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

Papers (1699)

Filter: Sort:
Platform algorithms that price, compress and police workers’ time create a novel power dynamic — 'algorithmic time politics' — the paper argues, which undermines worker health through a chain from fatigue to burnout and depression. The author urges that occupational-health risk assessments and algorithm design account for time-based harms.
n/a theoretical Full text openalex
Predation, acceleration, and loss of control: a multilevel theoretical framework for algorithmic time politics and the occupational health of platform workers · Qiuyu Fan · Jun 11, 2026
Older companies struggle to attract AI talent: a one-standard-deviation increase in firm age is linked to a 5.2% lower share of AI workers, though R&D spending and modern infrastructure can reverse the trend; diverse boards and slow decision-making in mature firms may further slow AI integration, while firms that do integrate AI talent see gains in valuation and efficiency.
low correlational Full text openalex
The AI workforce and firm maturity: old firms, new tech · Pattanaporn Chatjuthamard, Pornsit Jiraporn, Pandej Chintrakarn, Woraphon Wattan... · Jun 11, 2026
AI can boost municipal productivity, but only when local governments have the digital maturity and governance mechanisms to manage it; without readiness, algorithmic opacity, data dependence and coordination failures can amplify systemic risks across public services.
medium review_meta Summary openalex
AI Adoption in Local Government: Productivity, Systemic Risk, and Institutional Resilience: Evidence from a PRISMA 2020 Review · Abayomi Ogunrinde, Carmen De‐Pablos‐Heredero · Jun 11, 2026
AI can accelerate and broaden government audits—cutting detection timelines from years to days—but delivering those gains requires new data infrastructure, cross-functional teams, and strong governance; a structured 24–48 month rollout with executive sponsorship and iterative pilots is essential.
medium descriptive Full text openalex
Towards AI-Augmented Public Audit Systems: A Policy and Implementation Roadmap for Us Government Agencies and Multilateral Organizations · Nyiawung Fobellah Abetoh · Jun 11, 2026
Frontier agents have become far more capable and safer: task completion on WorkBench rose from 43% (GPT-4, Mar 2024) to 89% (Claude Opus 4.8, Jun 2026) and unintended harmful actions fell from 26% to 2.5%. At the same time, the rise of open-weight models has sharply reduced the cost of reaching top-tier performance, though rare irreversible errors such as mis-sending emails persist.
medium descriptive Full text arxiv
WorkBench Revisited: Workplace Agents Two Years On · Olly Styles · Jun 10, 2026
AGI need not be a single cliff-edge: four credible routes could lead from human-level AI to far more capable systems, and bottlenecks along these routes will shape whether change is incremental or explosive.
n/a theoretical Full text arxiv
From AGI to ASI · Tim Genewein, Matija Franklin, Alexander Lerchner, Laurent Orseau, Samuel Albani... · Jun 10, 2026
AI diagnostic predictions improve chest X‑ray interpretation most for lower‑ability clinicians and for those whose confidence better matches accuracy; a replication using 68 radiologists and 11,420 paired reads confirms larger incremental value from AI for these groups.
medium quasi_experimental Full text arxiv
Revisiting the ABCs of Working with AI: A Replication with Radiologists · Daniel Martin · Jun 10, 2026
A local, append-only project memory nudges AI coding assistants away from repeating past failures and preserves an auditable provenance trail; a two-month self-study over 10 projects (207 events) demonstrates feasibility though not yet rigorous proof of productivity gains.
low descriptive Full text arxiv
PROJECTMEM: A Local-First, Event-Sourced Memory and Judgment Layer for AI Coding Agents · Ripon Chandra Malo, Tong Qiu · Jun 10, 2026
Neither laissez-faire nor blanket IP protects the future of creative content for AI: permissive access starves creators of compensation, while strong rights blunt originality — and even a high‑quality model can erode itself by causing creators to homogenize their output. A data intermediary that internalizes cross-creator externalities and subsidizes novel contributions can both preserve incentives and improve long-run model quality.
n/a theoretical Full text arxiv
Market Design for AI: Beyond the Copyright Binary · Yan Dai, Maryam Farboodi, Negin Golrezaei, Sepehr Shahshahani · Jun 10, 2026
China’s AI surge is remaking management — but the gains depend on human wisdom. Across 18 papers, AI boosts efficiency and innovation when embedded in culturally aligned leadership, trust and networked resource orchestration, yet risks psychological alienation, ethical tensions and governance shortfalls if deployed without context‑sensitive oversight.
medium review_meta Summary openalex
Guest editorial: Digital age wisdom in Chinese management: applications and challenges of digital transformation and artificial intelligence · Tachia Chin, Chien-Liang Lin, Chris Rowley, Lei Huang · Jun 10, 2026
AI in postcolonial settings often entrenches dependency rather than delivering autonomy: a synthesis of 50 studies finds 'algorithmic colonialism'—driven by data, platform and governance lock-in—produces ethical harms and limited, structurally constrained forms of resistance.
medium review_meta Full text openalex
AI ethics in postcolonial contexts: a critical synthesis of infrastructures, power, and governance · Itoro Abraham · Jun 10, 2026
Online readers are weaponizing a new 'AI slop' label: accusations have exploded on Hacker News and Reddit and now function more as social gatekeeping than as accurate detection of AI-generated prose.
medium quasi_experimental Full text openalex
"That's AI Slop, You Bot!" Studying Accusations, Evidence, and Credibility in Online Discourse Towards LLM-Generated Comments · Jason Miklian, John E. Katsos · Jun 10, 2026
AI agents rival humans in methodological choices but betray bias at interpretation: LLMs reproduce or exceed human specification diversity on a policy dataset, yet targeted prompts can flip their final verdicts from near-zero to near-complete support while leaving coefficient distributions unchanged.
medium quasi_experimental Full text arxiv
AI Coding Agents in Social Science: Methodologically Diverse, Empirically Consistent, Interpretively Vulnerable · Meysam Alizadeh, Fabrizio Gilardi, Mohsen Mosleh, Enkelejda Kasneci · Jun 09, 2026
An automated LLM pre-mediator performs comparably to professional mediators on short-term preparation measures while inferring party preferences more accurately — cutting preference-inference error by 36%; prompt tuning also reduces excessive affirmation to match human mediator baselines.
medium rct Full text arxiv
Automated Mediator for Human Negotiation: Pre-Mediation via a Structured LLM Pipeline · Jamie Bergen, Sarit Kraus · Jun 09, 2026
Large language models outpace median human experts on a suite of bio-lab tasks, and one model produced working robot scripts that assembled DNA in lab tests, underscoring rapid productivity gains and heightened biosecurity concerns.
medium descriptive Full text arxiv
ABC-Bench: An Agentic Bio-Capabilities Benchmark for Biosecurity · Andrew Bo Liu, Samira Nedungadi, Bryce Cai, Alex Kleinman, Harmon Bhasin, Seth D... · Jun 09, 2026
A new benchmark reveals vision-language models struggle with engineering reasoning: across 696 problems in five disciplines, models often produce physically inconsistent multi-step solutions, though an 8-stage automated grader aligns tightly with human scores.
n/a descriptive Suspect text arxiv
Do VLMs Reason Like Engineers? A Benchmark and a Stage-wise Evaluation · Syed Wasiq, Syed Mohamad Tawseeq, Yashwant Pravinrao Bangde, Debaditya Roy · Jun 09, 2026
Generative-AI's performance-first stacks externalize large thermodynamic and material costs; this paper offers a Regenerative Socio-Technical roadmap that reframes AI infrastructure as a system constrained by planetary limits and proposes a sustainability-focused reference architecture to fill semiconductor Scope 3 and e-waste gaps.
n/a theoretical Full text arxiv
From Stacks to Circuits: A Regenerative Socio-Technical Roadmap for AI Infrastructure within Planetary Boundaries · Han-Teng Liao, Karen Ang · Jun 09, 2026
A compact grounding adapter lets general coding agents configure complex scientific simulators minutes instead of hours, matching expert-quality GEOS decks with a reported ~36× wall-clock speedup; validation, memory, retrieval and self-improvement components drive accuracy and make the approach transferable to OpenFOAM and LAMMPS.
medium descriptive Full text arxiv
SIGA: Self-Evolving Coding-Agent Adapters for Scientific Simulation · Matthew Ho, Brian Liu, Jixuan Chen, Audrey Wang, Lianhui Qin · Jun 08, 2026
CHAP formalises human–agent collaboration by turning informal edits, approvals and handoffs into structured, signed, auditable events; the protocol and reference implementation aim to make multi-agent, multi-human production workflows accountable and portable across teams.
n/a descriptive Full text arxiv
Collaborative Human-Agent Protocol (CHAP) · Arsalan Shahid, Gordon Suttie, Philip Black · Jun 08, 2026
Rapid AI expansion could add as much as 723 TWh of electricity demand by 2050 and force up to 226 GW of extra generation, driving higher local power costs and intermediate-year emissions unless efficiency and grid flexibility are scaled up; even moderate growth raises hourly firm-generation needs and pushes LCOE up by ~€35/MWh in key hubs.
medium descriptive Full text arxiv
Powering the Future of AI: Navigating the Trade-offs for Europe's Energy Transition and Net-Zero Goals · Mohammad Hemmati, Gbemi Oluleye, Vassilis M. Charitopoulos · Jun 08, 2026
Access to a proprietary, curated evidence corpus — not just better prompting or tools — determines how well an AI can value drug assets: adding the Noah AI corpus raised coverage and completeness-aware decision-utility nearly fourfold compared with public scaffolds, which mainly improved calibration but left a strict factual ceiling.
medium quasi_experimental Full text arxiv
AI Scientists Are Only as Good as Their Evidence: A Stratified Ablation of Proprietary Data and Reasoning Skills in Drug-Asset Valuation · Yinan Wang · Jun 08, 2026
A zero‑trust orchestration framework lets 'smart fabs' prove compliance with EU sustainability and carbon rules without revealing trade secrets by running virtual metrology and federated learning in hardware-isolated enclaves and exporting signed compliance tokens via secure data connectors.
low theoretical Full text arxiv
Trustworthy Smart Fabs via Professional Proxies: Scaling Safe and Sustainable by Design (SSbD) through Industrial Data Spaces · Han-Teng Liao, Chang-Yi Kao, Karen Ang · Jun 08, 2026
Agentic AI now handles the vast majority of routine cloud-network outages: a multi-agent system deployed at a hyperscaler autonomously resolves over 90% of common incidents, cutting the need for human intervention while retaining safety through authorization layers and rollback verification.
medium descriptive Full text arxiv
Autonomous Incident Resolution at Hyperscale: An Agentic AI Architecture for Network Operations · Arun Malik · Jun 08, 2026
Endogenous data makes automation self-reinforcing but slow: data both improves existing automation and widens the automation frontier, yet the share of tasks performed by labor falls only asymptotically as a power law. The model predicts explosive aggregate growth when capital is endogenous but persistent wage stagnation and a generic role for policy to reorient data accumulation toward socially valuable directions.
n/a theoretical Full text openalex
Data-Driven Automation · Maryam Farboodi, Andrew Koh, Anchi Xia · Jun 08, 2026
Adding LLM-generated predictions to regression adjustment modestly sharpens randomized-experiment estimates and will not worsen unbiased estimates, with the biggest gains when rich text or unstructured data are available.
medium theoretical Full text arxiv
AI-Assisted Variance Reduction in Randomized Experiments · David Arbour, Eli Ben-Michael, Avi Feller, Apoorva Lal, Lo-Hua Yuan · Jun 07, 2026