Oct 2026· Social Science Humanities and Sustainability Research
Ethics and Social Impacts of AI
Abstract
Artificial intelligence requires systematic reflection on the relationship between technological progress, collective responsibility and the dignity of the person, making a new social pact necessary. The article links dependence on fossil fuels to geopolitical imbalances, inflationary pressures and the ability of authoritarian regimes to support power policies, proposing international coordination, including at the UN, that reconciles energy security, climate protection and the transition to renewable sources and, in perspective, nuclear fusion. This analysis is accompanied by the criticism of an economy based exclusively on solvency, which excludes needs without adequate purchasing power from the production and distribution of essential goods. To combat poverty, inequality and vulnerability to war, an international agency capable of promoting socially useful work and coordinating the production and distribution of fundamental resources is envisaged. In this framework, artificial intelligence and robotics could contribute to the analysis of needs, the organization of work, agriculture, assistance and scientific research, as long as they remain subordinate to democratic decision-making and criteria of transparency, human control and non-discrimination. Finally, the technological comparison between the United States, China and Europe shows the need, especially for Europe, to strengthen investment, research and regulatory capacity. Energy, peace, social justice and innovation are thus brought back to the different dimensions of a single pact based on collective responsibility and human dignity.
GAOKAO-Bench is introduced, an intuitive benchmark that employs questions from the Chinese GAOKAO examination as test samples, including both subjective and objective questions that contribute a robust evaluation benchmark for future large language models and offers valuable insights into the advantages and limitations...
Xiaotian Zhang, Chun-yan Li, Yi Zong et al.· arXiv.org· 216 citations· ⚡17
An algorithm is proposed which leverages disentangled semantic factors to generate adversarial perturbation by altering controlled semantic attributes to fool the learner towards various "adversarial" targets.
Haonan Qiu, Chaowei Xiao, Lei Yang et al.· arXiv.org· 208 citations· ⚡17
This work proposes a HALO that directly maximizes the utility of generations instead of maximizing the log-likelihood of preferences, and suggests that the best loss depends on the inductive biases most appropriate for a given setting, an oft-overlooked consideration.
This work investigates the possibilities of using LLMs in a resume screening setting via a document retrieval framework that simulates job candidate selection and finds that the MTEs are biased, significantly favoring White-associated names in 85% of cases and female-associated names in only 11.1% of cases.
This work shows that orders of magnitude enhancement in performance could be obtained by a combination of hardware improvements and tight quantum-HPC integration and introduces high-performance architectures for quantum-probabilistic computing with custom-designed accelerators to tackle today's industry-scale classical...
Masoud Mohseni, Artur Scherer, K. Johnson et al.· arXiv.org· 121 citations· ⚡9
This paper presents a comprehensive overview of the Ultralytics YOLO family, emphasizing architectural evolution, benchmarking, deployment, and emerging directions from YOLOv5 through YOLO27, and examines detection, segmentation, depth, classification, pose, oriented detection, tracking, export, quantization, and deplo...