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

Papers (1699)

Filter: Sort:
Generative audiovisual AI could sharply cut the cost and time of content production, but unresolved data rights, governance gaps and authenticity concerns mean adoption and economic benefits are far from guaranteed; regulation, provenance systems and legal rulings will critically shape who captures value and whether society accepts synthetic media.
low review_meta Summary openalex
Ethical and societal challenges to the adoption of generative AI in communication. · Bruno Caldas Vianna, Carles Sora, Josep Torrelló · Jun 23, 2026
Generative audiovisual AI promises large efficiency gains in content production but will be hampered — and reshaped — by legal, ethical and trust frictions across training data, deployment and output, concentrating advantage with data- and deployment-rich firms and creating new markets for provenance and verification.
n/a review_meta Summary openalex
Ethical and societal challenges to the adoption of generative AI in communication. · Bruno Caldas Vianna, Carles Sora, Josep Torrelló · Jun 23, 2026
The rush to create and hoard internet-scale datasets for frontier AI hides mounting environmental and social costs: storage and curation of hundreds of thousands of datasets drive rising carbon burdens, while data-labour risks and representation harms disproportionately fall on workers in the Global South.
medium descriptive Full text openalex
How Hyper-Datafication Impacts the Sustainability Costs in Frontier AI · Sophia N. Wilson, Sebastian Mair, Mophat Okinyi, Erik B. Dam, Janin Koch, Raghav... · Jun 23, 2026
Countries that adopt AI tend to show higher income inequality, but strong education systems and richer economies blunt that effect; developing nations face a pronounced digital divide where AI adoption accompanies rising inequality.
low correlational Full text openalex
Analyzing the Impact of Artificial Intelligence Adoption on Economic Inequality: A Cross-Country Data-Driven Study · Vinayak Kumar, Raghu Raja Mehra · Jun 23, 2026
AI acts less like a neutral productivity tool and more like a rent-bearing asset, concentrating returns through access control and eroding the value of higher education; fixing labour-market harms requires tackling ownership and access, not just skill policies.
n/a theoretical Full text openalex
From human capital to asset ownership: AI as rentier asset · Gerbrand Tholen · Jun 23, 2026
AI is shifting governance from welfare-era and voluntary CSR models toward integrated, AI-enabled institutional governance that embeds monitoring, prediction and coordination across public and corporate domains. Theoretical synthesis contends this evolution reconceives the welfare state and corporate responsibility as components of a unified, sustainability-oriented socio-technical architecture.
n/a theoretical Summary openalex
Digital transformation of welfare state: the role of AI in the context of sustainable development paradigms within the evolution of corporate governance · О. Ю. Бобровська, Maksym Makarenko, Nataliia Andrushkevych, Oleksandra Niema, Na... · Jun 23, 2026
China's rollout of local government data platforms meaningfully lifted listed firms' digital-technology patenting, encouraging deeper application of digital technologies; the effect appears to operate by widening use cases and lowering uncertainty for corporate innovation.
medium quasi_experimental Full text openalex
Public Data and Digital Technology Innovation — Empirical Evidence from the Establishment of Chinese Government Data Platforms · Yue Wang · Jun 23, 2026
Static 'GPTs are GPTs' exposure scores have become a convenient proxy for AI’s labour risks, but they were never designed for policy decisions; relying on them without dynamic, usage and worker-centered measurement risks misleading who will be affected and when. Fixing this requires new measurement infrastructure, participatory research, and sustained coordination between researchers and policymakers.
n/a commentary Full text arxiv
AI Exposure Scores: what they measure, what they miss, and what comes next · Campbell Lund, Thomas Euyang, Zanele Munyikwa, Marzieh Fadaee · Jun 22, 2026
Human-AI teams succeed when systems enable coordination, clear expectations and adjustable autonomy rather than merely dividing tasks. Yet the literature is fragmented and skewed toward lab demonstrations, calling for more longitudinal, real-world studies to judge sustained workplace impact.
medium review_meta Full text semantic_scholar
From testbeds to high-stakes work: a review of Human-AI teaming domains and teaming factors · Shaida Kargarnovin, C. Hernandez, D. Reiners, C. Cruz-Neira, G. Bochenek, Waldem... · Fetched Jun 22, 2026
Firms that redesign how people and AI work together report markedly better returns: those intentionally architecting human–AI interactions are twice as likely to exceed AI investment targets and 2.5 times more likely to report superior financial performance. The result is drawn from a large Deloitte executive survey and illustrates associations, not proven causal effects.
low descriptive Summary semantic_scholar
Designing Human-Machine Collaboration: Strategic Imperatives for the AI-Powered Workplace · Jonathan H. Westover · Fetched Jun 22, 2026
Group-level selection of prompts drives cooperation among simulated LLM agents, while individual-level selection causes collective defection; a matching analytic model and robustness checks reveal a critical threshold for prosociality and model-specific anticipatory responses (notably in GPT-5.4).
medium theoretical Full text arxiv
Group Selection Promotes Prosocial Prompts in Populations of LLM Agents · Luis Celiktemel, Edward Eichhorn, Levin Brinkmann, Robin Schimmelpfennig, Aron V... · Jun 22, 2026
Procedural memory measurably improves LLM agents on enterprise tasks: a single refinement raises aggregate accuracy by 3.7–6.7 points, and skills learned from diverse multi-model traces transfer across models with 73.1% test accuracy; some skills generalize broadly while others become role-specific and lose transferability.
medium descriptive Full text arxiv
Managing Procedural Memory in LLM Agents: Control, Adaptation, and Evaluation · Julia Belikova, Rauf Parchiev, Evgeny Egorov, Grigorii Davydenko, Gleb Gusev, An... · Jun 22, 2026
LLM recommendations do not produce ubiquitous winner-takes-all outcomes: across 3,750 brand-free queries to three models, recommendation concentration was moderate and top-brand agreement across models under 42%, though displacement patterns varied sharply by industry.
medium descriptive Full text arxiv
Who Owns the AI Recommendation? A Multi-Industry Empirical Map of Brand Category Ownership Across Large Language Models · Dmitrij Żatuchin · Jun 22, 2026
A specialized IPO benchmark finds modern LLMs can handle long, complex S-1 filings: Alibaba Qwen 3.7 Max scores 79.4% on SpaceX IPO diligence questions, while Xiaomi MiMo-2.5 Pro offers near-comparable accuracy at a fraction of the cost; improvements rely on contextual retrieval tuned for long documents.
medium descriptive Full text arxiv
IPO Finance Agent: Evaluation of LLM Financial Analysts beyond Finance Agent v2, with Automated Rubric Generation -- the Case of the SpaceX (SPCX) IPO · Mostapha Benhenda · Jun 22, 2026
AI tools narrow the wage gap between junior and senior freelancers in coding, technical writing and design, compressing entry-level skill premiums and boosting the importance of supervision; simulations imply firms will need to redesign promotion paths as human intuition becomes more substitutable with algorithmic output.
medium quasi_experimental Full text openalex
THE ASYMMETRIC IMPACT OF GENERATIVE ARTIFICIAL INTELLIGENCE ON THE RETURNS TO COGNITIVE LABOR: A QUASI-EXPERIMENTAL STUDY OF HIGH-SKILLED FREELANCE MARKETS · Anika Singh · Jun 22, 2026
Advances in AI are frequently throttled by human capital and interfaces rather than model capabilities; organizations must redesign interfaces, retraining, and workflows to lift the practical performance ceiling. The paper proposes a socio-technical evaluation model and workforce-development principles to treat humans as core system components whose skills and processes shape AI productivity.
n/a theoretical Full text openalex
Optimizing Human Capital in AI-Enabled Architectures: A Systems Constraint and Capability Analysis · Jeremy A Schlegel, Jennifer Daffinee · Jun 22, 2026
Pooled data suggest co-authored AI pull requests merge less often, but that's an illusion: stratifying by agent and adding repo and PR-structure controls erases the apparent co-authorship advantage, indicating selection and structural confounders rather than a causal productivity gain.
medium correlational Full text arxiv
Beyond Simpson's Paradox: A Cascade of Confounders in AI Agent Pull-Request Co-Authorship · Haoran Yu, Xiaochong Jiang, Lifei Liu, Su Wang, Pin Qian, Yihang Chen · Jun 21, 2026
Long chains of AI actions fail exponentially in non-deterministic environments, threatening claims that compute scaling alone will unlock ASI-level economic performance; the authors propose a Supply Certainty Index and Determinism Maturity Model to measure and stage the environmental constraints that limit chainable automation.
n/a theoretical Full text arxiv
Grounded Scaling: Why Agentic AI Needs Deterministic Environments · Liang Ding, Xintong Wang · Jun 21, 2026
A three-tier oversight system for agentic code generation can preserve about 91% of coding speed while producing the compliance artifacts required by major regulators; the GAIE framework maps task-level oversight to regulatory impact to balance productivity and auditability.
low theoretical Full text arxiv
Governed AI-Assisted Engineering: Graduated Human Oversight for Agentic Code Generation in Regulated Domains · Richard Kang · Jun 21, 2026
Contrary to alarmist accounts, automation has often increased employment: highly automated countries in the reviewed literature show rising numbers and shares of workers, driven by productivity gains, higher real incomes and the creation of complementary and new jobs.
medium review_meta Summary openalex
New Technologies and Increase in Employment · Mehmet Arif Koşar · Jun 21, 2026
A misaligned AI can at most boost a human's correct-bit guesses by 50% relative to relying on the prior, and the gain shrinks for nearly independent information — yet a constructed six-bit example shows gains can exceed 25%, ruling out a universal 5/4 cap.
n/a theoretical Full text arxiv
Quantifying Theoretical AI Alignment Guarantees: Receiver-Utility Bounds in Bayesian Persuasion · Eric Yachbes, Eva Tardos · Jun 20, 2026
AI recommendations shorten resume review and blunt the benefits of extra scrutiny: without recommendations, extra viewing raises selection odds by about 3–4% and reviewers spend up to 55.6% longer inspecting resumes; completing an implicit-bias test beforehand reduces racial disparities in review time and predicts how people engage with AI screening.
medium rct Full text arxiv
Resume Screening, Fast and Slow: (Biased) AI Recommendations' Influence on Human Decision Making · Kyra Wilson, Mattea Sim, Anna-Maria Gueorguieva, Soham Chatterjee, Aylin Caliska... · Jun 20, 2026
AI improvements boost returns to AI capital, shift investment toward AI and generate a front-loaded output boom that slightly raises fertility; by contrast, longer lifespans increase saving, lower interest rates and fertility, and account for most macro volatility.
n/a theoretical Full text arxiv
Automation and Aging in General Equilibrium: AI Capital, Fertility, and the Return to Capital · James Wabenga Yango · Jun 20, 2026
After ChatGPT's release, clients on Upwork put less weight on workers' human-capital signals and more weight on price in AI-exposed job categories, with demand shifting toward lower-priced freelancers—evidence consistent with a commoditization effect of generative AI.
medium quasi_experimental Full text arxiv
Human Capital, AI, and Labor Commoditization · Auyon Siddiq, Niuniu Zhang · Jun 20, 2026
Firms pursuing human‑centric AI experience lower firm‑specific stock volatility in China, though the effect varies: stronger where digitalisation and executive ownership are higher, and weaker where operational efficiency and CEOs with IT backgrounds are greater.
medium correlational Full text openalex
Exploring the Relationship Between Human-Centric AI and firm Idiosyncratic Risks · Zhen-yuan Ralph Liu, Yu‐Ting Wang, Jia-jia Yan, Shivam Gupta, Mihalis Giannakis · Jun 20, 2026