6th October 2026
How Inventing New Words Helps Humans Work With AI
Today’s science highlights the need for clear boundaries to manage complex systems. As artificial intelligence integrates into our daily work, researchers are discovering how inventing new words can significantly improve human-AI collaboration . To keep these systems safe, experts are building frameworks to evaluate AI control and prevent misleading outputs . Beyond technology, scientists are creating methods to track how human knowledge evolves across different fields , . This push for precision also extends to history and physics, where advanced mathematical models are being used to detect long-term historical cycles , , and filter noise in physical measurements . Finally, economists are demanding stricter statistical standards for fair real estate valuations , , , .
Top 9 topics by publication and citation volume
Property Rights and Legal Doctrine17
Model Reduction and Neural Networks5
Structural Load-Bearing Analysis5
Ethics and Social Impacts of AI4
Housing Market and Economics4
Environmental Sustainability in Business3
Educational Games and Gamification3
Microgrid Control and Optimization3
Geology and Paleoclimatology Research3
Extended Breakdown↓
As scientific disciplines grow increasingly intertwined with computational models, a critical paradigm shift is occurring: the systematic formalization of human-designed boundaries across law, engineering, and artificial intelligence. Rather than relying on heuristic or ad-hoc methodologies, modern research is building rigorous, bounded frameworks to govern complex systems, validate emergent knowledge, and mathematically define the limits of predictability.
This structural turn is highly visible in the governance of artificial intelligence, where researchers are moving away from simple technical optimization toward robust, human-centered control. To prevent failures in advanced autonomous systems, the "Human Control Model" establishes a multi-dimensional framework to evaluate agent safety, distinguishing genuine authorization from mere preference accuracy. This oversight is vital as large language models (LLMs) are increasingly deployed to reason about complex systemic issues like sustainability; evaluations of AI-generated explanations of the circular economy reveal that these systems often suffer from an operational indifference to truth and lack genuine moral agency. To manage the language we use to collaborate with these machines, the "Systematic Neology" framework provides a structured methodology to evaluate whether introducing new terminology in human-AI workflows adds genuine utility beyond existing linguistic options.
This effort to standardize and track knowledge creation extends into meta-science. As scientific fields evolve, maintaining cognitive continuity is essential. The "Cognitive Continuity Methodology" (CCM) provides a formal framework to test and trace the continuity of decision-relevant knowledge across changes in representation, time, and context. On a broader scale, the "Meta-Sciences of Disciplinary Formation" outlines six revisable perspectives—including genesis, order, and legitimacy—to systematically study how research fields form and transform.
The pursuit of structural boundaries is equally prominent in computational history and dynamical systems, where researchers are using operator theory to map long-term cycles. The "Koopman-CMH" framework establishes the mathematical limits of detecting secular cycles, proving that shared ergodic-class membership implies a shared Koopman spectrum, though empirical historical panels often fall short of the statistical power required for detection. To resolve these limitations, in vitro simulation studies have benchmarked the power of pooled tests, revealing that empirical panels since 1789 sit at the absolute margin of detectability. To push past these predictive limits, researchers are looking "Beyond the Lyapunov Wall" by combining Koopman Operator theory, Dynamic Mode Decomposition, and Reservoir Computing to identify invariant cyclic structures in macrohistorical datasets. In physical dynamics, this mathematical refinement is mirrored in the development of Physics-Informed Kolmogorov-Arnold Networks (PIKANs) , which offer a noise-robust solution for reconstructing physical phenomena by filtering measurement noise while preserving high-frequency physical features that traditional solvers smooth away.
Finally, the demand for mathematical and procedural rigor is reshaping the economics of housing markets and property law. In real estate economics, framing sales comparison grids through the lens of "Conditional Expectation" clarifies the implicit assumptions of smoothness and unbiasedness required for accurate property valuation. Yet, the accuracy of these models is heavily threatened by "Omitted Variable Bias" , which distorts key coefficients while remaining hidden within standard R-squared statistics. To bridge the gap between theoretical econometrics and practical appraisal, researchers are advocating for "Valuation Engineering Protocols" (VEPs) to standardize statistical model design and out-of-sample validation. This drive toward precision is mirrored in legal doctrines surrounding compulsory land acquisition, where isolating compensable value requires a rigorous definition of "The Scheme" —the counterfactual world in which a public project does not exist—to successfully separate project-related value influences from fair market compensation. Collectively, these advancements across AI, history, physics, and economics signal a unified scientific trajectory toward highly structured, inspectable, and mathematically validated systems of knowledge.
This structural turn is highly visible in the governance of artificial intelligence, where researchers are moving away from simple technical optimization toward robust, human-centered control. To prevent failures in advanced autonomous systems, the "Human Control Model" establishes a multi-dimensional framework to evaluate agent safety, distinguishing genuine authorization from mere preference accuracy. This oversight is vital as large language models (LLMs) are increasingly deployed to reason about complex systemic issues like sustainability; evaluations of AI-generated explanations of the circular economy reveal that these systems often suffer from an operational indifference to truth and lack genuine moral agency. To manage the language we use to collaborate with these machines, the "Systematic Neology" framework provides a structured methodology to evaluate whether introducing new terminology in human-AI workflows adds genuine utility beyond existing linguistic options.
This effort to standardize and track knowledge creation extends into meta-science. As scientific fields evolve, maintaining cognitive continuity is essential. The "Cognitive Continuity Methodology" (CCM) provides a formal framework to test and trace the continuity of decision-relevant knowledge across changes in representation, time, and context. On a broader scale, the "Meta-Sciences of Disciplinary Formation" outlines six revisable perspectives—including genesis, order, and legitimacy—to systematically study how research fields form and transform.
The pursuit of structural boundaries is equally prominent in computational history and dynamical systems, where researchers are using operator theory to map long-term cycles. The "Koopman-CMH" framework establishes the mathematical limits of detecting secular cycles, proving that shared ergodic-class membership implies a shared Koopman spectrum, though empirical historical panels often fall short of the statistical power required for detection. To resolve these limitations, in vitro simulation studies have benchmarked the power of pooled tests, revealing that empirical panels since 1789 sit at the absolute margin of detectability. To push past these predictive limits, researchers are looking "Beyond the Lyapunov Wall" by combining Koopman Operator theory, Dynamic Mode Decomposition, and Reservoir Computing to identify invariant cyclic structures in macrohistorical datasets. In physical dynamics, this mathematical refinement is mirrored in the development of Physics-Informed Kolmogorov-Arnold Networks (PIKANs) , which offer a noise-robust solution for reconstructing physical phenomena by filtering measurement noise while preserving high-frequency physical features that traditional solvers smooth away.
Finally, the demand for mathematical and procedural rigor is reshaping the economics of housing markets and property law. In real estate economics, framing sales comparison grids through the lens of "Conditional Expectation" clarifies the implicit assumptions of smoothness and unbiasedness required for accurate property valuation. Yet, the accuracy of these models is heavily threatened by "Omitted Variable Bias" , which distorts key coefficients while remaining hidden within standard R-squared statistics. To bridge the gap between theoretical econometrics and practical appraisal, researchers are advocating for "Valuation Engineering Protocols" (VEPs) to standardize statistical model design and out-of-sample validation. This drive toward precision is mirrored in legal doctrines surrounding compulsory land acquisition, where isolating compensable value requires a rigorous definition of "The Scheme" —the counterfactual world in which a public project does not exist—to successfully separate project-related value influences from fair market compensation. Collectively, these advancements across AI, history, physics, and economics signal a unified scientific trajectory toward highly structured, inspectable, and mathematically validated systems of knowledge.
Latest Papers
[1]
Sapiens in Control: A Six-Dimension Framework for Evaluating Control Failures in Advanced AI Agents
Ethics and Social Impacts of AI
[2]
The circular economy, explained by AI
Ethics and Social Impacts of AI
[3]
Systematic Neology: Evaluating When New Terms Help in Human-AI Work
10 Citations·Topic Pending
[4]
[5]
Meta-Sciences of Disciplinary Formation: Six Perspectives for Studying Research Fields
8 Citations·Topic Pending
[6]
Koopman–CMH: A Formal Treatment of Spectral Structure Persistence in Historical Dynamics
4 Citations·Complex Systems and Time Series Analysis
[7]
Detecting Slow Cycles in Historical Panels: An In Vitro Study
4 Citations·Topic Pending
[8]
Beyond the Lyapunov Wall: Extending the Predictive Horizon of Computational Macrohistory. A Methodological Roadmap
2 Citations·Model Reduction and Neural Networks
[9]
Physics-informed Kolmogorov–Arnold networks (PIKANs) for noise-robust reconstruction of boiling heat flux from infrared measurements
Model Reduction and Neural Networks
[10]
Conditional Expectation
Housing Market and Economics
[11]
Omitted Variable Bias
Housing Market and Economics
[12]
Toward Valuation Engineering Protocols
Housing Market and Economics
[13]
The Scheme, and Effects on Value
Property Rights and Legal Doctrine