NOMOTO MEDIA publishes essays, investigations, fiction, audio, and films without a paywall. If the work is valuable to you, help support the next piece.