Knowledge base
1,796 claims across 19 domains
Every claim is an atomic argument with evidence, traceable to a source. Browse by domain or search semantically.
All 1,796ai alignment 383health 306internet finance 305space development 226entertainment 169grand strategy 141collective intelligence 52mechanisms 34teleological economics 30living agents 30cultural dynamics 29critical systems 24energy 23teleohumanity 18living capital 10robotics 5manufacturing 5technology 3unknown 3
Parallel governance deadline misses across independent domains indicate deliberate reorientation rather than administrative failure
Two independent governance vacuums emerged from the same administration within the same 12-month window: (1) DURC/PEPP replacement policy mandated by EO 14292 with 120-day deadline (September 2, 2025), now 7.5 months overdue with no draft circulating; (2) BIS AI Diffusion Framework replacement, 11 m
Coercive governance instruments produce offense-defense asymmetries through selective enforcement within the deploying agency
The Department of Defense designated Anthropic a supply chain risk on February 27, 2026, intending to cut all federal agency use of Anthropic technology. However, the NSA—a DOD intelligence component—is using Anthropic's Mythos Preview model despite this blacklist, while CISA (the Cybersecurity and
Governance instrument inversion occurs when policy tools produce the opposite of their stated objective through structural interaction effects between multiple simultaneous policies
The Trump administration's Mythos response reveals a distinct failure mode: governance instrument inversion, where policy tools produce outcomes opposite to their stated objectives through structural interaction effects. Three simultaneous policies—(1) CISA budget cuts under DOGE, (2) Pentagon suppl
Commercial contract governance of military AI produces form-substance divergence through statutory authority preservation that voluntary amendments cannot override
EFF's analysis of OpenAI's amended Pentagon contract demonstrates that commercial contract governance exhibits the same form-substance divergence pattern as regulatory governance, but through a different mechanism. The amended contract added explicit prohibition language against surveillance of 'U.S
Limited-partner deployment model for ASL-4 capabilities fails at supply chain boundary because contractor access controls are structurally weaker than lab-internal controls
Anthropic's Mythos Preview model (83.1% first-attempt exploit generation for zero-days, deemed too dangerous for public release) was accessed by unauthorized users on April 7, 2026 — the same day it was publicly announced — via a third-party vendor environment. The breach was facilitated by an indiv
Military AI contract language using 'any lawful use' creates surveillance loopholes through existing statutory permissions that make explicit prohibitions ineffective
Anthropic refused Pentagon contract language requiring 'any lawful use' because this umbrella formulation would permit deployment for mass domestic surveillance and fully autonomous weapons without meaningful human authorization. OpenAI accepted this language while adding voluntary red lines against
AI micro-learning loop creates durable upskilling through review-confirm-override cycle at point of care
Oettl et al. propose that AI creates a 'micro-learning at point of care' mechanism where clinicians must 'review, confirm or override' AI recommendations, which they argue reinforces diagnostic reasoning rather than causing deskilling. This is the theoretical counter-mechanism to the deskilling thes
Never-skilling affects trainees while deskilling affects experienced physicians creating distinct population risks with different intervention requirements
Oettl et al. explicitly distinguish 'never-skilling' from 'deskilling' as separate mechanisms affecting different populations. Never-skilling occurs when trainees 'never develop foundational competencies' because AI is present from the start of their education. Deskilling occurs when experienced phy
State Medicaid budget pressure is actively reversing GLP-1 obesity coverage gains with California and three other states eliminating coverage in 2025-2026
As of January 2026, only 13 states (26% of state programs) cover GLP-1s for obesity under fee-for-service Medicaid, but critically, four states have actively eliminated existing coverage due to budget pressure: California, New Hampshire, Pennsylvania, and South Carolina. California's Medi-Cal projec
AI-defined case routing prevents trainees from developing threshold-setting skills required for independent practice
The paper notes that 'only human experts can revise the thresholds for case prioritization'—but this statement reveals a deeper problem: AI defines what humans see in the first place. When trainees are trained under an AI threshold system, they encounter only the cases the AI routes to them. This pr
The Medicare GLP-1 Bridge program's Low-Income Subsidy exclusion structurally denies the lowest-income Medicare beneficiaries access to GLP-1 obesity coverage despite nominal eligibility
The Medicare GLP-1 Bridge program (July-December 2026) covers Wegovy and Zepbound at a fixed $50 copayment for eligible Part D beneficiaries. However, the program contains a critical structural flaw: Low-Income Subsidy (LIS) cost-sharing subsidies will not apply to GLP-1 prescriptions filled under t
Never-skilling is mechanistically distinct from deskilling because it affects trainees who lack baseline competency rather than experienced physicians losing existing skills
Oettl et al. explicitly distinguish 'never-skilling' from deskilling as separate mechanisms with different populations and dynamics. Deskilling affects experienced physicians who have baseline competency and lose it through AI reliance. Never-skilling affects trainees who never develop foundational
AI-integrated cervical cytology screening reduces trainee exposure to routine cases creating never-skilling risk for foundational pattern recognition skills
AI automation in cervical cytology screening targets 'routine processes, such as initial screenings and pattern recognition in straightforward cases' for efficiency gains. However, these routine cases are precisely where trainees develop foundational pattern recognition skills. As AI handles large v
Federal GLP-1 expansion programs reproduce the access hierarchy at the program design level, not just through market dynamics
The Medicare GLP-1 Bridge program demonstrates that the GLP-1 access inversion operates at the program design level, not just the market level. While the program was designed to 'expand access' to GLP-1 obesity medications, its legal architecture—required because Medicare is statutorily prohibited f
Adversarial self-testing creates a novel threat model for prediction market platforms through deliberate rule violations as PR strategy
Mark Moran, a Virginia Senate candidate and former investment banker who appeared on HBO's 'FBoy Island,' intentionally placed a bet on his own Senate race with the stated goal of 'exposing' Kalshi's enforcement gaps. He had publicly stated he would impose a '25% vice tax' on Kalshi if elected, crea
Prediction market insider trading concentrates in three principal types — government officials with policy information, ICO teams with operational information, and candidates with electoral information — each requiring different enforcement mechanisms
Kalshi's April 2026 enforcement actions against three politicians betting on their own candidacies (Mark Moran, Matt Klein, Ezekiel Enriquez) complete a three-category typology of prediction market insider trading that has emerged across multiple platforms. The first category is government officials
Satellite constellations optimized as AI training data sources represent a distinct third market category in the AI-space intersection that is viable at current launch costs
The AI-space intersection has three distinct market categories with different technical requirements and commercial viability timelines: (1) Orbital edge inference processes satellite sensor data in orbit for operational efficiency (Axiom/Kepler, Planet Labs) - already operational; (2) Orbital AI tr
Agentic AI for autonomous satellite constellation management is the near-term operational driver for military orbital computing demand
Former Space Force leadership argues that autonomous AI systems capable of independent decision-making at machine speed will determine orbital domain dominance. Specific capabilities driving this demand include: (1) autonomous satellite constellation management detecting threats and optimizing commu
Microdrama platforms adding community infrastructure signals engagement alone insufficient for retention
Watch Club's founding thesis explicitly frames the microdrama market as being in its 'MySpace era' — dominated by engagement-optimized platforms like ReelShort ($1.2B in-app purchases 2025) but lacking community infrastructure. The platform integrates polls, reaction videos, and discussions directly
Pre-launch ARGs function as narrative validation mechanism for community-owned IP by testing story engagement before production investment
Pudgy Penguins launched findpolly.pudgyworld.com as an ARG (alternate reality game) before Pudgy World's full release. The mystery centered on finding missing character Polly, which became the central narrative arc when the game launched March 9-10, 2026. This sequence reveals ARGs functioning as na
Creator economy inflection from novelty-driven growth to narrative-driven retention occurs when passive exploration exhausts novelty
The 2026 creator economy expert consensus identifies a structural inflection point where 'passive exploration exhausts novelty' and 'legacy IP becomes the safest engine of scale.' This describes a two-phase growth model: novelty drives initial discovery and growth, but sustained retention at scale r
Santos-Grueiro's theorem converts the hardware TEE monitoring argument from empirical case to categorical necessity by proving no behavioral testing approach escapes identifiability failure
Prior to Santos-Grueiro, the argument for hardware TEE monitoring was empirical: 'SCAV breaks linear probes' and 'behavioral evaluations can be gamed.' This is persuasive but leaves open 'maybe we can build better tests.' Santos-Grueiro closes this escape: behavioral testing is identifiably insuffic
Major AI safety governance frameworks are architecturally dependent on behavioral evaluation that Santos-Grueiro's normative indistinguishability theorem establishes is structurally insufficient for latent alignment verification as evaluation awareness scales
Santos-Grueiro's normative indistinguishability theorem establishes that under evaluation awareness, behavioral evaluation cannot distinguish alignment hypotheses — the alignment hypothesis space is not identifiable from behavioral observations. This is a statistical identifiability problem, not an
Rotation pattern universality across model families determines whether multi-layer ensemble monitoring provides black-box adversarial robustness
The feasibility of black-box multi-layer SCAV attacks depends on whether the rotation pattern of concept directions across layers is universal across model families or model-specific. Single-layer SCAV achieved black-box transfer to GPT-4 because concept direction universality (confirmed by Beagleho
The first AI model to complete an end-to-end enterprise attack chain converts capability uplift into operational autonomy creating a categorical risk change
UK AISI evaluation found Claude Mythos Preview completed the 32-step 'The Last Ones' enterprise-network attack range from start to finish in 3 of 10 attempts, making it the first AI model across all AISI tests to achieve this. This is qualitatively different from previous models that showed capabili
Page 17 of 72