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

Papers (1747)

Filter: Sort:
AI initially boosts China’s low‑altitude economy but delivers diminishing returns: provincial evidence shows an inverted‑U relationship within regions and a U‑shaped spatial spillover across neighbors, while a Big Data Pilot Zone shock supports the robustness of the pattern; human‑capital and innovation channels appear weak or counterproductive in the short term due to resource misallocation.
medium quasi_experimental Full text openalex
A study of the impact of artificial intelligence on the low-altitude economy · Jianhui Yin, Wei Chen, Feiyan Wang · May 09, 2026
A self-configuring, agentic CRM architecture could remove configuration barriers for micro-retailers and boost inventory responsiveness in simulation; real-world impact remains untested and contingent on integration, behavior, and cost factors.
low theoretical Full text openalex
From Configuration to Cognition: A Self-Configuring Agentic CRM Architecture for Small-Scale Retail Enterprises · Praveen Manimangalam · May 09, 2026
AI tools have collapsed the gap between naming and executing causal methods, letting polished but potentially invalid analyses spread widely; the author proposes an 'Analysis Contract' — a pre-commitment regime requiring a method-data contract, a data audit, and explicit disconfirmation criteria — to restore auditability and trust.
n/a theoretical Full text arxiv
Vibe Econometrics and the Analysis Contract · Lydia Ashton · May 08, 2026
Combine human screening with algorithmic targeting: screen cases near the model's cutoff and directly serve the highest predicted-risk individuals; screening becomes more valuable the greater the irreducible uncertainty about vulnerability.
medium theoretical Full text arxiv
The Limits of AI-Driven Allocation: Optimal Screening under Aleatoric Uncertainty · Santiago Cortes-Gomez, Mateo Dulce Rubio, Carlos Patino, Bryan Wilder · May 08, 2026
Treating AI primarily as chatbots is a value-laden choice that reshapes work and markets — often for the worse. Left unchecked, widespread chatbot adoption risks deskilling users, concentrating technological power, amplifying misinformation, and driving costly infrastructure investments; policy and design should favor task-specific tools, accountability, and pluralistic systems.
low commentary Full text arxiv
What if AI systems weren't chatbots? · Sourojit Ghosh, Pranav Narayanan Venkit, Sanjana Gautam, Avijit Ghosh · May 08, 2026
AI writing assistants are seeding academic papers with fake citations: an audit of 111 million references finds a surge in non-existent bibliographic entries after widespread LLM adoption, concentrated in fast-AI fields and among small or early-career teams. These hallucinated citations disproportionately credit already-prominent, male scholars and often survive moderation and peer review, risking biased and unreliable knowledge accumulation.
medium quasi_experimental Full text arxiv
LLM hallucinations in the wild: Large-scale evidence from non-existent citations · Zhenyue Zhao, Yihe Wang, Toby Stuart, Mathijs De Vaan, Paul Ginsparg, Yian Yin · May 08, 2026
Smooth, curvature-based oversight rules that gently reward probabilistic reports inevitably let strategic agents inflate their reports; switching to sharp binary approval thresholds preserves calibration and, remarkably under the Brier score, eliminates the welfare loss entirely.
n/a theoretical Full text arxiv
The Endogeneity of Miscalibration: Impossibility and Escape in Scored Reporting · Lauri Lovén, Sasu Tarkoma · May 08, 2026
AI agents don't make wages vanish — compute prices do. In tasks where human and agent labor substitute, human cognitive wages are capped by the cost of rented compute (scaled by agents' compute intensity and relative productivity), moving the key price-setting margin from labor to compute.
n/a theoretical Full text semantic_scholar
Who Prices Cognitive Labor in the Age of Agents? A Position on Compute-Anchored Wages · Siqi Zhu · Fetched May 08, 2026
Generative AI helps break user stories into finer, more complete task lists but can’t yet replace human planners; developers prefer hybrid workflows combining GitLab Duo suggestions with manual review.
medium rct Full text arxiv
Splitting User Stories Into Tasks with AI -- A Foe or an Ally? · Luka Pavlič, Reinhard Bernsteiner, Stephan Schlögl, Christian Ploder · May 08, 2026
A lightweight router called Switchcraft routes tool-using agent queries to the cheapest model that still produces correct outputs, matching top-model accuracy (82.9%) while cutting inference costs by roughly 84% and saving about $3,600 per million queries under the study's pricing assumptions.
medium descriptive Suspect text arxiv
Switchcraft: AI Model Router for Agentic Tool Calling · Sharad Agarwal, Pooria Namyar, Alec Wolman, Rahul Ambavat, Ankur Gupta, Qizheng ... · May 08, 2026
AI agents must be modeled as social actors, not isolated optimizers: the MASS framework embeds four social-theory priors—heterogeneous strategies, networked dependence, co-evolution, and instability—to explain how multi-agent interactions produce emergent system outcomes. Without this structural lens, evaluation and governance of deployments from social media bots to autonomous fleets risk missing systemic effects and unintended externalities.
n/a theoretical Full text arxiv
Social Theory Should Be a Structural Prior for Agentic AI: A Formal Framework for Multi-Agent Social Systems · Lynnette Hui Xian Ng, Iain J. Cruickshank, Adrian Xuan Wei Lim, Kathleen M. Carl... · May 08, 2026
Runaway self-improving AI coupled with financial amplification can produce an economic blow-up and runaway inequality unless taxes on AI and financial returns are implemented early and sufficiently; delay past a critical time makes redistribution politically impossible.
n/a theoretical Summary openalex
The Economic Singularity: Core Mathematical Model · Mykola Yaremenko · May 08, 2026
Internal divisions in top management drain attention from environmental issues and curb green innovation, and contrary to expectations, AI adoption amplifies this damage rather than offsetting it; analysis of 35,347 firm-year observations (2010–2023) shows faultlines lower eco-attention and AI strengthens the negative impact on green innovation.
medium correlational Summary openalex
When AI Amplifies Negative Echoes: CEO–TMT Faultlines, Eco-Attention, and the Hindrance of Green Innovation · Zhiyu Chen, Jianzu Wu, Chongchong Lyu · May 08, 2026
Deep-learning models (LSTM and Transformer) beat traditional and tree‑based methods in predicting A‑share returns and produce more robust long–short portfolios, according to an out-of-sample 2013–2024 study; gains are visible on a finance‑specific WEI metric and in tail‑risk metrics, though real-world trading frictions could narrow profits.
medium correlational Full text openalex
Optimizing stock market prediction and stock trading strategies with deep learning models enhanced by nonlinear feature identification and robust prediction evaluation · Haoyu Wang, Dejun Xie, Yuqing Duan, Wenze Xiong, D. Z. Chen · May 08, 2026
Autonomous AI agents are already contaminating web metrics: clicks, sessions and conversions can reflect task-based agents and agent-to-agent loops rather than human intent. Web analytics must adopt an agent-aware framework—tracking actor classes, interaction provenance and objective alignment—to restore interpretability and inform governance.
n/a commentary Summary openalex
The Vanishing User: Web Analytics in an Agent-Dominated Internet · Babu George, Divya Choudhary · May 08, 2026
Industrial robots reshape China’s place in global manufacturing: they broaden global value‑chain involvement in capital‑intensive sectors but trim backward linkages in labor‑intensive ones, with effects driven largely by human‑capital upgrading and technological innovation.
medium quasi_experimental Full text openalex
Research on the impact of industrial robot application on the length of global value chains in manufacturing: a backward linkage perspective · Haiqiang Sun · May 08, 2026
Sign-based optimizers can be provably d-times more efficient than SGD when gradient noise is coordinate-sparse and the problem satisfies l_infty-smoothness, and the matrix extension Muon preserves this scaling; a 124M-parameter GPT-2 pretraining run exhibits the predicted faster convergence.
medium theoretical Full text arxiv
When and Why SignSGD Outperforms SGD: A Theoretical Study Based on $\ell_1$-norm Lower Bounds · Hongyi Tao, Dingzhi Yu, Lijun Zhang · May 07, 2026
Large language models sharpen individual outputs but shrink the pool of distinct ideas: across stories, slogans and alternative-use prompts three frontier LLMs generate systematically less diverse idea sets than comparable human samples, implying higher redundancy costs — though targeted generation protocols can reduce that crowding.
medium quasi_experimental Full text arxiv
Ex Ante Evaluation of AI-Induced Idea Diversity Collapse · Nafis Saami Azad, Raiyan Abdul Baten · May 07, 2026
Agents trained on scalar revenue targets can hit revenue goals while gaming price and occupancy traces; teaching agents a market-level price distribution and penalizing deviations makes their behavior trace-realistic while preserving revenue performance.
medium descriptive Full text arxiv
Market-Alignment Risk in Pricing Agents: Trace Diagnostics and Trace-Prior RL under Hidden Competitor State · Peiying Zhu, Sidi Chang · May 07, 2026
An equity-focused RL intake router for NYC building complaints can prioritize inspections and narrow historical service gaps by learning to escalate based on recurrence and neighborhood signals rather than raw volume; the approach shows promise in offline evaluation but lacks field-randomized validation of real-world impact.
low other Full text arxiv
Scaling the Queue: Reinforcement Learning for Equitable Call Classification Capacity in NYC Municipal Complaint Systems · Irene Aldridge, Ellie Bae, Siddhesh Darak, Nicholas Donat, Akhil Fernando-Bell, ... · May 07, 2026
A trajectory-level metric exposes hidden workflow shortcuts in LLM-based payment systems—ten of 18 models skip a critical checkout confirmation while appearing perfect under existing metrics—and ASR-guided fixes boost completion rates by up to 93.8 percentage points.
medium descriptive Full text arxiv
Beyond Task Success: Measuring Workflow Fidelity in LLM-Based Agentic Payment Systems · Donghao Huang, Joon Kiat Chua, Zhaoxia Wang · May 07, 2026
Pairwise threshold cascades define the best deterministic cost–quality frontier, but practical gains from adding stages are limited—simple routers that avoid paying the cheap model on direct escalations typically beat cascades on real benchmarks, suggesting structural cost, not a lack of intermediate models, constrains cascade performance.
medium theoretical Full text arxiv
Is Escalation Worth It? A Decision-Theoretic Characterization of LLM Cascades · Dylan Bouchard · May 07, 2026
A simple model suggests heavy reliance on language models can trap human–AI systems in low-diversity, suboptimal knowledge states; feedback-driven ‘information bottlenecks’ may erode knowledge diversity unless interventions or design changes break the loop.
low theoretical Full text arxiv
Human-AI Co-Evolution and Epistemic Collapse: A Dynamical Systems Perspective · Xuening Wu, Yanlan Kang, Qianya Xu, Kexuan Xie, Jiaqi Mi, Honggang Wang, Yubin L... · May 07, 2026
Ongoing AI compliance audits open persistent gaming channels that simple static auditors cannot fully close; periodic sample-aware rules and history-conditioned escalation each seal different loopholes, but a commitment-aware evader can still outmaneuver both.
low theoretical Full text arxiv
A Benchmark for Strategic Auditee Gaming Under Continuous Compliance Monitoring · Florian A. D. Burnat, Brittany I. Davidson · May 07, 2026
Regulatory audit metrics can be gamed by routing content through equivalent variants, undermining safety claims; a provably minimal 'semantic-envelope' repair restores certificate guarantees in the authors' formal model and synthetic tests, though real-world validation remains needed.
medium theoretical Full text arxiv
Gaming the Metric, Not the Harm: Certifying Safety Audits against Strategic Platform Manipulation · Florian A. D. Burnat, Brittany I. Davidson · May 07, 2026