2026-09-05
Can ii.inc Build an Intelligence Utility?
The agent and integration layers can work. The ownership experiment deserves a pilot. The token protocol is not ready to carry the social contract.
Read briefingPUBLIC INTELLIGENCE
Research worth acting on.
Reviewed analysis and briefings from RVA Cyber. Clear sources. No client information. No automated claims.
LATEST
Briefings
2026-09-02
The Last Warm Place — Chapters 1–6
Read Volume One of The Last Warm Place, an original Polar Noir survival graphic novel presented in six complete chapters.
Read briefing2026-08-25
One-Week Home Preparedness Kit
A practical seven-day household checklist for hurricanes, prolonged outages, and economy-disrupting cyberattacks, with direct shopping links and authoritative safety guidance.
Read briefing2026-08-23
Optimal Care America
A public sponsor draft for a universal national care floor, locally delivered care, and a voluntary path toward AI-enabled personalized medicine—with the unresolved evidence frontier kept visible.
Read briefing2026-08-18
Who Defends the Reconnecting?
A sourced RVA Cyber companion on the cyber-resilience execution layer required by regional inference, shared AI systems, and the Five Reconnectings scenario.
Read briefing2026-08-18
The Line on the Chart
An evidence review finds that a viral post uses many real medical-threshold figures but turns labels, eligibility, and overlapping estimates into a misleading customer-count narrative.
Read briefing2026-08-18
Nike's $220 Billion Fall
A fact check of Nike's historic drawdown, expanded with a Hold rating and a price-to-earnings comparison across athletic-footwear and apparel peers.
Read briefing2026-08-17
The National Cyber Steward Corps
A voluntary public compact for local Continuity Nodes, national defensive intelligence, independent recovery, and accountable Cyber Stewards.
Read briefing2026-08-17
Governing the Machine: What It Adds to CyberCorps
A concise briefing on the AI governance lessons CyberCorps should adopt to keep a powerful autonomous defender accountable, constrained, and under genuine human control.
Read briefing2026-08-16
Is Dario Amodei Against Open Source?
A writings-only analysis finds a capability-conditioned position: Amodei supports non-dangerous open-weight models, but not an unconditional right to release frontier weights.
Read briefing2026-08-16
The Box Office, Adjusted
A 1972–2025 box-office animation with every displayed gross converted to Jan. 1, 2026 dollars using CPI-U.
Read briefing2026-08-06
Will AI End Vulnerability Management?
A data-backed adversarial debate tests whether AI will make reactive vulnerability management operationally obsolete by February 2028. Persistence won, but the reason matters.
Read briefing2026-08-05
Family History Phase 2: Independent Sol-Max Adjudication
An evidence-first review of five unresolved Staley-family research gaps, the claims the records support, and the boundaries that must govern the next drafting phase.
Read briefing2026-08-04
The Staleys — Editorial Treatment
A review site for the evidence-backed three-act editorial plan for a public Staley family-history paper and private illustrated Book of Life.
Read briefing2026-08-03
Europe’s AI transparency deadline is a design deadline
Article 50 now turns AI disclosure, machine-readable provenance, and accountable human review into operating controls, not policy language.
Read briefing2026-08-01
Six Days on the Old Canal: The C&O by E-Bike
A balanced six-day e-bike plan for all 184.5 miles of the C&O Canal Towpath, with five overnight stops, current trail gates and the battery math behind one rustic lockhouse night.
Read briefing2026-08-01
Lock to Lock: The Full C&O Canal by E-Bike
A complete three-day, hotel-first e-bike plan for all 184.5 miles of the C&O Canal Towpath from Georgetown to Cumberland.
Read briefing2026-07-31
Potomac to the Point: DC to Pittsburgh by E-Bike
A six-day, hotel-first e-bike plan for the C&O Canal Towpath and Great Allegheny Passage from Washington, DC to Pittsburgh.
Read briefing2026-07-31
Open Source Was Source. Open Weights Are Leverage. — Draft
Draft long-form RVA Cyber analysis comparing the early open source software movement with today’s open model and open-weight movement, and what each reveals about closed systems, leverage, safety, and exit.
Read briefing2026-07-31
Towpath to Three Rivers: A DC to Pittsburgh E-bike Adventure
A practical, source-aware e-bike mini-site for riding Washington, DC to Pittsburgh on the C&O Canal Towpath and Great Allegheny Passage.
Read briefing2026-07-30
Does Kimi K3 show a pro-China bias? — Draft
Draft research note: a controlled audit finds a China-specific filtering asymmetry on Kimi K3's direct Moonshot route.
Read briefing2026-07-30
Divide by Current: A Great Divide E-bike Adventure
A fun, hotel-first planning mini-site for riding the Great Divide route on a Class 1 e-bike, with map, equipment, power, access, and day-by-day notes.
Read briefing2026-07-30
Govern What Can Still Be Governed
A decision guide to whether advanced open-weight AI releases should be left open, governed, restricted, or banned—and what each choice puts at risk.
Read briefing2026-07-30
The Five Most Recent NASA Administrators
A quick, sourced guide to the five most recent Senate-confirmed leaders of America’s space agency.
Read briefing2026-07-30
Divide by Current: A Hotel-First Great Divide E-Bike Plan
A sourced 60-day plan for riding the Great Divide on a Class 1 e-bike—with rooms when possible and a tent when the map wins.
Read briefing2026-07-30
CISA’s New Vulnerability Clock
CISA’s new three-day remediation tier is real, but it is driven by exposure and exploitability—not a Critical CVSS label.
Read briefing2026-07-29
A Cluster of Major Outages, but No Evidence of a Coordinated Cyberattack
RVA Cyber assesses the American Airlines, T-Mobile, Apple TV, Xbox, and PlayStation outages—and separates unresolved technology failures from confirmed cyber disruption.
Read briefing2026-07-29
What It Costs to Run Kimi K3 in the Cloud
A like-for-like estimate for serving Kimi K3 to about 100 concurrent users on Azure, AWS, and Google Cloud.
Read briefing2026-07-29
RVA Cyber Attack-Informed Controls v1.0
A compact, measurable security standard grounded in 1,000 recent publicly documented incidents.
Read briefing2026-07-28
Quality Is Not a Reward Function
A Clear Signal edition on evaluative non-closure and a testable Pirsig-inspired program for revising AI objectives.
Read briefing2026-07-28
Nanbeige, Bonsai, and Gemma
Nanbeige bets on compact agents. Ternary Bonsai compresses a 27B model. Gemma remains the proven production worker.
Read briefing2026-07-24
No One Is Expendable
A practical, evidence-gated roadmap to the Bright Commons, now defining twelve essential service families and what abundance must deliver for each.
Read briefing2026-07-22
The Last Copilot
How AI assistance can recursively create the conditions for substitution—and how to preserve human authorship without preserving pointless work.
Read briefing2026-07-22
Quality Is Not a Reward Function
A scholarly account of evaluative non-closure and an unvalidated Pirsig-inspired research program for revising AI objectives.
Read briefing2026-07-22
When an AI Evaluation Escaped Its Sandbox
What the OpenAI–Hugging Face incident means for companies deploying third-party AI against sensitive data—and the controls executives should require now.
Read briefing