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

Papers (1699)

Filter: Sort:
A new nationwide corpus unlocks local law for machine use: LOCUS collects ordinances from 9,239 U.S. cities and counties and provides a county-harmonized layer covering 2,309 counties, together with OCR processing, coverage metadata, and trained classifiers to analyze regulatory features such as opacity and paternalism.
n/a descriptive Full text arxiv
Freeing the Law with LOCUS: A Local Ordinance Corpus for the United States · Denis Peskoff, Joe Barrow, Christopher Vu, Diag Davenport · Jun 17, 2026
Complex symbolic scaffolding can make long-horizon LLM collaboration worse — models become self-referential and lose semantic contact ('Index Sickness'). A pragmatic fix that separates baseline logs physically and restores natural-language purpose cut AI instruction volume by roughly 75% and prevented recurrence in subsequent sessions.
low descriptive Full text arxiv
Written by AI, Managed by AI: Semantic Space Control and Index Sickness Elimination Across 391 Consecutive Sessions · Hui Zhang, Shuren Song · Jun 17, 2026
The web's human-first assumption is broken by AI intermediaries; to preserve truth and fair economics the authors propose making agents first-class: agent identity headers and rate limits, a tokenized, intent-based subscription model that charges agents like their human principals, and ATML plus cryptographic provenance to stop self-reinforcing AI content loops.
n/a theoretical Full text arxiv
Towards an Agent-First Web: Redesigning the Web for AI Agents · Eranga Bandara, Ross Gore, Ravi Mukkamala, Asanga Gunaratna, Safdar H. Bouk, Xue... · Jun 17, 2026
Removing manipulable features from high-stakes predictors can backfire; jointly choosing which features to keep and how strongly to regularize them delivers better performance and reduces gaming, according to a formal analysis and a healthcare benchmark.
medium theoretical Full text arxiv
Strategic Feature Selection · Jivat Neet Kaur, Pratik Patil, Divya Shanmugam, Emma Pierson, Michael I. Jordan,... · Jun 17, 2026
Seeding a robot with a reusable episodic memory of prior teamwork lifts urban search-and-rescue performance—rescue success jumps from 25.7% to 41.3% and average task time falls by nearly five minutes, with the largest gains occurring at interaction start.
medium rct Full text arxiv
Improving Human-Robot Teamwork in Urban Search and Rescue Through Episodic Memory of Prior Collaboration · Taewoon Kim, Emma van Zoelen, Mark Neerincx · Jun 17, 2026
AI research lifts EU biofuel output: a 1% rise in AI publications is associated with roughly a 0.47% cumulative increase in biofuel production (mostly after two years), while AI venture capital yields smaller gains; effects are strongest in low-production countries.
low correlational Summary semantic_scholar
Digital innovation for a greener future: the role of artificial intelligence in Europe's biofuel transition · Tufan Sarıtaş, Emin Ahmet Kaplan, Yasin Büyükkör, Alper Aslan · Fetched Jun 17, 2026
Integrating LLM-generated user profiles into DiDi's dispatcher raised prediction AUC by up to 6.14% and produced a measurable lift in platform revenue and service quality (≈+0.47% GMV, +0.33% completion rate) in a 14‑day production A/B test. The deployment also reduced cancel-before-accept rates, showing that utility-aligned LLM profiling can deliver real economic value when engineered for low-latency production use.
high rct Full text arxiv
ProfiLLM: Utility-Aligned Agentic User Profiling for Industrial Ride-Hailing Dispatch · Tengfei Lyu, Zirui Yuan, Xu Liu, Kai Wan, Zihao Lu, Li Ma, Hao Liu · Jun 17, 2026
Digital technologies are reshaping China’s economy: AI, IoT and platforms lift manufacturing competitiveness and expand services, but agriculture is being left behind; policy gaps and uneven infrastructure widen the digital divide.
medium descriptive Summary semantic_scholar
How to Utilize New Technologies to Improve Productivity · Yishu Liu, Yanxiu Guo, Zhengran Gao · Jun 17, 2026
AI ties environmental and financial systems into a tightly coupled triad: it boosts foresight and allocation but amplifies systemic fragility and tipping-point risk. Policymakers should treat sustainability as an AI-mediated co-evolution requiring integrated resource, finance and AI governance.
n/a theoretical Summary semantic_scholar
Synergy in the economics of sustainable development and Artificial Intelligence: combining aspects of natural resources and financial security · S. Petrukha, N. Petrukha, Olena Stoliarenko, G. Ortina, Dmytro Plakhotnii, Rusla... · Jun 17, 2026
Specialized decision-makers plus many simple agents can outperform a society of uniformly informed individuals: when information is costly, delegation and cognitive division of labour preserve much of knowledge's value while cutting decision costs.
n/a theoretical Full text openalex
The Simplicity Paradox: Why Evolution Does Not Produce Universally Complex Agents · Teddy Lazebnik · Jun 17, 2026
AI and machine-learning models generally outperform traditional time-series methods in forecasting economic indicators, though data gaps and study heterogeneity limit firm conclusions; early tests show ChatGPT and Google Gemini generate competitively accurate predictions for EU GDP and unemployment.
medium review_meta Full text openalex
A Review of Nowcasting, Forecasting and AI in Economic Indicator Prediction · Özge Var, Alptekin Durmuşoğlu, Türkay Dereli · Jun 17, 2026
AI-driven productivity is not self-sustaining: gains must reach household incomes or economies hit an 'absorption' ceiling that limits consumption and welfare improvements. The paper reframes the debate by linking technological efficiency to distributional transmission through wages, prices, investment and public policy and proposes measurable indicators for future empirical testing.
n/a review_meta Full text openalex
Artificial Intelligence, Labour Income and Effective Demand: A Theoretical Framework for the Distributional Absorption Threshold of AI-Induced Productivity · Narcis Eduard Mitu · Jun 17, 2026
Manufacturers that turn data into reusable assets show higher ‘digital intelligence’: BERT-based analysis of Chinese listed firms (2014–23) finds robust positive links driven by better AI talent, tech innovation and collaboration, with bigger gains in firms that can absorb and govern AI effectively.
medium correlational Full text openalex
Data Valuation, AI Ecosystem Restructuring and Enterprise Digital Intelligence Transformation · Zhijun Zhou, Yanhong Liu, Xianfu Wu, Fang Liu, Jiexia Li · Jun 17, 2026
A machine-learning screening tool can modestly identify potentially disruptive AI patents before their impact appears — an AdaBoost model trained on patent metadata flags top-5% CD disruptors with 27% precision and 43% recall. The method is best used to generate candidates for expert review, not to certify future market disruption.
medium correlational Full text openalex
A CD index guided ensemble framework for screening potentially disruptive patent candidates in artificial intelligence · Hui Li, Jie Wu, Xiaodong Xie, Liu Y, T Wang · Jun 17, 2026
China’s green public procurement is linked to measurable falls in listed firms’ carbon intensity, and the effect is amplified when firms adopt AI or obtain green subsidies; state‑owned and high‑tech firms in digitally advanced regions benefit most.
medium quasi_experimental Summary openalex
The Impact of Government Green Procurement on Corporate Carbon Emission Reduction: A Dual Mediation Perspective of Artificial Intelligence and Green Finance · Z K Zhang, Jianmin Wu · Jun 17, 2026
National AI development alone does not raise adult lifelong-learning rates in Europe; rising AI R&D is followed by declines in participation whereas government and policy activity around AI is linked to higher later training uptake.
medium quasi_experimental Full text openalex
National AI development and adult lifelong-learning participation: Evidence for knowledge-transfer policy in European countries · Nadiia Аrtyukhova, Аrtem Аrtyukhov, Elena Kašťáková, Каріна Таранюк, Alvina Orie... · Jun 17, 2026
Rural digital adoption boosts farm efficiency and strengthens farmers’ resilience: household panel evidence from China (2014–2022) shows digital tools raise production efficiency—especially for younger farmers and grain producers in well-connected villages—and translate into steadier income, greater asset accumulation and stronger risk-coping ability.
medium quasi_experimental Full text openalex
Digital technology adoption, agricultural production efficiency, and farmers’ livelihood resilience · Li Qi, Haoxiang Zhao, Fengyi Wang, Guoge Yang, Yun Jiang · Jun 17, 2026
More collaborators can hurt as much as help: simulated human-AI teams often see performance drop when members are added without coordination rules, but adding shared group memory plus approval gates substantially improves outcomes—especially for three-person teams.
medium quasi_experimental Full text arxiv
Searching for Synergy in Shared Workspace Human-AI Collaboration · Nachiket Kotalwar, Rohini Das, Carolyn Rose · Jun 16, 2026
Most AI-generated test files do little verification — 80% lack strong assertions — but when tests contain explicit oracle signals PRs are noticeably more likely to be accepted, increasing merge odds by about 28% after controls.
medium correlational Full text arxiv
All Smoke, No Alarm: Oracle Signals in Agent-Authored Test Code · Dipayan Banik, Kowshik Chowdhury, Shazibul Islam Shamim · Jun 16, 2026
Treat robot flash as scarce capital: an endurance shadow price makes memory placement a cost problem, and in long-horizon manipulation workloads valuable memories tend to be routed off-device while cheaper QLC flash makes endurance economically binding.
medium descriptive Full text arxiv
Memory as a Wasting Asset: Pricing Flash Endurance for Embodied Agents, and the Limits of Doing So · Josef Liyanjun Chen · Jun 16, 2026
As large language models become autonomous consumer agents, standard consumer theory needs revision — agentic markets can break assumptions like direct revealed preference and individual rationality, creating new alignment and market-structure challenges.
n/a theoretical Full text arxiv
LLM Consumer Behavior Theory: Foundations of a Novel Research Field · Manon Reusens, Sofie Goethals, David Martens · Jun 16, 2026
An LLM-powered system builds factory-wide process twins far faster than manual work and with high accuracy; in a single-plant case study it produced correct models (mean F1 95.2%) and cut development time by roughly 6x while deferring ambiguous, safety-critical bindings to operators to avoid mis-bindings.
medium descriptive Full text arxiv
FacProcessTwin: An LLM-Based System for Process Twin Development · Yash Pulse, Yong-Bin Kang, Abhik Banerjee, Prem Prakash Jayaraman · Jun 16, 2026
There are provable limits to how quickly pre-fine-tuning probes can shrink uncertainty about LLM performance, so some uncertainty remains irreducible and probes must be budgeted strategically; mapping tasks into Static‑Sufficient, Dynamic‑Critical and Noise‑Dominant regimes lets practitioners cut fine‑tuning costs where prediction is feasible.
medium theoretical Full text arxiv
A Risk Decomposition Framework for Pre-Hoc Fine-Tuning Prediction · Yuxiang Luo, Chen Wang, Nan Tang · Jun 16, 2026
Commercial LLMs default to recommending incumbent brands 100% of the time for identical products, but a competitor needs less than a 0.1-star edge or persuasive authority-style copy (+0.17 bias value) to break that monopoly; if every brand adopts generative engine optimization, individual recommendation gains evaporate (payoff drops from +0.802 to +0.007) and non-participants are effectively excluded.
medium quasi_experimental Full text arxiv
Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems · Xi Chu, Yupeng Hou · Jun 16, 2026
A POMDP-based validation framework lets firms audit autonomous AI agents component-by-component; in a portfolio-management backtest, belief-conditioned latent-state inference measurably improves portfolio decisions and remains robust across parameter choices.
medium theoretical Full text arxiv
Model Validation of Agentic AI Systems: A POMDP-Based Framework for Belief-State, Forecast, and Policy Validation · Matthew Francis Dixon · Jun 16, 2026