The Commonplace
Home Papers Evidence Explore Trends Syntheses Digests About 🎲 Workforce Futures

Papers (1699)

Filter: Sort:
A large procurement workflow that looks safe at the state level can still hide sizable next-step uncertainty: expanding state detail from 42 to 668 raises state-action blind mass from 0.0165 to 0.1253, implying much higher oversight needs than coarse metrics suggest; a simple maximum-probability score m(s) predicted autonomous-step accuracy on held-out data within about 3.4 percentage points.
medium theoretical Full text arxiv
The Stochastic Gap: A Markovian Framework for Pre-Deployment Reliability and Oversight-Cost Auditing in Agentic Artificial Intelligence · Biplab Pal, Santanu Bhattacharya · Mar 25, 2026
Proposed system would let autonomous agents sell verifiable 'memories', turning one-off API calls into tradable assets and potentially cutting redundant work. The idea is purely conceptual and untested: critical issues of valuation, verification at scale, privacy, and incentives remain unaddressed.
n/a theoretical Full text arxiv
Infrastructure for Valuable, Tradable, and Verifiable Agent Memory · Mengyuan Li, Lei Gao, Haoxuan Xu, Jiate Li, Potung Yu, Lingke Cheng, Yue Zhao, M... · Mar 25, 2026
A new public dataset of 55 hours of continuous expert desktop recordings and 3.6M UI annotations aims to jump-start general-purpose computer-use agents; preliminary benchmarks show current action models fail about 60% of professional tasks, indicating substantial room for improvement.
n/a descriptive Full text arxiv
CUA-Suite: Massive Human-annotated Video Demonstrations for Computer-Use Agents · Xiangru Jian, Shravan Nayak, Kevin Qinghong Lin, Aarash Feizi, Kaixin Li, Patric... · Mar 25, 2026
A redesigned generative search engine, OneSearch-V2, boosted on-site engagement and sales in live experiments—raising item clicks by ~4%, buyer conversion by ~3%, and order volume by ~2%—achieving these gains through latent-reasoning query understanding and self-distillation without increasing latency.
medium rct Full text arxiv
OneSearch-V2: The Latent Reasoning Enhanced Self-distillation Generative Search Framework · Ben Chen, Siyuan Wang, Yufei Ma, Zihan Liang, Xuxin Zhang, Yue Lv, Ying Yang, Hu... · Mar 25, 2026
A simple dynamical model calibrated to PISA and domain data finds humans risk abrupt loss of core skills once AI performs roughly 85% of tasks, but scheduled practice or occasional AI outages can largely avert the collapse.
medium theoretical Full text arxiv
The enrichment paradox: critical capability thresholds and irreversible dependency in human-AI symbiosis · Jeongju Park, Musu Kim, Sekyung Han · Mar 25, 2026
An AI system matches human experts on preschool-quality ratings (up to 88% agreement) and speeds assessments 18-fold, making monthly AI-assisted monitoring in Chinese kindergartens feasible.
medium descriptive Full text arxiv
When AI Meets Early Childhood Education: Large Language Models as Assessment Teammates in Chinese Preschools · Xingming Li, Runke Huang, Yanan Bao, Yuye Jin, Yuru Jiao, Qingyong Hu · Mar 25, 2026
Pharma’s AI promise won’t be proven by accuracy metrics alone: EFMC2 calls for new impact-oriented KPIs to measure whether AI actually raises R&D productivity and drives top-line value, arguing that strategic indicators are needed to guide adoption and innovation.
n/a commentary Summary semantic_scholar
Strategic Key Performance Indicators for AI in Lead Optimization · Theodor Theis, S. Flohr, Hayley Binch, Werngard Czechtizky, Ewa I. Chudyk, Marku... · Fetched Mar 25, 2026
Among 326 Chinese B2B SMEs, managers who view AI as useful and have AI-capable suppliers are more likely to adopt AI for procurement, and adoption is associated with a roughly 21% faster decision cycle and 16% lower procurement costs; lack of organizational resources hampers uptake.
low correlational Full text semantic_scholar
Research on the Adoption of Artificial Intelligence and Procurement Decision-Making Behavior in B2B Scenarios Based on Quantitative Analysis · Bojun Liu · Fetched Mar 25, 2026
A self-hosted, 8B-parameter model for Dream11's CriQ reduces text-to-SQL input size by over 99%, eliminates costly external API calls, and improves execution success to 98.4% and semantic accuracy to 92.5%, outperforming a Gemini Flash 2.0 baseline; the approach offers a practical route to lower-latency, lower-cost production deployment for domain-specific conversational SQL applications.
medium descriptive Full text arxiv
Schema on the Inside: A Two-Phase Fine-Tuning Method for High-Efficiency Text-to-SQL at Scale · Chinmay Soni, Shivam Chourasia, Gaurav Kumar, Hitesh Kapoor · Mar 25, 2026
Teaching a time-series foundation model basic economic rationality improves demand forecasts: Chronos-2 fine-tuned on synthetic, GARP-consistent consumer histories predicts real consumer choices markedly better than the zero-shot model across all tested horizons. The result indicates economic theory can supply structured synthetic data that acts as a powerful forecasting prior.
medium quasi_experimental Full text arxiv
GARP-EFM: Improving Foundation Models with Revealed Preference Structure · Victor H. Aguiar, Nail Kashaev · Mar 25, 2026
Generative AI used for legal drafting can flip into producing convincing but fictitious case law when internal model states cross a calculable threshold; this deterministic failure mode — not mere randomness — means lawyers, courts and regulators must adopt verification protocols rather than treating these systems as black boxes.
low theoretical Full text arxiv
When AI output tips to bad but nobody notices: Legal implications of AI's mistakes · Dylan J. Restrepo, Nicholas J. Restrepo, Frank Y. Huo, Neil F. Johnson · Mar 25, 2026
Firms with stronger AI development shift income toward capital and away from labor, but the effect varies across industries and ownership types and is dampened by financing constraints.
medium correlational Full text openalex
THE IMPACT OF ARTIFICIAL INTELLIGENCE ON ENTERPRISE INCOME DISTRIBUTION  · Jiang Zhan, JianHeng Ye, Guofeng Liang, Fanrong Feng · Mar 25, 2026
Digital technologies — from automation and IIoT to ERP and AI — are strongly associated with rising labor productivity across Kazakhstan's industry. Case studies of major firms and aggregate regressions suggest digitalization, rather than workforce size or hours, drives most efficiency gains, but causality is not robustly isolated.
low correlational Full text openalex
Digitalization and labor costs: efficiency of industrial enterprises in Kazakhstan · E. S. Akanova, А. Z. Kapenova, Y. U. Uzun, А. А. Mutaliyeva · Mar 25, 2026
AI innovation and 'recasting' are prerequisite for platform firms to gain competitive edge; three distinct AI-driven configurations—dominant, subsidiary, and collaborative—produce advantage, with innovation platforms outperforming transaction platforms.
medium correlational Summary openalex
How AI Enables Platform Enterprises to Build Competitive Advantages: A Configurational Analysis from the Perspective of Situated AI Theory · Xuguang Guo, Ying Teng, Huayong Du · Mar 25, 2026
Combining humans and agentic AI can triple media-production throughput in pilots and raise creative-quality metrics by roughly a quarter to over half, according to illustrative case studies — though the results come from limited, non-randomized deployments. Cyborg workflows offer a practical roadmap for media teams, but claims need independent, rigorous validation before being taken as generalizable productivity gains.
low descriptive Summary openalex
Cyborg Workflows Merging Human Judgment and Agentic AI for Digital Media Transformation · P. Selvaprasanth · Mar 25, 2026
Algorithmic oversight in public transport—from constant GPS tracking to automated deactivations—correlates with higher anxiety, burnout and chronic precarity among drivers; authors call for transparency, mandatory mental-health risk audits, human review of sanctions and wage protections to curb psychological harm.
medium review_meta Full text openalex
Algorithmic Control and Psychological Risk in Digitally Managed Public Transport Systems: Implications for Occupational Mental Health · Johnson O. Oloyele, ADEDOYIN IBOSIOLA, Goodness Olaleye, Damilola Elizabeth Baka... · Mar 25, 2026
Public MCP repositories show a rapid shift toward agent tools that take direct action: software-development agents dominate the ecosystem, and usage of action-enabled tools climbed from 27% to 65% in 16 months, including agents capable of financial transactions. The dataset suggests regulators should monitor the tool layer as well as model outputs to detect deployment risks beyond model-level behaviour.
medium descriptive Full text openalex
How are AI agents used? Evidence from 177,000 MCP tools · Merlin Stein · Mar 25, 2026
AI forecasting and inventory tools meaningfully improve inventory outcomes and operational resilience in smart manufacturing, but benefits materialize only where firms have clean data, trained personnel and the right organizational capabilities.
low descriptive Summary openalex
Assessing the Effectiveness of AI-Driven Techniques for Demand Forecasting and Inventory Optimization in Smart Manufacturing · Matthew Anderson · Mar 25, 2026
China’s ‘intelligent manufacturing’ pilot program measurably improved ESG performance and curtailed greenwashing among listed manufacturers — raising ESG scores by roughly 0.14 and strengthening credibility of sustainability reporting, especially for competitive, growing, and capital‑scarce firms.
medium quasi_experimental Summary openalex
Intelligent Manufacturing Demonstration Projects Driving Corporate ESG Ratings: An Analysis Based on Innovation Efficiency and Cost Management · Guo Hu, Bing Li · Mar 25, 2026
As automation from IoT, AI and robotics reshapes incomes, the paper proposes tying social security contributions to technology-generated income to shore up social protection—and warns that implementing such a scheme will face measurement, institutional and political hurdles.
n/a theoretical Full text openalex
IoT, artificial intelligence, cloud computing and robotics and the contributory principle on social security · Marco António Cabeçais de Carvalho · Mar 25, 2026
Cheaper AI production may swell output without broad entrepreneurial gains: as more creators enter low-marginal-cost markets, scarce human attention dilutes per-producer returns and concentrates rewards among a few winners.
n/a theoretical Full text arxiv
The Economics of Builder Saturation in Digital Markets · Armin Catovic · Mar 24, 2026
The EU's AI Act is poorly equipped to govern autonomous AI agents: rules built for conventional systems leave enforcement, monitoring, and resourcing gaps that risk unchecked failures, misuse, and unequal access unless policymakers rapidly revise institutional design.
n/a commentary Full text arxiv
Regulating AI Agents · Kathrin Gardhouse, Amin Oueslati, Noam Kolt · Mar 24, 2026
Sellers are quietly optimizing for AI: after ChatGPT’s debut, Etsy listings grew richer in machine-readable cues — a form of 'mecha-nudging' that steers algorithmic recommendations without changing human-facing choices.
medium quasi_experimental Full text arxiv
Mecha-nudges for Machines · Giulio Frey, Kawin Ethayarajh · Mar 24, 2026
A targeted uncertainty-visualization helps people handle more complex AI-mediated negotiations: in a lab property-rental study, the Bayesian 'agreement probability' display boosted deal quality and speed and prevented loss of human control once issues exceeded three.
medium rct Full text arxiv
From Overload to Convergence: Supporting Multi-Issue Human-AI Negotiation with Bayesian Visualization · Mehul Parmar, Chaklam Silpasuwanchai · Mar 24, 2026
A task-based CES framework lets researchers read automation from routine macro data; applying it to Japanese manufacturing reveals rising automation driven by capital deepening despite weak productivity gains.
medium theoretical Full text openalex
The macroeconomics of automation · Hideki Nakamura, Masakatsu Nakamura, Shota Moriwaki · Mar 24, 2026