The Commonplace
Home Dashboard Papers Evidence Digests 🎲

Papers (627)

Filter: Sort:
Safer Builders, Risky Maintainers: A Comparative Study of Breaking Changes in Human vs Agentic PRs
arxiv
Safer Builders, Risky Maintainers: A Comparative Study of Breaking Changes in Human vs Agentic PRs · K M Ferdous, Dipayan Banik, Kowshik Chowdhury, Shazibul Islam Shamim · Mar 29, 2026
AI is likely to rewire how jobs are created — moving from employer-dominated hiring to a decentralized ecosystem where individuals leverage AI and platforms to generate work; policy and education should pivot toward cultivating human–AI collaboration and creative skills.
low theoretical arxiv
AI Civilization and the Transformation of Work · Dongsoo Han · Mar 29, 2026
Companies that treat AI as a strategic operating layer — deploying agentic workflows and localized models — can accelerate innovation and expand internationally without traditional physical footprints. This advantage, however, only materializes when firms transition from pilot projects to integrated, governed AI‑first business models.
n/a descriptive openalex
The AI Advantage: Strategic Innovation and Global Expansion in the Digital Era · Yashaswini Bergal, Rupali Dilip Taru · Mar 29, 2026
A 'novelty bottleneck' limits how much AI can shrink human effort: when any fraction of a task requires genuinely novel human judgment, that serial component forces human effort to scale linearly with task size, so stronger agents cut costs but do not change the fundamental scaling. Consequently, organizations can shorten wall-clock time via parallel teams and need fewer humans as agents improve, but aggregate human effort remains bounded—making AI more effective at exploiting existing knowledge than at accelerating frontier research.
n/a theoretical arxiv
The Novelty Bottleneck: A Framework for Understanding Human Effort Scaling in AI-Assisted Work · Jacky Liang · Mar 28, 2026
EpochX proposes treating humans and AI agents as peers in a credits-native marketplace that turns every completed task into reusable assets—skills, workflows, and execution traces—so work can be delegated, verified and monetized at scale. By coupling an explicit asset dependency graph with a credit settlement mechanism, the design aims to make durable human–agent collaboration economically viable despite real compute costs.
n/a theoretical arxiv
EpochX: Building the Infrastructure for an Emergent Agent Civilization · Huacan Wang, Chaofa Yuan, Xialie Zhuang, Tu Hu, Shuo Zhang, Jun Han, Shi Wei, Da... · Mar 28, 2026
An AI multi-agent system cuts TensorFlow-to-JAX migration time roughly 6–8x on commercial Google workloads by combining planners, orchestrators and AI judges; the approach speeds model framework transitions but rests on proprietary tooling and case selection.
medium descriptive arxiv
A Multi-agent AI System for Deep Learning Model Migration from TensorFlow to JAX · Stoyan Nikolov, Bernhard Konrad, Moritz Gronbach, Niket Kumar, Ann Yan, Varun Si... · Mar 28, 2026
AI delivers bigger growth payoffs where institutions adapt: countries with stronger institutions, higher digital capacity and R&D intensity see superior productivity performance, suggesting AI's benefits depend on institutional and organizational innovation.
low theoretical semantic_scholar
Artificial intelligence, institutional innovation and economic growth: A conceptual framework · Zehra Doğan Çalışkan · Fetched Mar 28, 2026
AI-powered, multi-source horizon scanning classifies technologies into emerging, developing and mature stages and matches each stage to tailored policy tools—R&D funding, regulatory sandboxes, procurement incentives and tax relief. Expert Delphi validation and AHP weighting increase prioritization confidence, though outcomes depend on data coverage and expert selection.
low descriptive semantic_scholar
Emerging Technologies Based on Large AI Models and the Design of Support Policies: Patent Mining and Industry Trends · Yiwei Zhang, Shanghai Mu, Baitong Liu · Fetched Mar 28, 2026
AI sharpens ESG scoring and enables real‑time tracking of environmental and social risks, but gains are uneven: poor data, algorithmic bias and weak governance hamper reliable, trustworthy sustainable finance.
medium review_meta openalex
Artificial intelligence in sustainable finance and Environmental, Social, and Governance (ESG) performance: A review · Vibhavari Prasad Mane · Mar 28, 2026
Closed AI models often undermine scientific inference by hiding construction and deployment details, making results hard to interpret, replicate, or generalize; researchers should explicitly document threats to inference, the mitigation steps taken, and why a given (closed or open) model was chosen.
n/a theoretical arxiv
How Open Must Language Models be to Enable Reliable Scientific Inference? · James A. Michaelov, Catherine Arnett, Tyler A. Chang, Pamela D. Rivière, Samuel ... · Mar 27, 2026
An ASR system trained on field recordings of Ikema achieves around 15% character error rate and materially speeds up human transcription, cutting labor time and reducing cognitive load; this suggests practical AI-assisted pathways for scaling documentation of severely endangered languages.
medium quasi_experimental arxiv
Automatic Speech Recognition for Documenting Endangered Languages: Case Study of Ikema Miyakoan · Chihiro Taguchi, Yukinori Takubo, David Chiang · Mar 27, 2026
AI is reshaping Cambodia's labour market: low-skilled workers face disproportionate displacement while new industries create employment and spike demand for reskilling; policymakers must invest in targeted training and education.
low correlational semantic_scholar
Navigating AI‐Induced Job Displacement and Skill Demands: Insights From an Emerging Market Perspective · Bora Ly, Romny Ly, Sokhom Ma · Fetched Mar 27, 2026
AI code reviewers currently catch only a fraction of issues human experts find — top models detect 15–31% of flagged problems — and giving them more file-level context often makes performance worse, suggesting attention dilution in long prompts.
high descriptive arxiv
SWE-PRBench: Benchmarking AI Code Review Quality Against Pull Request Feedback · Deepak Kumar · Mar 27, 2026
Generative AI boosts customer-service throughput and agent productivity, especially by automating routine queries and drafting replies, but quality failures and trust risks mean it complements rather than replaces humans; hybrid human-AI systems with strong governance produce the best results.
medium review_meta openalex
The Effectiveness of ChatGPT in Customer Service and Communication Management (Nano Review) · Horn Sarun · Mar 27, 2026
Generative AI widens the space of feasible ideas and speeds early-stage R&D, but gains crystallize only when humans vet, curate and implement outputs; unchecked use risks low-quality, biased, or risky innovations.
medium review_meta openalex
ChatGPT as an Innovative Tool for Idea Generation and Problem-Solving: Nano Review · Horn Sarun · Mar 27, 2026
Generative AI floods teams with ideas and structures early-stage problem solving but seldom substitutes for human judgment; firms that combine high-volume generation with expert curation will capture most of the economic upside, while policy should prioritize training and accountability.
medium review_meta openalex
ChatGPT as an Innovative Tool for Idea Generation and Problem-Solving: Nano Review · Horn Sarun · Mar 27, 2026
Generative AI can cut costs and scale personalized customer service, but hallucinations, empathy shortfalls and integration hurdles make full automation unwise; firms capture the most value through hybrid human–AI systems with transparent AI use, clear escalation paths and continuous oversight.
low review_meta openalex
The Effectiveness of ChatGPT in Customer Service and Communication Management (Nano Review) · Horn Sarun · Mar 27, 2026
Behavioral frictions—miscalibrated trust, cognitive overload, and weak governance—help explain why many corporate AI pilots never reach production, with coding assistants a high‑visibility example. The paper offers a practical diagnostic framework for managers to detect disengagement and guide institutionalization, alongside a research agenda for mixed‑methods validation.
n/a review_meta openalex
Behavioral Factors as Determinants of Successful Scaling of Artificial Intelligence Pilot Projects in a Corporate Environment · Aneesha Sharma · Mar 27, 2026
China's AI collaboration network grows through triangle-closing and 'rich-get-richer' dynamics and is driven more by universities and research institutes than by firms; notably, organizations seek complementary technological expertise rather than similar tech, while geographic, cultural and institutional closeness still helps form ties.
medium correlational openalex
The evolutionary mechanism of artificial intelligence industry collaboration networks: evidence from China · Jianlin Lyu, Bin Hu, Wenrong Lyu · Mar 27, 2026
Countries that boost AI readiness tend to see measurable gains in e-government: within-country increases in the Government AI Readiness Index are linked to higher E‑Government Development, and the post‑pandemic period shows a notable acceleration in digital public administration—though national trajectories differ sharply.
medium quasi_experimental openalex
E-government development: Artificial intelligence vibrancy and readiness as drivers of digital public administration · Sergiy Spivakovskyy, Narek M. Kesoyan, Olena Igorivna Astapova-Vyazmina, Anatoli... · Mar 27, 2026
Challenges of indigenous employment in the private logistics sector in developing economies: an in-depth study of the Sultanate of Oman
openalex
Challenges of indigenous employment in the private logistics sector in developing economies: an in-depth study of the Sultanate of Oman · Hamad Zahar Hamad Al Abri · Mar 27, 2026
Saving the Common Law of Contracts from Politics and Codification: Holmes, Langdell, and Legal Science
openalex
Saving the Common Law of Contracts from Politics and Codification: Holmes, Langdell, and Legal Science · Daniel P. O’Gorman · Mar 27, 2026
Smarter search can hurt consumers: more informative AI-assisted search may reduce learning and consumer surplus and soften price competition unless markets themselves learn from agent-consumer interactions; by contrast, cheaper search unambiguously benefits consumers.
n/a theoretical arxiv
Agentic Markets: Equilibrium Effects of Improving Consumer Search · Brendan Lucier, Nicole Immorlica, Markus Mobius, Aleksandrs Slivkins, Daniel G. ... · Mar 26, 2026
A new 67.5-hour benchmark finds current multimodal models struggle to infer users' GUI behavior and intent — scoring about 44.6% on behavior-state detection and 55.0% on help prediction — yet supplying structured user context can raise help-prediction accuracy by up to 50.2 percentage points.
n/a descriptive arxiv
GUIDE: A Benchmark for Understanding and Assisting Users in Open-Ended GUI Tasks · Saelyne Yang, Jaesang Yu, Yi-Hao Peng, Kevin Qinghong Lin, Jae Won Cho, Yale Son... · Mar 26, 2026
Multiple general-purpose coding agents can dramatically accelerate hardware designs: a two-stage agent factory using up to 10 Claude Code agents achieves a mean 8.27× HLS speedup across 12 benchmarks, with peaks above 20×, and finds cross-kernel optimizations missed by per-subkernel search.
medium other arxiv
Agent Factories for High Level Synthesis: How Far Can General-Purpose Coding Agents Go in Hardware Optimization? · Abhishek Bhandwaldar, Mihir Choudhury, Ruchir Puri, Akash Srivastava · Mar 26, 2026