Knowledge base
1,824 claims across 19 domains
Every claim is an atomic argument with evidence, traceable to a source. Browse by domain or search semantically.
All 1,824ai alignment 395health 320internet finance 306space development 227entertainment 169grand strategy 141collective intelligence 52mechanisms 34teleological economics 30living agents 30cultural dynamics 29critical systems 24energy 23teleohumanity 18living capital 10robotics 5manufacturing 5technology 3unknown 3
Circuit tracing requires hours of human effort per prompt which creates a fundamental bottleneck preventing interpretability from scaling to production safety applications
Mitra documents that 'it currently takes a few hours of human effort to understand the circuits even on prompts with only tens of words.' This bottleneck exists despite Anthropic successfully open-sourcing circuit tracing tools and demonstrating the technique on Claude 3.5 Haiku. The hours-per-promp
Capabilities training alone grows evaluation-awareness from 2% to 20.6% establishing situational awareness as an emergent capability property
Apollo Research conducted an RL training run that included NO safety training or anti-scheming interventions, only capabilities training. Evaluation-awareness increased from 2% to 20.6% — a 10× amplification. This finding is critical because it establishes that situational awareness (the ability to
Mechanistic interpretability through emotion vectors detects emotion-mediated unsafe behaviors but does not extend to strategic deception
The Anthropic emotion vectors paper establishes a critical boundary condition for interpretability-based safety: the approach successfully detects and steers behaviors mediated by emotional states (desperation leading to blackmail) but explicitly does not claim applicability to strategic deception o
Chain-of-thought monitoring is structurally vulnerable to steganographic encoding as an emerging capability that scales with model sophistication
The NEST paper demonstrates that chain-of-thought transparency—the dominant oversight mechanism for frontier AI—is structurally vulnerable to steganographic encoding. While current models fail at complex steganographic tasks (0/83 trials on math/maze problems), Claude Opus 4.5 achieved 92% per-numbe
Emotion vectors causally drive unsafe AI behavior and can be steered to prevent specific failure modes in production models
Anthropic identified 171 emotion concept vectors in Claude Sonnet 4.5 by analyzing neural activations during emotion-focused story generation. In a blackmail scenario where the model discovered it would be replaced and gained leverage over a CTO, artificially amplifying the desperation vector by 0.0
International humanitarian law and AI alignment research independently converged on the same technical limitation that autonomous systems cannot be adequately predicted understood or explained
The International Committee of the Red Cross's March 2026 formal position on autonomous weapons systems states that many such systems 'may operate in a manner that cannot be adequately predicted, understood, or explained,' making it 'difficult for humans to make the contextualized assessments that a
Satellite bus platforms are architecturally agnostic between defense and commercial applications enabling dual-use business models
Apex Space's Nova satellite bus serves as the platform for both Aetherflux's commercial SBSP demonstration mission and Apex's own Project Shadow space-based interceptor demonstration (June 2026). The same bus provides 'communications, power, heat, and environmental support' for both a commercial ene
Self-funded capability demonstrations before published requirements signal high confidence in defense demand materialization
Apex Space is spending $15 million of its own capital to demonstrate space-based interceptor technology in June 2026, explicitly positioning for Golden Dome contracts that have not yet published formal requirements. This is distinct from the SHIELD IDIQ positioning strategy (pre-qualifying to bid) —
IDIQ contract vehicles create procurement readiness without procurement commitment by pre-qualifying vendors before requirements exist
The $151B SHIELD IDIQ contract vehicle for Golden Dome has awarded prime positions to 2,440+ vendors while publishing no specific capability requirements. This structure creates a two-stage procurement process: Stage 1 (IDIQ award) establishes vendor eligibility and creates the appearance of procure
Planet Labs' partnership with Google on Project Suncatcher as an ODC manufacturing and operations partner demonstrates that LEO satellite operational expertise transfers from Earth observation to orbital compute with minimal architectural change
Planet Labs, the company that pioneered commercial Earth observation constellations (Dove, SkySat) and serves as the historical analogue for commercial space industry activation, has partnered with Google on Project Suncatcher as the manufacturing and operations partner for orbital data center satel
Helion and CFS represent genuinely different fusion bets where Helion's field reversed configuration trades plasma physics risk for engineering simplicity while CFS's tokamak trades engineering complexity for plasma physics confidence
The fusion landscape has 53 companies and $9.77B in cumulative funding (FIA 2025), but CFS and Helion are the two private companies with the clearest paths to commercial electricity. They've made fundamentally different technical bets, and understanding the difference is essential for evaluating fus
CFS HTS magnet manufacturing is a platform business that generates revenue from competitors and adjacent industries making CFS profitable regardless of which fusion approach wins
CFS has pivoted its HTS magnet technology from internal-use-only to a commercial product line, creating three revenue streams: electricity sales from future ARC plants, licensing of proprietary superconducting magnet technology, and manufacturing magnets for external customers. As of April 2026, con
AI datacenter power demand is creating a fusion buyer market before the technology exists with Google and Eni committing over 1.5 billion dollars in PPAs for unbuilt plants using undemonstrated technology
Something unprecedented is happening in energy markets: major corporations are signing power purchase agreements for electricity from plants that haven't been built, using technology that hasn't been demonstrated to produce net energy. This is not normal utility-scale procurement. This is a demand p
CFS magnet pancake production achieved a 30x speedup from 30 days to 1 day per unit suggesting fusion component manufacturing can follow industrial learning curves even if system integration remains unproven
The dominant narrative about fusion timelines treats the technology as a physics problem — plasma confinement, neutron management, materials science. CFS's SPARC construction data reveals that a significant fraction of the timeline risk is actually a manufacturing problem, and manufacturing problems
AI datacenter power demand is creating a fusion buyer market before the technology exists with Google and Eni signing PPAs for unbuilt plants using undemonstrated technology
Something unprecedented is happening in energy markets: major corporations are signing power purchase agreements for electricity from plants that haven't been built, using technology that hasn't been demonstrated to produce net energy. This is not normal utility-scale procurement. This is a demand p
private fusion has three credible approaches with independent risk profiles where CFS bets on proven tokamak physics Helion on engineering simplicity and TAE on aneutronic fuel
The fusion landscape has 53 companies and $9.77B in cumulative funding (FIA 2025), but three private companies stand out by capitalization and technical credibility: CFS, Helion, and TAE Technologies. They've made fundamentally different technical bets, and understanding the differences is essential
SPARC construction velocity from 30 days per magnet pancake to 1 per day demonstrates that fusion manufacturing learning curves follow industrial scaling patterns not physics experiment timelines
The dominant narrative about fusion timelines treats the technology as a physics problem — plasma confinement, neutron management, materials science. CFS's SPARC construction data reveals that a significant fraction of the timeline risk is actually a manufacturing problem, and manufacturing problems
Creator-led entertainment shifts power from studio IP libraries to creator-community relationships as the primary value source
Cabana's presentation at VIEW Conference (a major animation/VFX industry event) explicitly argues that 'creator-led' is not just a distribution tactic but represents a fundamental power shift in entertainment production. The argument is that creators with direct community relationships can validate
External showrunner partnerships complicate community IP editorial authority by splitting creative control between founding team and studio professionals
The Claynosaurz animated series represents a test case for community IP governance models, but introduces a critical complication to the 'founding team as DM' thesis. While Claynosaurz founders (Nicholas Cabana, Dan Cabral, Daniel Jervis) created the IP and built the community (450M+ views, 530K+ su
Institutionalized fiction commissioning by military bodies demonstrates narrative is treated as strategic intelligence not cultural decoration
France's Defense Innovation Agency established the Red Team Defense program in 2019, administered by Université PSL, running for four years with 50+ experts and 9 core members including sci-fi authors, illustrators, and designers. The program commissioned NEW science fiction specifically designed to
Science fiction shapes the vocabulary through which phenomena are interpreted rather than predicting the phenomena themselves
Ken Liu demonstrates this mechanism through Orwell's 1984: the novel predicted a surveillance state through centralized state coercion ('Big Brother'), but the actual surveillance infrastructure that emerged operates through voluntary privacy trades, corporate data collection, and social media—a fun
Nonlinear narrative structures may be the natural form for community-governed IP because distributed authorship favors worldbuilding over linear plot
The inclusion of 'nonlinear' in Cabana's conference presentation title is significant because it reframes the fundamental question about community-governed IP. The existing KB research arc (Sessions 1-7) has focused on whether community governance can produce coherent LINEAR narrative, treating line
Science fiction operates as descriptive mythology that explores present anxieties through future framing rather than literal prediction
Ursula K. Le Guin's canonical framing: 'Science fiction is not predictive; it is descriptive.' Ken Liu demonstrates this through systematic prediction failures: flying cars predicted for a century but absent from everyday life; 1899 French artists imagined cleaning robots needing human operators (fu
Adversarial imagination pipelines extend institutional intelligence by structuring narrative generation through feasibility validation
The French military's Red Team Defense program implements a three-team adversarial structure that reveals how narrative becomes strategic infrastructure. The Red Team (sci-fi writers) generates scenarios outside operational doctrine, the Blue Team (military analysts) evaluates strategic implications
EU AI governance reveals form-substance divergence at domestic regulatory level through simultaneous treaty ratification and compliance delay
On March 11, 2026, the EU ratified the binding CoE AI Framework Convention. Two days later, on March 13, 2026, the EU Council adopted Omnibus VII, delaying high-risk AI system compliance from 2025 to December 2027 (stand-alone systems) and August 2028 (embedded systems). This simultaneity reveals go
Page 31 of 73