1^st June 2026

New Geometric Framework Aims to Replace Dark Matter

This week's research shifts toward precise mathematical solutions. In physics, a new geometric theory proposes replacing dark matter entirely, explaining galaxy rotations without invisible particles , while a related model unifies dark matter and energy . In technology, engineers developed a mathematical control loop to prevent AI systems from generating false information . The societal impacts of AI are also under scrutiny: generative AI threatens democratic information sharing , and students relying on AI to manage stress are experiencing higher burnout . Finally, materials science saw advances in stable, highly efficient solar panels , and safer batteries that resist internal damage , , alongside new AI tools predicting dementia risk after strokes .

Top 10 topics by publication and citation volume

Ethics and Social Impacts of AI15

Perovskite Materials and Applications12

Cosmology and Gravitation Theories12

AI in Service Interactions10

Artificial Intelligence in Healthcare and Education10

Advanced Battery Materials and Technologies10

Management and Organizational Studies9

Dementia and Cognitive Impairment Research9

Education, Innovation and Language Studies9

Embodied and Extended Cognition9

Extended Breakdown↓

Building on last week's focus on clinical AI workflows and nature-inspired materials, the latest scientific literature demonstrates a significant shift toward highly formalized, geometric, and molecular-level solutions. Across disciplines—ranging from AI ethics and cosmology to battery chemistry and cognitive science—researchers are moving past general conceptual frameworks to implement rigorous, mathematically grounded, and structurally precise interventions.

Ethics, Democracy, and Governance in the Agentic AI Era

The rapid integration of artificial intelligence into critical societal infrastructures has prompted a profound re-evaluation of ethical, legal, and regulatory frameworks. In military operations, the transition to the "agentic era" is marked by the deployment of lethal autonomous weapons systems (LAWS), drone swarms, and AI-assisted cyber warfare [P10]. Researchers argue that this rapid pace of development far outstrips international regulatory capacities, placing a unique historical responsibility on regions with strong traditions of peace, such as Latin America, to advocate for meaningful human control [P10].

On the domestic front, generative AI poses a structural threat to democratic governance by shifting the paradigm from the "rule of law" to the "rule of algorithm" . By constructing and manipulating the informational environment rather than merely classifying it, generative AI dissolves public answerability and fragments the shared factual ground necessary for democratic deliberation . To mitigate these systemic risks, researchers have proposed risk-tiered governance checklists designed to scale compliance measures directly to the actual risks posed by AI systems, thereby balancing safety with the pace of innovation [P67]. Furthermore, when AI is integrated into intimate human domains like healthcare, care ethics frameworks reveal that dominant deontology- or efficiency-based models fail to capture the relational and evaluative aspects of caregiving [P19]. AI and robotic systems reorganize care by shifting human attentiveness to sensing and responsibility to oversight, potentially reducing the conditions necessary for authentic, empathetic care [P19].

Advancements in Perovskite Solar Cells and Optoelectronics

In materials science, perovskite solar cells (PSCs) continue to see rapid efficiency and stability breakthroughs. Addressing the poor interfacial compatibility and intrinsic defects of all-inorganic CsPbBr3 semi-transparent cells, researchers successfully utilized atomic layer deposition (ALD) with ultrapure water as a co-reactant to grow uniform nickel oxide (NiOx) hole transport layers [P3]. When combined with ultrasonic spray-coated PEABr surface passivation, these cells achieved a power conversion efficiency (PCE) of 7.00% while maintaining high visible transmittance [P3].

To push efficiency boundaries further, the molecular engineering of covalent organic frameworks (COFs) has emerged as a key strategy. Introducing N-cationic radicals into electron-rich COFs has enabled a certified PCE of 25.98% and outstanding thermal durability, retaining 88% of initial performance after 1000 hours of aging at 85°C . Interfacial adhesion has also been dramatically improved through the design of novel benzofuran-substituted self-assembled molecules (SAMs), which enhance mechanical strength at the hole-transport-layer/perovskite interface, yielding highly stable modules under rigorous thermal cycling . Concurrently, the strategic development of lead-free alternatives, particularly tin-based PSCs, is accelerating to address toxicity concerns while striving to match lead-based performance [P75].

Alternative Cosmological Frameworks and Geometric Gravity

Cosmology is undergoing a theoretical renaissance as researchers propose novel geometric replacements for dark matter and dark energy to resolve persistent cosmic tensions. Hu's Cosmic Spacetime Origin Theory (HCSOT) offers a minimal modified gravity framework based on a single scalar-tensor field that unifies dark matter and dark energy through density-dependent gravitational coupling . This theory has pre-registered precise weak-lensing predictions for the upcoming Euclid DR1 data release, offering a clear path to empirical validation . In a parallel breakthrough, Information Tension Theory proposes a purely geometric replacement for dark matter . By deriving the Information Tension tensor as a geometric stress on a 240-dimensional Stiefel manifold, this framework successfully reproduces flat galaxy rotation curves and Bullet Cluster lensing without invoking dark matter halos, a theory supported by statistical signatures in the Trinity 2.0 dataset .

AI Service Interactions and the "Cognitive Footprint"

The deployment of AI in service sectors is shifting from passive assistance to active behavioral modification. Field experiments in the hospitality industry demonstrate that robot concierges utilizing guilt-framed messages can significantly increase sustainable guest behaviors, such as towel and linen reuse, by leveraging anticipated guilt and perceived social presence [P17]. However, as humans interact more deeply with conversational agents, they leave behind a "Cognitive Footprint"—a unique trace of questions, decisions, and corrections that serves as a highly valuable, yet sensitive, 21st-century identity signature [P21].

AI Integration in Healthcare, Education, and the Risk of Dependence

In healthcare, the GIN-McMaster Guideline Development Checklist has been extended to provide standards for integrating AI into clinical guideline development, ensuring transparency and methodological rigor [P38]. However, the educational sector reveals a darker side of AI integration. A large-scale study of college students found that academic stress frequently drives students toward maladaptive AI dependence . This cognitive offloading undermines self-efficacy, reduces mastery experiences, and ultimately mediates a pathway toward heightened academic burnout and anxiety, highlighting the urgent need for autonomy-supportive educational scaffolding .

Next-Generation Solid-State Battery Technologies

Solid-state lithium metal batteries are closer to practical realization thanks to innovative electrolyte designs. To prevent interfacial degradation in high-voltage, nickel-rich lithium batteries, researchers developed a deep-eutectic polymer electrolyte (p-DEPE) that establishes a competitive hydrogen-bonding coordination environment, successfully inhibiting cathode decomposition and oxygen release . Additionally, the suppression of lithium dendrites has been achieved through the molecular engineering of dual-ion regulated COFs . By integrating lithiophilic methoxy groups and anionophilic imidazolium species into a single framework, this design immobilizes anions to prevent space-charge polarization, achieving a high Li+ transference number of ~0.72 and enabling stable cycling with high-voltage cathodes .

Early Stratification and Biomarkers in Dementia Research

Dementia research is benefiting from both biological and computational breakthroughs. The clinical introduction of blood-based biomarkers (BBMs) offers a scalable, low-cost diagnostic tool for Alzheimer's disease, though it requires careful shared decision-making frameworks to prevent misdiagnosis and address health disparities [P15]. Complementing these biomarkers, multimodal machine learning models integrating clinical, demographic, and neuroimaging data have demonstrated remarkable accuracy (up to 98.13% internally) in the early risk stratification of post-stroke cognitive impairment (PSCI), paving the way for personalized preventative interventions .

Geometric Foundations and Hybrid Paradigms in Extended Cognition

Finally, the boundary between human and machine cognition is blurring, necessitating new theoretical foundations. The concept of "hybrid thinking" describes a joint cognitive process where humans and AI agents simultaneously elaborate the same chain of reasoning, introducing benefits like accelerated discovery alongside severe risks of cognitive atrophy and silent transfer bias [P83]. To address the catastrophic failure mode of epistemic drift (or hallucination) in these systems, researchers have introduced the Anti-Drift Cognitive Control Loop (ADCCL) . This non-stochastic governance layer mathematically bounds AI reasoning trajectories within a stable topological manifold, providing a structural guarantee of hallucination-free, high-integrity sovereign intelligence .

Selected Papers
* P10
* P19
* 4
* 2
* 3
* 5
* 6
* 10
* 1
* 7
* 8
* 9