Digital hopes, real power: From connection to collective action
If the Arab Spring was defined by optimism about what the internet could do, the years since have been marked by a more sober understanding of what it takes to defend it
If the Arab Spring was defined by optimism about what the internet could do, the years since have been marked by a more sober understanding of what it takes to defend it
Analyzing CES and CPS survey data from 2012 to 2024, I evaluate whether the adoption of an opt-out AVR policy increases registration and turnout across the electorate, as well as among groups specifically targeted by these policies - movers and young people. Difference in difference estimates suggest that AVR increases registration by 1.8 percentage points among the full electorate and 3.5 percentage points among young voters but does not impact individual-level turnout…
Abstract This paper proposes a constellation pattern design optimization framework that considers satellite deployments. The proposed framework embeds the feasibility and practicality of multi-satellite deployment into the design space by including finite dwell time availability, J 2 -driven right ascension of the ascending node (RAAN) drift for orbital plane distribution, and the required completion time window. The resulting design problem is solved via a two-stage…
LLM agents retrieve untrusted content and act through tools, creating indirect prompt-injection risks that can cause unauthorized actions or persistent state changes. Existing automated red-teaming largely optimizes payloads for pre-specified scenarios, leaving latent vulnerabilities across the agent's behavior space unexplored. We present ASPIRE, an Agentic Safety & Prompt Injection Red-teaming Engine for open-ended, behavior-level vulnerability discovery. ASPIRE maintains…
AI agents increasingly rely on modular third-party "skills" that are dynamically selected by skill routers to execute complex tasks. While recent studies highlight the threat of prompt injections embedded in these skills, existing evaluations often assume settings where the malicious skill is already selected for execution. We show that this assumption can substantially overestimate attack success. In realistic multi-skill environments, injected skills must first compete for…
Large language model (LLM) agents are increasingly deployed in tool-augmented environments, but their reliance on external inputs makes them highly vulnerable to prompt injection attacks that can hijack task objectives. Existing safety alignment methods rely on static expert trajectories or preference optimization, limiting their ability to generalize to adaptive attack patterns. In this work, we propose Self-Reflection Fine-Tuning (SRFT), a training framework that enables…
This study develops an integrated LLM–fuzzy decision-support and digital twin framework for adaptive agricultural robotics, combining semantic reasoning, fuzzy inference, multi-objective optimization, and feedback-driven state updating. Numerical sensor data, operational information, and textual field knowledge are transformed into validated decision rules and model-ready parameters, enabling coordinated decisions across economic, operational, crop-stress, and…
Background: Augmented reality (AR) glasses with waveguide optics and facial recognition may enhance patient identification and clinical workflows. However, their usability among physicians has not yet been evaluated. Objective: The objective of this study was to assess the usability and acceptability of prototype AR glasses with facial recognition in a simulated outpatient setting. Methods: In May 2025, 14 urologists from a single institution tested prototype AR glasses…
Ahead of September 2026's OAS meeting of foreign ministers, this blog by Manuel Orozco examines the legal and diplomatic options available to hold Nicaragua accountable for its authoritarian rule and cross-border aggression, without violating the principle of nonintervention.
Driven by data-enabled technologies, precision agriculture is evolving into an innovative way to increase agricultural productivity and sustainability. Conventional agricultural practices face several challenges, such as poor soil quality, inappropriate crop selection, improper resource utilization, and unpredictable climate conditions. To address these limitations, this paper proposes a hybrid machine learning framework for sustainable crop optimization and predictive…
Large language model (LLM)-based agents increasingly rely on external tools and content, exposing them to indirect prompt injection (IPI). This threat has motivated a wide range of defenses, among which training-based defenses are often regarded as most reliable. However, existing training-based defenses are typically optimized on a static distribution of explicit injections. They learn surface-form cues rather than the boundary between serving the user and obeying an…
Introduction: Nursing home residents are at high risk for underrecognized pain and dehydration, particularly those with cognitive impairment who cannot reliably self-report symptoms. Consumer-grade smartwatches capable of continuous physiological monitoring may help detect these conditions, but their feasibility in frail older adults remains underexplored, and evaluation strategies based on random splits of repeated time windows can yield optimistically biased estimates. We…
We present a stylized model of a venture capital firm being active in a specific market, where all possible projects compete.These projects have a payoff structure such that only the best few alternatives generate a positive return.The payoff depends on the relative ranking of a project's quality in comparison with all other projects.This discontinuous payoff structure is motivated by the internet economy, where classical price competition is often not possible, and market…
Typed decision models (TDMs), such as Jev, answer a question about an input with a probability, an option, or an ordinal level. Web services execute these decisions automatically above a confidence threshold. An attacker who writes part of the input can change a decision that the service then executes. Current adversarial benchmarks cannot measure this risk, because they judge generated text. A TDM writes no text, and its wrong decisions have a valid format. We present…
Excessive nitrogen (N) fertilization is a major challenge to sustainable agriculture, adversely affecting crop productivity, soil quality, and environmental health, emits a potent greenhouse gas nitrous oxide (N2O), and nearly 80% of sectoral emissions linked to nutrient inputs applied during crop production. The escalating use of N fertilizers further amplifies N2O emissions, undermining soil sustainability and contributing to environmental degradation. So, the hypothesis…
This study takes the agricultural robot drive system as the research object and carries out optimization research from two dimensions: mechanical structure and intelligent control. At the mechanical level, redundant degrees of freedom are eliminated through topological analysis of mechanism freedom, a 'short-chain direct-drive' transmission scheme is introduced to reduce mechanical losses, and finite element topology optimization is combined to achieve lightweight design of…
Metaheuristic optimization increasingly requires search operators with functionally justified biological grounding rather than superficial metaphors. This paper introduces Sunn Pest Optimization (SPO), a novel algorithm derived from the bioecology of the cereal pest Eurygaster integriceps. SPO couples directional Lévy flight exploration with differential swarm exploitation while utilizing a tiered stagnation hierarchy and a spatial diversity-override mechanism to manage…
Activation steering modifies LLM behavior at inference time, but identifying where and how strongly to steer remains manual. We introduce Deep Noir, a framework that uses Logit Lens convergence and causal head-level attribution to autonomously discover optimal steering parameters. Across three scales (1B x 3, 2-3B x 2, and 7-9B x 4), our engine achieves 16.7 percentage-point improvement on spam at 1B (standard deviation 4.7; 39 runs), with gains increasing to 21 to 42…
ABSTRACT This study investigates volatility spillovers and minimum‐variance portfolio allocation among Islamic finance, sustainable, and technological innovation markets across four major global turbulences (COVID‐19 pandemic, Russia–Ukraine conflict, 2023 banking crisis, and 2025 US policy shifts) from July 07, 2021, to April 17, 2025. The study employs the Diebold‐Yilmaz and DCC‐GARCH models. Results show that volatility connections (conditional correlations) are generally…
Although smart home ecosystems have grown rapidly in scale and interaction complexity, bottlenecks in device pairing, scene automation, cross-device coordination, and feature discovery persist across major platforms. Existing usability research often relies on expert evaluation, questionnaires, or task-level summaries that lack the behavioral granularity needed to prioritize specific interaction nodes. Positioned as a Human–Computer Interaction and usability-engineering…
In response to the lack of high-quality annotated corpora in the current smart home field and the insufficient accuracy and interpretability of existing intent classification models in complex home scenarios, this study proposes an intent classification model system that integrates unsupervised topic discovery and deep learning classification. Firstly, a Chinese smart home corpus containing 120,000 real conversations with a dual-layer annotation structure is constructed, and…
In 2025, Jennifer Kavanaugh and Christopher McCallion wrote, “Armed Neutrality for Ukraine Is NATO’s Least Poor Option,” where they made the case for armed neutrality over full NATO membership as the wiser course for both Ukraine and NATO. A year later, we asked them to revisit their arguments. Image: MidjourneyIn your 2025 article, you argued that NATO membership and binding U.S. or European security guarantees for Ukraine were effectively unworkable, and that armed…
ABSTRACT Post‐aid scholarship suggests cooperation among non‐state actors in middle‐income regions has become less Northern‐anchored and more geographically dispersed. At the same time, optimistic readings of polycentric climate governance imply that decentralisation may loosen North–South hierarchies. We test this optimism using a partnership network of 195 climate‐focused non‐governmental organisations across 21 Latin American and Caribbean countries. We find that slightly…
Abstract Energy and climate assessments often contrast a few narrative scenarios, limiting insight into interacting uncertainties. We map a scenario space for a whole energy system model using more than 4,500 cost optimal runs, then analyse the ensemble with global sensitivity analysis, scatter plot diagnostics and scenario discovery. This workflow identifies influential drivers, reveals thresholds and regime switching, and distinguishes robust from contingent technology…
Vertical farming (VF) systems require massive amounts of electricity, primarily for artificial lighting, resulting in high carbon emissions and challenging profitability. Although integrating on-site renewable energy (RE) offers a promising pathway, its effectiveness depends on the joint design of cultivation intensity and energy infrastructure, which remains underexplored. This study develops a multi-objective VF-RE co-design framework that jointly optimises cultivation…
Against the global carbon neutrality agenda, renewable energy subsidies have become a core policy tool for China’s energy transition, yet related policies have faced repeated anti-subsidy challenges in international trade for over a decade. This paper examines the compliance dilemma of China’s renewable energy subsidy policy under the WTO Agreement on Subsidies and Countervailing Measures (SCM Agreement), drawing on three landmark trade disputes: two WTO cases (DS419…