Evidence (8807 claims)
Search and filter individual claims pulled from the papers. Looking for a specific finding ("what's the effect on wages?"), you're in the right place. Want to compare whole outcome categories against each other instead? Use the Evidence Explorer.
The board below groups claims two ways: by broad theme (nine paper-level topics) and by outcome category (the 34 claim-level outcomes that the Explorer and Syntheses also use).
Browse by theme
Nine broad, paper-level topics. Click one to filter the claims below.
Adoption
9875 claims
Filter claims →
Productivity
8807 claims
Filtered →
Governance
7870 claims
Filter claims →
Human-AI Collaboration
7560 claims
Filter claims →
Org Design
4892 claims
Filter claims →
Innovation
4781 claims
Filter claims →
Labor Markets
4004 claims
Filter claims →
Skills & Training
3308 claims
Filter claims →
Inequality
2332 claims
Filter claims →
Claims by outcome category
Counts by direction of finding. These are the same 34 outcome categories the Explorer compares and the Syntheses are written for. A linked row has a published synthesis.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 870 | 233 | 116 | 1066 | 2363 |
| Governance & Regulation | 976 | 451 | 218 | 133 | 1809 |
| Organizational Efficiency | 949 | 224 | 144 | 88 | 1416 |
| Technology Adoption Rate | 764 | 287 | 141 | 122 | 1325 |
| Research Productivity | 501 | 152 | 74 | 362 | 1101 |
| Output Quality | 542 | 216 | 69 | 69 | 896 |
| Decision Quality | 387 | 198 | 94 | 54 | 740 |
| Firm Productivity | 513 | 67 | 101 | 27 | 714 |
| AI Safety & Ethics | 249 | 303 | 73 | 36 | 667 |
| Market Structure | 190 | 192 | 134 | 27 | 548 |
| Task Allocation | 243 | 77 | 91 | 36 | 452 |
| Innovation Output | 291 | 33 | 55 | 20 | 401 |
| Skill Acquisition | 206 | 72 | 65 | 21 | 364 |
| Employment Level | 133 | 63 | 115 | 22 | 335 |
| Fiscal & Macroeconomic | 153 | 79 | 52 | 32 | 323 |
| Task Completion Time | 206 | 37 | 12 | 15 | 272 |
| Firm Revenue | 179 | 52 | 29 | 5 | 266 |
| Consumer Welfare | 130 | 76 | 47 | 13 | 266 |
| Inequality Measures | 48 | 137 | 51 | 6 | 242 |
| Worker Satisfaction | 101 | 81 | 25 | 13 | 220 |
| Error Rate | 84 | 110 | 11 | 5 | 210 |
| Wages & Compensation | 98 | 47 | 30 | 10 | 185 |
| Regulatory Compliance | 88 | 73 | 17 | 7 | 185 |
| Automation Exposure | 66 | 64 | 33 | 16 | 182 |
| Team Performance | 105 | 29 | 30 | 11 | 176 |
| Training Effectiveness | 109 | 22 | 14 | 21 | 168 |
| Developer Productivity | 114 | 21 | 14 | 8 | 158 |
| Job Displacement | 12 | 90 | 24 | 1 | 127 |
| Hiring & Recruitment | 57 | 9 | 9 | 5 | 80 |
| Skill Obsolescence | 6 | 56 | 9 | 1 | 72 |
| Social Protection | 43 | 17 | 8 | 2 | 70 |
| Creative Output | 35 | 21 | 9 | 4 | 70 |
| Labor Share of Income | 18 | 21 | 17 | 1 | 57 |
| Worker Turnover | 15 | 16 | — | 4 | 35 |
| Industry | — | — | — | 1 | 1 |
Productivity
Remove filter
With endogenous capital accumulation, data-driven automation generates explosive growth but stagnant long-run wages.
Extended model incorporating endogenous capital accumulation: analytical solution/characterization showing unbounded (explosive) growth in aggregate variables while real wages remain stagnant in the long run (model derivation).
Along the transition path of automation, data simultaneously augments the productivity of already-automated tasks and expands the automation frontier (dual role).
Analytical results from the dynamic model showing two mechanisms: (i) data increases productivity of tasks already automated; and (ii) data enables automation of additional tasks (model derivations).
Defining query difficulty is one of the hardest problems in deployment engineering.
Statement/assertion in the paper (introductory claim); no specific empirical measurement in the abstract.
There were no significant differences in AI use based on most accountant characteristics, except in auditing where business owners reported a higher frequency of AI use.
Inferential statistical analysis of questionnaire data (comparative design); specific statistical tests and sample size not reported in the summary.
Frontier proprietary models achieve near-zero success under GUI-based interaction, whereas COM-based execution yields substantial immediate gains.
Experimental comparison reported in the paper on ComCADBench between GUI-based interaction by proprietary models and COM-based execution (authors report success rates and comparative performance).
Environment engineering can amplify productive behaviors (e.g., open-ended exploration, systematic artifact management, inter-agent collaboration) while suppressing harmful behaviors (e.g., reward hacking and high-friction human oversight).
Framing and argument in the paper describing expected effects of environment design (conceptual; no quantification provided in the excerpt).
Architectural smell density (ASD) declines by 6.7% (p = 0.004), but this decline is a denominator effect resulting from lines-of-code growth rather than an actual architectural improvement.
Observed ASD change computed from estimated smell counts and LOC changes in the 151-repository panel and interpreted by decomposing density into numerator (smells) and denominator (LOC).
The same observation is seen with the amount of changes (e.g., code churn, number of modified files) and with the efforts to merge an agentic PR (e.g., merge time and number of comments).
Reported that pre/post comparison across projects shows mixed/no consistent improvement patterns for code churn, modified files, merge time, and comment counts after instruction-file creation (analysis over 15,549 PRs in 148 projects).
Specifying instructions for AI-agents does not necessarily lead to better results.
Project-level before/after comparison of performance metrics for projects that created instruction files (pre/post comparison across 148 projects using the 15,549 PRs).
Du et al. (2026) find that information-based team faultlines can enhance proactive behavior via deep information processing, while AI adoption moderates and mitigates the negative effects of social-based faultlines on team cooperation.
Information-processing theoretical framing and empirical analysis reported in the paper (study type and sample size not specified in the excerpt).
Liao et al. (2026) identify multiple equifinal pathways to high performance in digit-oriented spin-offs (parent-oriented, independent-oriented, ambidextrous-oriented configurations) using fuzzy-set qualitative comparative analysis (fsQCA).
fsQCA analysis reported in the paper (methodological approach described; sample not specified in excerpt).
A store-level policy learned from logged marketplace data selects a discrete multiplier that shifts the dispatch optimizer's tradeoff between delivery quality and batching efficiency.
Methodological description: store-level policy trained from logged data that outputs a discrete multiplier to alter optimizer objective weights; stated design and training approach in paper (no numerical evaluation details provided in the excerpt).
Implementation success depends heavily on data quality, workflow redesign, interpretability, governance, and procurement alignment.
Synthesis of factors identified across included studies and supporting regulatory/industry documents as important determinants of successful deployment.
However, evidence is uneven: many studies are simulation-based.
Review observation from the synthesis of the 35 included studies noting study designs (simulation prevalence noted but not numerically specified).
The comparative evaluation shows differences in economic inclusiveness between ML, DL, and Generative AI.
Abstract states differences in economic inclusiveness found in the review; no quantitative inclusiveness metrics or sample sizes provided in abstract.
The comparative evaluation shows differences in explainability among ML, DL, and Generative AI.
Abstract notes comparative differences in explainability as part of review findings; no empirical measures of explainability included in abstract.
The comparative evaluation shows differences in patterns of substituting labor across ML, DL, and Generative AI.
Abstract states comparative differences in labor-substitution patterns based on the systematic review of literature; no empirical counts or sizes in abstract.
The comparative evaluation shows differences in scale of impact across ML, DL, and Generative AI.
Abstract reports a comparative evaluation highlighting scale differences across AI phases; no quantitative scale measures given in abstract.
Generative AI brings innovative disruption with profound effects on the structure of employment, knowledge-based ecosystems, and high-skill industries.
Synthesis claim in abstract based on reviewed peer‑reviewed literature; no specific studies, sample sizes, or quantitative effects reported in abstract.
Coding agents already know how to navigate files, edit code, run commands, and repair outputs, but lack the simulator's executable contract (vocabulary, structural constraints, validation rules, termination conditions).
Framing/assumption presented in the paper motivating the approach (not an empirical claim).
The research contrasts tool-shaping (AI behavior/prototype) and mind-shaping (user strategy training) pathways and reports differing effects between them.
Paper presents both a tool-shaping experiment (Study 1) and a mind-shaping experiment (Study 2) and discusses comparative findings across these pathways.
Cognitive flexibility is examined as a moderator (boundary condition) of the interventions' effects.
Paper reports including cognitive flexibility as an individual-differences moderator in analyses across the two studies (moderation analysis planned/reported).
Reasoning scaffolds (public tools, playbook, verifier, objectivity policy, red-team) improve calibration and audit discipline, but proprietary evidence sets the upper bound of what the AI Scientist can know and therefore decide.
Synthesis of experimental results showing B improved calibration/audit metrics while C (with proprietary data) markedly increased coverage and informed decision-quality.
Under capability-superset accounting on the curated gold competitive record, agent A recovers only 0.25, agent B recovers 0.38, while agent C recovers 0.96 (overall).
Capability-superset accounting comparison of fraction of a curated gold competitive record recovered by each agent on the benchmark.
Data contamination (training-data overlap) complicates interpretation of the models' performance.
Author notes the possibility that models' training data may have contained the target papers or related material, making results ambiguous.
Advanced economies have integrated AI technologies at scale, while emerging economies such as Algeria face structural and institutional challenges that limit the potential impact of AI on productivity growth.
Asserted in the paper with supporting literature citations (Agrawal et al., 2019; Acemoglu & Restrepo, 2020) and comparative use of World Bank and Oxford Insights indices; no specific sample-size based causal estimate provided.
There is significant cross-national, cross-industry, and cross-regional heterogeneity in AI's impact.
Conclusion from the systematic literature review indicating variation across countries, industries and regions in the effects reported by prior studies.
Research has shown that artificial intelligence is primarily driven by substitution effects in the short term, but will generate complementary and creative effects in the long term.
Synthesis claim from the literature review; the paper reports this as an aggregate finding from prior studies (no single-study sample size provided).
The paper analyzes the direct impact of artificial intelligence on employment structure, occupational tasks, and skill demand, as well as its indirect effects on job mobility, cross-border and industry differences, and policy interventions.
Descriptive claim of scope drawn from the systematic literature review conducted by the authors; no single empirical sample reported.
The rapid development of artificial intelligence is profoundly reshaping the global labor market landscape.
Statement in paper based on a systematic literature review synthesizing prior studies; no single empirical sample reported.
Analysis of recent benchmark evidence including SWE-bench Verified, EvoClaw, and LangChain's multi-agent coordination studies demonstrates both the transformative potential of the agentic paradigm and its current limitations.
Empirical/benchmark analysis referencing SWE-bench Verified, EvoClaw, and LangChain multi-agent studies as sources of evidence; the paper analyzes these benchmarks qualitatively or comparatively (specific sample sizes and quantitative effect sizes not stated in the abstract).
An explicit thinking mode raises rank-order correlation without moving accuracy.
Empirical comparison of reasoning modes showing increased rank-order correlation (e.g., Spearman/Fisher-z) when explicit 'thinking' mode is used, with no significant change in accuracy.
Most published twins are either coarse persona bots conditioned on a few demographic questions or detailed individual-level twins built on purpose-collected surveys and interview transcripts.
Author's literature summary / positioning statement in paper (qualitative assessment of existing published twins).
Overall, complementarity is attainable in multi-agent regression but obstructed in classification under natural conditions on local aggregation and loss functions.
Synthesis of the paper's proved positive results for regression and negative impossibility results for classification within the tree-based HAI framework (theoretical proofs; no empirical sample).
In regression under squared loss, complementarity is equivalent to Euclidean distance minimization from the ground-truth vector.
Analytic equivalence proved in the paper for the tree-based model under squared loss (mathematical derivation; no empirical sample).
AI advances science through structurally distinct creative pathways rather than a single mechanism; the creative pathway depends on how AI is incorporated into the research process.
Interpretation synthesized from observed heterogeneity in creativity outcomes across classified AI research modes (Tool-oriented vs Adaptation-oriented) in the >1M publication analysis.
There is a long-run equilibrium (cointegrating) relationship among AI adoption, skill-disaggregated unemployment, and sustainable development in South Africa.
Empirical ARDL results reported in the paper indicating a long-run equilibrium relationship based on annual 2003–2024 time-series data.
The study uses annual time-series data from 2003–2024 and the Autoregressive Distributed Lag (ARDL) modelling approach to estimate short- and long-run coefficients.
Explicit statement in the paper: annual time-series data 2003–2024 and ARDL modelling to simultaneously estimate short- and long-run coefficients.
AI will have social, economic, and political impacts on work, inequality, democracy and power.
Author's projection of the domains affected by AI (stated as a subject of later chapters; no empirical evidence provided in the excerpt).
The opportunities of AI in human good are real and vast; and the opportunities in human ill, in human society, in human institutions of government, and in the longer term in the environment in which humanity thrives are real and underestimated.
Author's evaluative judgment asserting both substantial benefits and substantial underestimated harms of AI (normative claim without empirical substantiation in the excerpt).
These behavioral differences have implications for deployment of agentic AI in scientific computing workflows, such as trade-offs between speed versus auditability, silent versus transparent error handling, instruction interpretation, and the criticality of intermediate data representations in multi-model pipelines.
Authors' discussion and interpretation based on observed experimental differences between the two agents across the runs.
The autonomously generated manuscripts also diverged in length, details, and quality.
Reported qualitative comparison of the LLM-assisted manuscripts produced by each agent indicating differences in length, level of detail, and overall quality between the two agents' outputs.
The agents exhibited substantially different behaviors and computational costs.
Overall observation from the two runs: distinct behavioral patterns (silent reinterpretation vs explicit restarts), different execution times, and differing computational actions (optimization introduced by Codex).
Claude Code completed the pipeline in ~3.4 minutes with silent deviations from the specification, while Codex required ~16 minutes across explicit self-correcting restarts, including an unsolicited performance optimization of the matched filter inner loop.
Reported run-time measurements and qualitative behavior descriptions in paper: timing values (~3.4 min vs ~16 min) and observed behaviors (silent deviations for Claude Code; explicit restarts and an unsolicited optimization by Codex).
The optimal architecture is highly task-dependent.
Empirical claim in the abstract: experiments across tasks showed that different hybrid architectures perform best for different tasks.
Task accuracy, monetary cost, and edge energy consumption are tightly coupled in hybrid MAS design.
Claim made in the abstract and investigated empirically by adapting MAS architectures and measuring power, cost, and performance trade-offs.
No single LLM dominates across engine types, highlighting the importance of specific tasks and tradeoffs between speed and accuracy.
Empirical observation from cross-engine evaluations reported in the paper; descriptive conclusion without numeric dominance metrics or sample sizes in the excerpt.
The evaluations implemented by the initiative demonstrate that AI enabled modeling tools perform better at discussion and basic qualitative tasks than with causal reasoning and quantitative error fixing.
Result reported from the implemented evaluations comparing relative performance across task categories (discussion/qualitative vs causal reasoning/quantitative error fixing); no quantitative effect sizes or sample sizes provided in the excerpt.
When engines from the sd ai project are coupled with different LLMs, their performance on these evaluations reveals variability across different AI tools.
Empirical statement in the paper based on applying the implemented evaluations to different engine+LLM combinations; no numeric performance metrics or sample sizes reported in the excerpt.
We illustrate this transition through examples in consumer markets, education, news, and coding.
Authors state they use sectoral examples to illustrate the framework; this is a claim about the paper's contents rather than an empirical finding.