Why we think this paper is great for you:
This paper offers a practical, measurable approach to addressing poverty using advanced machine learning techniques. It demonstrates how technology can provide crucial insights for ending poverty in underserved regions.

Abstract
Accurate, fine-grained poverty maps remain scarce across much of the Global South. While Demographic and Health Surveys (DHS) provide high-quality socioeconomic data, their spatial coverage is limited and reported coordinates are randomly displaced for privacy, further reducing their quality. We propose a graph-based approach leveraging low-dimensional AlphaEarth satellite embeddings to predict cluster-level wealth indices across Sub-Saharan Africa. By modeling spatial relations between surveyed and unlabeled locations, and by introducing a probabilistic "fuzzy label" loss to account for coordinate displacement, we improve the generalization of wealth predictions beyond existing surveys. Our experiments on 37 DHS datasets (2017-2023) show that incorporating graph structure slightly improves accuracy compared to "image-only" baselines, demonstrating the potential of compact EO embeddings for large-scale socioeconomic mapping.

AI Summary

• Graph-based models, particularly the 'Ego Graphs' approach, modestly improve poverty prediction accuracy (R2 0.569) over image-only baselines by leveraging local spatial context within plausible DHS displacement radii. [2]
• Compact AlphaEarth satellite embeddings enable large-scale graph neural network training for continental-scale socioeconomic mapping with significantly reduced computational cost (45.6 MB for AlphaEarth vs. [2]
• 51.5 GB for Sentinel-2). [2]
• The proposed 'fuzzy label' loss function, designed to account for DHS coordinate displacement, unexpectedly degrades model performance, indicating challenges in robustly learning under spatial uncertainty with current candidate location datasets. [2]
• Incorporating auxiliary unlabeled settlement locations from the GeoNames database into graph structures provides complementary information, enhancing model performance beyond predictions based solely on single-site image embeddings. [2]
• DHS coordinate displacement (up to 10 km in rural areas) introduces substantial spatial uncertainty, which current methods struggle to robustly address, highlighting a critical area for future methodological refinement. [2]
• Future refinements for fuzzy-labeling should integrate population density priors and actual DHS sampling frames to more accurately define candidate locations and their likelihoods, mitigating biases from incomplete auxiliary datasets like GeoNames. [2]
• Refining graph edge definitions beyond simple Euclidean distance, such as incorporating estimated travel time derived from road networks or terrain models, could more accurately capture real-world accessibility and socioeconomic interactions. [2]
• AlphaEarth embeddings: Compact, 64-dimensional representations of local land use and environmental context derived from satellite imagery, serving as the primary input for predictive models. [2]
• International Wealth Index (IWI): A standardized, survey-based measure of material wealth, ranging from 0 to 100, used as the target variable for socioeconomic prediction. [2]
• Fuzzy label loss function: A probabilistic loss function formulated to explicitly account for spatial uncertainty due to DHS coordinate displacement by weighting candidate settlement locations according to their displacement likelihood. [2]

Why we think this paper is great for you:
You will find this paper highly relevant as it introduces a comprehensive index for quantifying societal well-being beyond traditional economic metrics. It provides a valuable tool for understanding and promoting a healthy society at a granular level.

Abstract
Quantifying human flourishing, a multidimensional construct including happiness, health, purpose, virtue, relationships, and financial stability, is critical for understanding societal well-being beyond economic indicators. Existing measures often lack fine spatial and temporal resolution. Here we introduce the Human Flourishing Geographic Index (HFGI), derived from analyzing approximately 2.6 billion geolocated U.S. tweets (2013-2023) using fine-tuned large language models to classify expressions across 48 indicators aligned with Harvard's Global Flourishing Study framework plus attitudes towards migration and perception of corruption. The dataset offers monthly and yearly county- and state-level indicators of flourishing-related discourse, validated to confirm that the measures accurately represent the underlying constructs and show expected correlations with established indicators. This resource enables multidisciplinary analyses of well-being, inequality, and social change at unprecedented resolution, offering insights into the dynamics of human flourishing as reflected in social media discourse across the United States over the past decade.

Why we think this paper is great for you:
This research provides empirical insights into how subtle social pressures can impact different groups, particularly focusing on gendered responses. It offers a deeper understanding of factors influencing participation and empowerment.

Abstract
This study analyzes whether subtle variations in the survey questionnaire phrasing influence participant engagement and whether these effects differ by gender. Building on theories of social pressure and politeness norms, it is hypothesized that presumptive phrasing would reduce engagement compared to appreciative phrasing and baseline phrasing (H1), and this effect would be more pronounced among women (H2). Mixed-effects regression models showed no significant treatment effects on any outcome and no evidence of gender moderation for 164 participants and 492 observations. The only robust finding was a small negative baseline sentiment across all participants, independent of any treatment or gender. The findings contribute to refining theoretical expectations about the conditions in which linguistic framing and gender norms shape behaviour.

Why we think this paper is great for you:
This paper explores the complexities of empowerment within AI systems, offering a nuanced perspective on how assistive AI can truly serve human goals. It directly addresses the concept of empowerment in a technological context.

Abstract
Empowerment, a measure of an agent's ability to control its environment, has been proposed as a universal goal-agnostic objective for motivating assistive behavior in AI agents. While multi-human settings like homes and hospitals are promising for AI assistance, prior work on empowerment-based assistance assumes that the agent assists one human in isolation. We introduce an open source multi-human gridworld test suite Disempower-Grid. Using Disempower-Grid, we empirically show that assistive RL agents optimizing for one human's empowerment can significantly reduce another human's environmental influence and rewards - a phenomenon we formalize as disempowerment. We characterize when disempowerment occurs in these environments and show that joint empowerment mitigates disempowerment at the cost of the user's reward. Our work reveals a broader challenge for the AI alignment community: goal-agnostic objectives that seem aligned in single-agent settings can become misaligned in multi-agent contexts.

Why we think this paper is great for you:
This paper delves into the ethical considerations of generative AI in collaborative settings, which is crucial for ensuring technology contributes positively to societal well-being. It examines how AI agents influence moral deliberation.

Abstract
Generative AI is increasingly positioned as a peer in collaborative learning, yet its effects on ethical deliberation remain unclear. We report a between-subjects experiment with university students (N=217) who discussed an autonomous-vehicle dilemma in triads under three conditions: human-only control, supportive AI teammate, or contrarian AI teammate. Using moral foundations lexicons, argumentative coding from the augmentative knowledge construction framework, semantic trajectory modelling with BERTopic and dynamic time warping, and epistemic network analysis, we traced how AI personas reshape moral discourse. Supportive AIs increased grounded/qualified claims relative to control, consolidating integrative reasoning around care/fairness, while contrarian AIs modestly broadened moral framing and sustained value pluralism. Both AI conditions reduced thematic drift compared with human-only groups, indicating more stable topical focus. Post-discussion justification complexity was only weakly predicted by moral framing and reasoning quality, and shifts in final moral decisions were driven primarily by participants' initial stance rather than condition. Overall, AI teammates altered the process, the distribution and connection of moral frames and argument quality, more than the outcome of moral choice, highlighting the potential of generative AI agents as teammates for eliciting reflective, pluralistic moral reasoning in collaborative learning.

Why we think this paper is great for you:
This paper is highly relevant for understanding how to build trust and foster effective collaboration between humans and AI systems. It provides insights into the dynamics of AI reasoning and human knowledge in decision-making.

Abstract
Effective human-AI collaboration requires humans to accurately gauge AI capabilities and calibrate their trust accordingly. Humans often have context-dependent private information, referred to as Unique Human Knowledge (UHK), that is crucial for deciding whether to accept or override AI's recommendations. We examine how displaying AI reasoning affects trust and UHK utilization through a pre-registered, incentive-compatible experiment (N = 752). We find that revealing AI reasoning, whether brief or extensive, acts as a powerful persuasive heuristic that significantly increases trust and agreement with AI recommendations. Rather than helping participants appropriately calibrate their trust, this transparency induces over-trust that crowds out UHK utilization. Our results highlight the need for careful consideration when revealing AI reasoning and call for better information design in human-AI collaboration systems.

Why we think this paper is great for you:
This paper explores how consumer attention and pricing influence welfare outcomes, offering an economic perspective on resource allocation and societal well-being. It provides a theoretical framework for understanding market dynamics.

Abstract
To choose between two discrete goods, a consumer pays attention to only those with prices below a threshold. From these, she chooses her most preferred good. We assume consumers in a population have the same preference but may have different thresholds. Similar models of bounded rationality have been studied in the empirical marketing literature. We fully characterize the model, and using observational choice data alone, we identify the welfare implications of a price change. The behavioral content of our model overlaps with an important class of random utility models, but the welfare implications are meaningfully different. The distribution of equivalent variation under our model first-order stochastically dominates that under the random utility model.