Abstract
Imagine decision-makers uploading data and, within minutes, receiving clear, actionable insights delivered straight to their fingertips. That is the promise of the AI Data Scientist, an autonomous Agent powered by large language models (LLMs) that closes the gap between evidence and action. Rather than simply writing code or responding to prompts, it reasons through questions, tests ideas, and delivers end-to-end insights at a pace far beyond traditional workflows. Guided by the scientific tenet of the hypothesis, this Agent uncovers explanatory patterns in data, evaluates their statistical significance, and uses them to inform predictive modeling. It then translates these results into recommendations that are both rigorous and accessible. At the core of the AI Data Scientist is a team of specialized LLM Subagents, each responsible for a distinct task such as data cleaning, statistical testing, validation, and plain-language communication. These Subagents write their own code, reason about causality, and identify when additional data is needed to support sound conclusions. Together, they achieve in minutes what might otherwise take days or weeks, enabling a new kind of interaction that makes deep data science both accessible and actionable.

Abstract
This review examines how AI technologies are transforming democratic representation, focusing on citizen participation and algorithmic decision-making. The analysis reveals that AI technologies are reshaping democratic processes in fundamental ways: enabling mass-scale deliberation, changing how citizens access and engage with political information, and transforming how representatives make and implement decisions. While AI offers unprecedented opportunities for enhancing democratic participation and governance efficiency, it also presents significant challenges to democratic legitimacy and accountability. Social media platforms' AI-driven algorithms currently mediate much political discourse, creating concerns about information manipulation and privacy. Large Language Models introduce both epistemic challenges and potential tools for improving democratic dialogue. The emergence of Mass Online Deliberation platforms suggests possibilities for scaling up meaningful citizen participation, while Algorithmic Decision-Making systems promise more efficient policy implementation but face limitations in handling complex political trade-offs. As these systems become prevalent, representatives may assume the role of architects of automated decision frameworks, responsible for guiding the translation of politically contested concepts into technical parameters and metrics. Advanced deliberation platforms offering real-time insights into citizen preferences will challenge traditional representative independence and discretion to interpret public will. The institutional integration of these participation mechanisms requires frameworks that balance the benefits with democratic stability through hybrid systems weighting different forms of democratic expression.

Abstract
Reinforcement learning (RL) marks a fundamental shift in how artificial intelligence is applied in healthcare. Instead of merely predicting outcomes, RL actively decides interventions with long term goals. Unlike traditional models that operate on fixed associations, RL systems learn through trial, feedback, and long-term reward optimization, introducing transformative possibilities and new risks. From an information fusion lens, healthcare RL typically integrates multi-source signals such as vitals, labs clinical notes, imaging and device telemetry using temporal and decision-level mechanisms. These systems can operate within centralized, federated, or edge architectures to meet real-time clinical constraints, and naturally span data, features and decision fusion levels. This survey explore RL's rise in healthcare as more than a set of tools, rather a shift toward agentive intelligence in clinical environments. We first structure the landscape of RL techniques including model-based and model-free methods, offline and batch-constrained approaches, and emerging strategies for reward specification and uncertainty calibration through the lens of healthcare constraints. We then comprehensively analyze RL applications spanning critical care, chronic disease, mental health, diagnostics, and robotic assistance, identifying their trends, gaps, and translational bottlenecks. In contrast to prior reviews, we critically analyze RL's ethical, deployment, and reward design challenges, and synthesize lessons for safe, human-aligned policy learning. This paper serves as both a a technical roadmap and a critical reflection of RL's emerging transformative role in healthcare AI not as prediction machinery, but as agentive clinical intelligence.

Abstract
As deep learning (DL) technologies advance, their application in automated visual inspection for Class III medical devices offers significant potential to enhance quality assurance and reduce human error. However, the adoption of such AI-based systems introduces new regulatory complexities--particularly under the EU Artificial Intelligence (AI) Act, which imposes high-risk system obligations that differ in scope and depth from established regulatory frameworks such as the Medical Device Regulation (MDR) and the U.S. FDA Quality System Regulation (QSR). This paper presents a high-level technical assessment of the foresee-able challenges that manufacturers are likely to encounter when qualifying DL-based automated inspections within the existing medical device compliance landscape. It examines divergences in risk management principles, dataset governance, model validation, explainability requirements, and post-deployment monitoring obligations. The discussion also explores potential implementation strategies and highlights areas of uncertainty, including data retention burdens, global compliance implications, and the practical difficulties of achieving statistical significance in validation with limited defect data. Disclaimer: This publication is in-tended solely as an academic and technical evaluation. It is not a substitute for le-gal advice or official regulatory interpretation. The information presented here should not be relied upon to demonstrate compliance with the EU AI Act or any other statutory obligation. Manufacturers are encouraged to consult appropriate regulatory authorities and legal experts to determine specific compliance pathways.

Abstract
Auditability is defined as the capacity of AI systems to be independently assessed for compliance with ethical, legal, and technical standards throughout their lifecycle. The chapter explores how auditability is being formalized through emerging regulatory frameworks, such as the EU AI Act, which mandate documentation, risk assessments, and governance structures. It analyzes the diverse challenges facing AI auditability, including technical opacity, inconsistent documentation practices, lack of standardized audit tools and metrics, and conflicting principles within existing responsible AI frameworks. The discussion highlights the need for clear guidelines, harmonized international regulations, and robust socio-technical methodologies to operationalize auditability at scale. The chapter concludes by emphasizing the importance of multi-stakeholder collaboration and auditor empowerment in building an effective AI audit ecosystem. It argues that auditability must be embedded in AI development practices and governance infrastructures to ensure that AI systems are not only functional but also ethically and legally aligned.

Abstract
As generative artificial intelligence (AI) tools become widely adopted, large language models (LLMs) are increasingly involved on both sides of decision-making processes, ranging from hiring to content moderation. This dual adoption raises a critical question: do LLMs systematically favor content that resembles their own outputs? Prior research in computer science has identified self-preference bias -- the tendency of LLMs to favor their own generated content -- but its real-world implications have not been empirically evaluated. We focus on the hiring context, where job applicants often rely on LLMs to refine resumes, while employers deploy them to screen those same resumes. Using a large-scale controlled resume correspondence experiment, we find that LLMs consistently prefer resumes generated by themselves over those written by humans or produced by alternative models, even when content quality is controlled. The bias against human-written resumes is particularly substantial, with self-preference bias ranging from 68% to 88% across major commercial and open-source models. To assess labor market impact, we simulate realistic hiring pipelines across 24 occupations. These simulations show that candidates using the same LLM as the evaluator are 23% to 60% more likely to be shortlisted than equally qualified applicants submitting human-written resumes, with the largest disadvantages observed in business-related fields such as sales and accounting. We further demonstrate that this bias can be reduced by more than 50% through simple interventions targeting LLMs' self-recognition capabilities. These findings highlight an emerging but previously overlooked risk in AI-assisted decision making and call for expanded frameworks of AI fairness that address not only demographic-based disparities, but also biases in AI-AI interactions.

Abstract
Artificial intelligence simultaneously transforms human capital production in schools and its demand in labor markets. Analyzing these effects in isolation can lead to a significant misallocation of educational resources. We model an educational planner whose decision to adopt AI is driven by its teaching productivity, failing to internalize AI's future wage-suppressing effect on those same skills. Our core assumption, motivated by a pilot survey, is that there is a positive correlation between these two effects. This drives our central proposition: this information failure creates a skill mismatch that monotonically increases with AI prevalence. Extensions show the mismatch is exacerbated by the neglect of unpriced non-cognitive skills and by a school's endogenous over-investment in AI. Our findings caution that policies promoting AI in education, if not paired with forward-looking labor market signals, may paradoxically undermine students' long-term human capital, especially if reliance on AI crowds out the development of unpriced non-cognitive skills, such as persistence, that are forged through intellectual struggle.

Abstract
One of the enduring challenges in education is how to empower students to take ownership of their learning by setting meaningful goals, tracking their progress, and adapting their strategies when faced with setbacks. Research has shown that this form of leaner-centered learning is best cultivated through structured, supportive environments that promote guided practice, scaffolded inquiry, and collaborative dialogue. In response, educational efforts have increasingly embraced artificial-intelligence (AI)-powered digital learning environments, ranging from educational apps and virtual labs to serious games. Recent advances in large language models (LLMs) and neuro-symbolic systems, meanwhile, offer a transformative opportunity to reimagine how support is delivered in digital learning environments. LLMs are enabling socially interactive learning experiences and scalable, cross-domain learning support that can adapt instructional strategies across varied subjects and contexts. In parallel, neuro-symbolic AI provides new avenues for designing these agents that are not only adaptive but also scalable across domains. Based on these remarks, this paper presents a multi-agent, neuro-symbolic framework designed to resolve the aforementioned challenges. The framework assigns distinct pedagogical roles to specialized agents: an RL-based 'tutor' agent provides authoritative, non-verbal scaffolding, while a proactive, LLM-powered 'peer' agent facilitates the social dimensions of learning. While prior work has explored such agents in isolation, our framework's novelty lies in unifying them through a central educational ontology. Through case studies in both college-level and middle school settings, we demonstrate the framework's adaptability across domains. We conclude by outlining key insights and future directions for advancing AI-driven learning environments.

Abstract
A fundamental question in cognitive science concerns how social norms are acquired and represented. While humans typically learn norms through embodied social experience, we investigated whether large language models can achieve sophisticated norm understanding through statistical learning alone. Across two studies, we systematically evaluated multiple AI systems' ability to predict human social appropriateness judgments for 555 everyday scenarios by examining how closely they predicted the average judgment compared to each human participant. In Study 1, GPT-4.5's accuracy in predicting the collective judgment on a continuous scale exceeded that of every human participant (100th percentile). Study 2 replicated this, with Gemini 2.5 Pro outperforming 98.7% of humans, GPT-5 97.8%, and Claude Sonnet 4 96.0%. Despite this predictive power, all models showed systematic, correlated errors. These findings demonstrate that sophisticated models of social cognition can emerge from statistical learning over linguistic data alone, challenging strong versions of theories emphasizing the exclusive necessity of embodied experience for cultural competence. The systematic nature of AI limitations across different architectures indicates potential boundaries of pattern-based social understanding, while the models' ability to outperform nearly all individual humans in this predictive task suggests that language serves as a remarkably rich repository for cultural knowledge transmission.

Abstract
This study presents a lightweight, domain-informed AI model for predicting indoor temperatures in naturally ventilated schools and homes in Sub-Saharan Africa. The model extends the Temp-AI-Estimator framework, trained on Tanzanian school data, and evaluated on Nigerian schools and Gambian homes. It achieves robust cross-country performance using only minimal accessible inputs, with mean absolute errors of 1.45{\deg}C for Nigerian schools and 0.65{\deg}C for Gambian homes. These findings highlight AI's potential for thermal comfort management in resource-constrained environments.

Abstract
As artificial intelligence (AI) becomes foundational to enterprise infrastructure, organizations face growing challenges in accurately assessing the full economic implications of AI deployment. Existing metrics such as API token costs, GPU-hour billing, or Total Cost of Ownership (TCO) fail to capture the complete lifecycle costs of AI systems and provide limited comparability across deployment models. This paper introduces the Levelized Cost of Artificial Intelligence (LCOAI), a standardized economic metric designed to quantify the total capital (CAPEX) and operational (OPEX) expenditures per unit of productive AI output, normalized by valid inference volume. Analogous to established metrics like LCOE (levelized cost of electricity) and LCOH (levelized cost of hydrogen) in the energy sector, LCOAI offers a rigorous, transparent framework to evaluate and compare the cost-efficiency of vendor API deployments versus self-hosted, fine-tuned models. We define the LCOAI methodology in detail and apply it to three representative scenarios, OpenAI GPT-4.1 API, Anthropic Claude Haiku API, and a self-hosted LLaMA-2-13B deployment demonstrating how LCOAI captures critical trade-offs in scalability, investment planning, and cost optimization. Extensive sensitivity analyses further explore the impact of inference volume, CAPEX, and OPEX variability on lifecycle economics. The results illustrate the practical utility of LCOAI in procurement, infrastructure planning, and automation strategy, and establish it as a foundational benchmark for AI economic analysis. Policy implications and areas for future refinement, including environmental and performance-adjusted cost metrics, are also discussed.

Abstract
This paper introduces and overviews a multidisciplinary project aimed at developing responsible and adaptive multi-human multi-robot (MHMR) systems for complex, dynamic settings. The project integrates co-design, ethical frameworks, and multimodal sensing to create AI-driven robots that are emotionally responsive, context-aware, and aligned with the needs of diverse users. We outline the project's vision, methodology, and early outcomes, demonstrating how embodied AI can support sustainable, ethical, and human-centred futures.

Abstract
Communicating the risks and benefits of AI is important for regulation and public understanding. Yet current methods such as technical reports often exclude people without technical expertise. Drawing on HCI research, we developed an Impact Assessment Card to present this information more clearly. We held three focus groups with a total of 12 participants who helped identify design requirements and create early versions of the card. We then tested a refined version in an online study with 235 participants, including AI developers, compliance experts, and members of the public selected to reflect the U.S. population by age, sex, and race. Participants used either the card or a full impact assessment report to write an email supporting or opposing a proposed AI system. The card led to faster task completion and higher-quality emails across all groups. We discuss how design choices can improve accessibility and support AI governance. Examples of cards are available at: https://social-dynamics.net/ai-risks/impact-card/.

Abstract
As artificial intelligence rapidly transforms society, developers and policymakers struggle to anticipate which applications will face public moral resistance. We propose that these judgments are not idiosyncratic but systematic and predictable. In a large, preregistered study (N = 587, U.S. representative sample), we used a comprehensive taxonomy of 100 AI applications spanning personal and organizational contexts-including both functional uses and the moral treatment of AI itself. In participants' collective judgment, applications ranged from highly unacceptable to fully acceptable. We found this variation was strongly predictable: five core moral qualities-perceived risk, benefit, dishonesty, unnaturalness, and reduced accountability-collectively explained over 90% of the variance in acceptability ratings. The framework demonstrated strong predictive power across all domains and successfully predicted individual-level judgments for held-out applications. These findings reveal that a structured moral psychology underlies public evaluation of new technologies, offering a powerful tool for anticipating public resistance and guiding responsible innovation in AI.

Abstract
Fairness in recommender systems (RSs) is commonly categorised into group fairness and individual fairness. However, there is no established scientific understanding of the relationship between the two fairness types, as prior work on both types has used different evaluation measures or evaluation objectives for each fairness type, thereby not allowing for a proper comparison of the two. As a result, it is currently not known how increasing one type of fairness may affect the other. To fill this gap, we study the relationship of group and individual fairness through a comprehensive comparison of evaluation measures that can be used for both fairness types. Our experiments with 8 runs across 3 datasets show that recommendations that are highly fair for groups can be very unfair for individuals. Our finding is novel and useful for RS practitioners aiming to improve the fairness of their systems. Our code is available at: https://github.com/theresiavr/stairway-to-fairness.

Abstract
Machine learning models must balance accuracy and fairness, but these goals often conflict, particularly when data come from multiple demographic groups. A useful tool for understanding this trade-off is the fairness-accuracy (FA) frontier, which characterizes the set of models that cannot be simultaneously improved in both fairness and accuracy. Prior analyses of the FA frontier provide a full characterization under the assumption of complete knowledge of population distributions -- an unrealistic ideal. We study the FA frontier in the finite-sample regime, showing how it deviates from its population counterpart and quantifying the worst-case gap between them. In particular, we derive minimax-optimal estimators that depend on the designer's knowledge of the covariate distribution. For each estimator, we characterize how finite-sample effects asymmetrically impact each group's risk, and identify optimal sample allocation strategies. Our results transform the FA frontier from a theoretical construct into a practical tool for policymakers and practitioners who must often design algorithms with limited data.

Abstract
The growing complexity of urban mobility and the demand for efficient, sustainable, and adaptive solutions have positioned Intelligent Transportation Systems (ITS) at the forefront of modern infrastructure innovation. At the core of ITS lies the challenge of autonomous decision-making across dynamic, large scale, and uncertain environments where multiple agents traffic signals, autonomous vehicles, or fleet units must coordinate effectively. Multi Agent Reinforcement Learning (MARL) offers a promising paradigm for addressing these challenges by enabling distributed agents to jointly learn optimal strategies that balance individual objectives with system wide efficiency. This paper presents a comprehensive survey of MARL applications in ITS. We introduce a structured taxonomy that categorizes MARL approaches according to coordination models and learning algorithms, spanning value based, policy based, actor critic, and communication enhanced frameworks. Applications are reviewed across key ITS domains, including traffic signal control, connected and autonomous vehicle coordination, logistics optimization, and mobility on demand systems. Furthermore, we highlight widely used simulation platforms such as SUMO, CARLA, and CityFlow that support MARL experimentation, along with emerging benchmarks. The survey also identifies core challenges, including scalability, non stationarity, credit assignment, communication constraints, and the sim to real transfer gap, which continue to hinder real world deployment.

Abstract
Bus bunching remains a challenge for urban transit due to stochastic traffic and passenger demand. Traditional solutions rely on multi-agent reinforcement learning (MARL) in loop-line settings, which overlook realistic operations characterized by heterogeneous routes, timetables, fluctuating demand, and varying fleet sizes. We propose a novel single-agent reinforcement learning (RL) framework for bus holding control that avoids the data imbalance and convergence issues of MARL under near-realistic simulation. A bidirectional timetabled network with dynamic passenger demand is constructed. The key innovation is reformulating the multi-agent problem into a single-agent one by augmenting the state space with categorical identifiers (vehicle ID, station ID, time period) in addition to numerical features (headway, occupancy, velocity). This high-dimensional encoding enables single-agent policies to capture inter-agent dependencies, analogous to projecting non-separable inputs into a higher-dimensional space. We further design a structured reward function aligned with operational goals: instead of exponential penalties on headway deviations, a ridge-shaped reward balances uniform headways and schedule adherence. Experiments show that our modified soft actor-critic (SAC) achieves more stable and superior performance than benchmarks, including MADDPG (e.g., -430k vs. -530k under stochastic conditions). These results demonstrate that single-agent deep RL, when enhanced with categorical structuring and schedule-aware rewards, can effectively manage bus holding in non-loop, real-world contexts. This paradigm offers a robust, scalable alternative to MARL frameworks, particularly where agent-specific experiences are imbalanced.

Abstract
The learning from practice paradigm is crucial for developing capable Agentic AI systems, yet it is severely hampered by inefficient experience generation, a bottleneck especially pronounced in complex benchmarks like GAIA. To address this, we introduce AWorld, an open-source system engineered for large-scale agent-environment interaction. By distributing tasks across a cluster, AWorld accelerates experience collection by 14.6x compared to standard single-node, sequential execution. This critical speedup makes extensive reinforcement learning practical and scalable. Leveraging this capability, we trained a Qwen3-32B-based agent that significantly outperforms its base model, increasing its overall GAIA accuracy from 21.59% to 32.23%. On the benchmark's most challenging levels, our agent achieves a score of 16.33%, surpassing the performance of leading proprietary models. Our open-source system and resulting agent provide a practical blueprint for a complete agentic AI training pipeline, from efficient interaction to demonstrable model improvement.

Abstract
Reasoning models are the new generation of Large Language Models (LLMs) capable of complex problem solving. Their reliability in solving introductory physics problems was tested by evaluating a sample of n = 5 solutions generated by one such model -- OpenAI's o3-mini -- per each problem from 20 chapters of a standard undergraduate textbook. In total, N = 408 problems were given to the model and N x n = 2,040 generated solutions examined. The model successfully solved 94% of the problems posed, excelling at the beginning topics in mechanics but struggling with the later ones such as waves and thermodynamics.

Abstract
Physics-Informed Neural Networks (PINNs) present a transformative approach for smart grid modeling by integrating physical laws directly into learning frameworks, addressing critical challenges of data scarcity and physical consistency in conventional data-driven methods. This paper evaluates PINNs' capabilities as surrogate models for smart grid dynamics, comparing their performance against XGBoost, Random Forest, and Linear Regression across three key experiments: interpolation, cross-validation, and episodic trajectory prediction. By training PINNs exclusively through physics-based loss functions (enforcing power balance, operational constraints, and grid stability) we demonstrate their superior generalization, outperforming data-driven models in error reduction. Notably, PINNs maintain comparatively lower MAE in dynamic grid operations, reliably capturing state transitions in both random and expert-driven control scenarios, while traditional models exhibit erratic performance. Despite slight degradation in extreme operational regimes, PINNs consistently enforce physical feasibility, proving vital for safety-critical applications. Our results contribute to establishing PINNs as a paradigm-shifting tool for smart grid surrogation, bridging data-driven flexibility with first-principles rigor. This work advances real-time grid control and scalable digital twins, emphasizing the necessity of physics-aware architectures in mission-critical energy systems.

Abstract
The scalable computing revolution of the late '80s through mid- '00s forged a new technical and economic model for computing that delivered massive societal impact, but its economic benefit has driven scalability to sizes that are now exhausting the energy grid's capacity. Our time demands a new revolution in scalable energy, mirroring in key ways the scalable computing revolution; e.g., compelling economic forces, use of mass-market components, overcoming foibles of those components, judicious use of physical locality, and the the difficult integration into an effective system. The offgrid AI approach closely fits this mold, combining local mostly renewable generation and storage to power an AI data center, starting offgrid. Obstacles to delivering this approach are social, technical, and project, but the potential is massive. I argue that the offgrid-AI approach needs pioneers among both system developers and AI-data-center operators to move it quickly from concept to large-scale deployment.

Abstract
Research and usage of artificial intelligence, particularly generative and large language models, have rapidly progressed over the last years. This has, however, given rise to issues due to high energy consumption. While quantum computing is not (yet) mainstream, its intersection with machine learning is especially promising, and the technology could alleviate some of these energy challenges. In this perspective article, we break down the lifecycle stages of large language models and discuss relevant enhancements based on quantum algorithms that may aid energy efficiency and sustainability, including industry application examples and open research problems.

Abstract
The stability of ice cream during melting is a critical factor for consumer's acceptance and product quality. With the commonly added stabilizer to improve texture, structure and slower melting as the factors to analyze. This report explores the effects of locust bean gum, guar gum, maltodextrin, and carrageenan on the melting behavior of homemade ice cream. The main objective was to assess how these additives influence melting resistance and to identify a more cost-effective recipe formulation. Ice cream samples incorporating each additive were prepared and subjected to melting tests under controlled conditions. Timelapse recordings were used to capture and analyze the progression of melting over time. Python and OpenCV is used for process and analysis. Observations revealed that all samples retained a foam-like structure even after melting, suggesting the stabilizers contributed to the formation of a stable air-cell matrix. Furthermore, when the melted samples were re-frozen and subsequently melted again, they displayed increased sturdiness, indicating improved resilience of the ice cream structure. Comparative analysis of the different stabilizers highlighted variations in their effectiveness, with some offering stronger melting resistance and structural support than others. Overall, the findings provide insights into the functional roles of commonly used food additives in ice cream formulation. By evaluating both performance and cost, this study demonstrates the potential for developing recipes that balance durability with economic efficiency, contributing to practical applications in both small-scale and commercial ice cream production.