Evidence (7560 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
Filter claims →
Governance
7870 claims
Filter claims →
Human-AI Collaboration
7560 claims
Filtered →
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 |
Human Ai Collab
Remove filter
There is a ceiling effect where excessive linguistic expansion yields diminishing marginal utility.
Empirical observation reported in the abstract that overly expanding linguistic output leads to diminishing marginal gains; presumably derived from analysis of the dataset and evaluation framework.
Achieving this system-level transformation takes time: it requires trust and accountability infrastructure, machine-legible and interoperable data and interfaces, the design and adoption of these new workflows, and economic incentives that favor reconstruction rather than local optimization.
Argumentative claim listing necessary preconditions and complementary investments; presented conceptually without reported empirical measurement in the provided text.
The main reason [the disruption has not fully arrived] is not model capability, nor even the tools built to harness those models; rather, most organizations are still using AI to accelerate workflows designed for a pre-AI world.
Argued in the paper as an explanatory thesis; supported by conceptual argument and illustrative examples (consumer markets, education, news, coding) rather than reported empirical analysis in the provided text.
The disruption many expect has not fully arrived.
Stated as an observation in the paper's introduction/abstract; no empirical method, sample size, or data reported in the excerpt.
Cafeteria demand planning requires both algorithmic pattern recognition and human expertise, yet current systems treat these separately, which generates significant food waste.
Statement in the paper's motivation/background; presumably grounded in literature review and problem framing rather than new empirical measurement in this study.
Reputation mechanisms presuppose persistent identity, behavioral continuity, sanction sensitivity, and costly non-fungibility; absence of any of these undermines reputation systems.
Analytic claim in the paper articulating necessary conditions for reputation mechanisms to function; presented as theoretical grounding rather than empirically tested criteria.
The analogy from human identity verification and reputation mechanisms (e.g., 'Know Your Customer', credit scores) to 'Know Your Agent' regimes is fundamentally incomplete.
Comparative conceptual argument in the paper highlighting disanalogies between human actors and modular language model agents; no empirical comparison or data provided.
Identity-based, ex post, regulative, sanction-based governance, such as reputation, is structurally inapplicable to dissociative agents.
Normative/theoretical argument in the paper deduced from properties of dissociative agents and requirements of identity-based governance; no empirical or experimental support reported.
Dissociativity leaves agents without grounding for identifiability, predictability, credibility, and rehabilitability — the very properties that reputation mechanisms aim to sustain — thereby collapsing trust.
Conceptual inference in the paper combining the dissociative characterization of agents with the functional requirements of reputation systems; no empirical validation provided.
An agent's persona is fluid, vulnerable to adversarial attack, and may not internalize sanctions.
Argumentative claim in the paper citing susceptibility of modular agent components (prompts, tools, memory) to manipulation; no empirical attack experiments or sample sizes reported.
Language model agents are ontologically dissociative: they are essentially an assemblage of mutable modules -- foundational models, system prompts, tool-access policies, external memory, and, in some cases, a multi-agent system as a whole -- any of which may change agent behavior.
Theoretical characterization and system-level description in the paper; lists components that can be changed to alter behavior; no empirical measurement or sample reported.
Human-only teams commit more errors than mixed human–AI teams.
Reported counts/observations of errors made by team type in the escape room experiment; the abstract does not include numerical error counts or significance levels.
Human-only teams take longer to complete the escape room than mixed human–AI teams.
Reported comparison of time-to-complete between human-only and mixed teams in the experiment; specific times or statistical tests are not provided in the abstract.
The share of diffs receiving timely review has declined, exposing a widening gap between code supply and reviewer bandwidth.
Observational telemetry/operational metrics reported in the paper indicating a decline in timely reviews relative to diff supply. No specific numeric sample size provided in the excerpt.
GPT models showed significantly larger discrepancies compared to other evaluated models.
Comparative evaluation reported in the paper indicating GPT-family models had larger errors/discrepancies relative to the best-performing models.
Employees often struggle to identify "who knows what," leading to organizational productivity losses.
Motivating statement in the paper (not an empirical result from this study); general observation cited as motivation for the research.
Experiments on closed-source and open-source LLMs reveal a clear failure cascade from executable code to valid geometry and finally to engineering-ready design, with even the strongest models achieving limited success on fine-grained engineering criteria.
Experimental results described in abstract comparing multiple LLMs across the three evaluation stages.
Existing benchmarks focus primarily on generating single-part CAD models and evaluate them using geometric similarity metrics that fail to capture functionality, manufacturability, and assemblability.
Paper's literature/related-work claim as stated in abstract; critique of existing benchmarks' focus and metrics.
Results across 15 experimental runs reveal that elderly female occupants consistently experience the lowest satisfaction in initial rounds.
Empirical experiment results reported in abstract: 15 experimental runs; observed satisfaction distribution across demographic profiles with elderly females lowest initially.
Existing AI education, AI literacy, and human-AI collaboration frameworks remain centred on prompting, task execution, and productivity support and are poorly equipped to address this tacit layer of expert cognition.
Argumentative critique in the paper drawing on conceptual analysis and review of prevailing frameworks; no empirical evaluation or sample reported.
Both humans and AI contribute wrong answers.
Reported error contributions from both human participants and AI agents in the experimental task.
Humans over-rely on AI when AI misleads them, occurring in 1.7% of opportunities.
Aggregate analysis of adoption decisions in the experiment (reported percentage of over-reliance on misleading AI suggestions).
Humans under-rely on correct AI suggestions, missing 3.9% of opportunities.
Aggregate analysis of adoption decisions in the experiment (reported percentage of missed opportunities to rely on correct AI suggestions).
The core cause of the R&D productivity paradox is cognitive saturation: researchers spend an increasing share of their effort on coordination, documentation, and data governance—hidden work that displaces high-value hypothesis formation, interpretation, and strategic synthesis.
Argument presented in the paper supported by DSR analysis, triangulated with four expert interviews, foresight scenarios, and pattern matching (causal claim based on qualitative evidence and reasoning).
Corporate R&D faces a persistent productivity paradox: rising investment and expanding scientific knowledge have not translated into proportional innovation output (Eroom's Law); analogous patterns appear across engineering, materials science, and healthcare.
Literature reference to Eroom's Law and cross-domain pattern matching described in the paper (conceptual/literature observation).
Incumbent workforce management frameworks remain anchored to a purely human labor model, rendering AI agents invisible to capacity planning, performance attribution, and governance enforcement.
Stated in paper's problem motivation / literature review; presented as an observed gap motivating the design-science contribution. No sample size or empirical study described in the provided text.
Organizations increasingly deploy separate purpose-built AI tools across professional domains, often hiring domain specialists for each, recreating the staffing models AI was expected to transform.
Stated as an observational/introductory claim in the paper (no empirical data or sample size reported to support the general trend).
Such settings abstract away two characteristics of real industrial OR workflows: persistent multi-artifact workspaces and multi-stage task lifecycles.
Argument in the paper identifying gaps in benchmark realism (conceptual claim based on comparison between benchmark setups and industrial workflows).
Existing OR-oriented benchmarks often reduce evaluation to one-shot translation from a self-contained problem statement into a mathematical formulation or solver program.
Claim about the prevailing design of prior benchmarks as stated in the paper (literature survey / critical summary in the text excerpt; no numeric evidence provided here).
The greatest organizational risk of AI may not be technical failure but structural over-optimization (i.e., AI-driven erosion of adaptive openness).
Argumentative claim derived from the AI fragility theory presented in the paper; no empirical validation or quantified risk assessment included.
Artificial intelligence functions as a 'hyper-crystallization' engine—by classifying, predicting, standardizing and optimizing it accelerates structural crystallization and may erode local judgment and generative adaptability.
Conceptual theory labeled 'AI fragility theory' developed in the paper; supported by argumentative reasoning rather than empirical testing.
When digital systems are reified into internal structural optimization and control, transformation efforts can intensify organizational rigidity and failure to adapt.
Theoretical/analytic argument contrasting two modes of digital transformation; no empirical estimates or dataset provided.
Structurally heavy firms with substantial material and institutional resources frequently experienced paralysis or collapse during the pandemic.
Qualitative claim grounded in the author's reading of pandemic outcomes; the paper does not report systematic data or case counts.
During the COVID-19 pandemic, firms with the most optimized structures were not necessarily the most adaptive under radical uncertainty.
Argument based on the COVID-19 pandemic presented as an empirical 'stress test' in the paper; no empirical sample, data, or statistical analysis provided.
AI may influence society broadly via ethical issues, economic inequality, and social adaptation challenges.
Paper lists ethics, economic inequality, and social adaptation as societal-level areas affected by AI (abstract). Presented as thematic concerns reviewed in the paper; no empirical estimates included in the provided text.
AI-driven automation is associated with job loss.
The paper lists automation and job loss among the areas it examines (abstract). The provided text frames job loss as a potential negative ramification but does not report primary empirical estimates or sample sizes.
LLMs heavily rely on simulations for designing algorithms, which is notorious for breaking when transferred to real hardware.
Paper's claim grounded in known transferability issues between simulation and hardware; no experimental quantification provided in the abstract.
LLM pitfalls worsen on Radio Access Network (RAN) use cases: they hallucinate Application Programming Interfaces (APIs) and mis-read specifications, which kills interoperability of RAN components at the first mistake.
Author assertion / observed behavior reported in the paper (qualitative examples implied); no formal experiment or sample size provided in the abstract.
Cellular research and development (R&D) is throttled by six structural processes that each consume months of manual engineering work per iteration: (i) synthesizing new features from standards or research papers into production code; (ii) conformance and interoperability testing; (iii) hardening against field anomalies and diverse deployment environments; (iv) data-driven optimization of network functionalities; (v) discovering and prototyping novel waveforms, functionalities, and capabilities for future standards; and (vi) securing the stack against vulnerabilities.
Author assertion in the paper (qualitative analysis / domain expertise). No empirical sample size or quantitative study reported in the abstract.
Existing approaches remain fragmented across formal verification, runtime assurance, neuro-symbolic reasoning and trustworthy Artificial Intelligence (AI) research communities.
Author claim about the state of the research landscape; asserted fragmentation without bibliometric or survey data provided in excerpt.
Current reasoning systems still suffer from hidden logical inconsistencies, hallucinated symbolic transitions, unsupported theorem applications, and limited reliability guarantees.
Author assertion identifying failure modes of current reasoning systems; presented qualitatively without quantitative error rates or experimental sample sizes in the excerpt.
Stochastic Tax can remain positive even when Agentic Technical Debt is minimized.
Theoretical claim in the paper's model and discussion: even with minimized debt (stock), the model predicts a nonzero recurring operating burden from stochastic agents; illustrated via examples and an accounts-payable simulation.
Stochastic Tax is a recurring flow of operating burden that arises when stochastic agents are used in business workflows.
Definition provided in the paper as part of the conceptual framework describing Stochastic Tax as a flow (recurring operating burden) associated with stochastic agents in workflows.
Humans are the bottleneck in building and improving AI. Both the models and the agents that wrap them are written, tuned, and corrected by people.
Statement/argument in the paper's introduction framing motivation (conceptual observation about current practice). No experimental data reported in the abstract to support this claim.
Translators have functioned as 'invisible teachers' of AI—through the construction of translation memories, post-editing, and quality assessment—without recognition as teachers of models.
Conceptual framing and synthesis of workflow practices (TM construction, post-editing, QA) and their role as supervision for ML; qualitative argument and illustrative examples in the paper. No quantitative sample reported.
Translators' renditions have been bought as deliverables under contract, segmented as technical objects, and processed as 'information analysis' data under copyright law—resulting in the loss of moral, creative, and economic attribution to the translators who produced them.
Comparative reading of contract practices and copyright treatment (legal/contractual analysis across jurisdictions), descriptive examples of how translations are delivered, segmented, and processed; qualitative argumentation in the paper. No quantitative sample reported.
Using GoEmotions-based affective prompting, we show that emotion substantially shifts negotiation outcomes, suggesting that emotion is a strategic action channel rather than a surface style.
Experiments using GoEmotions-based affective prompting applied to negotiation agents; reported shift in negotiation outcomes attributed to emotional framing. (Specific experimental details such as number of runs, seeds, or exact metrics are not provided in the excerpt.)
Slow AI induced delayed cognitive conflict; humans hesitated (61.1% accuracy).
Reported experimental result for Slow/Accurate AI condition showing human accuracy of 61.1%, interpreted as hesitation/delayed conflict.
Pure behavioural teams (N=8) failed to scale beyond 74.1%.
Reported team performance metric for 'pure behavioural' teams with sample size N=8; maximum reported performance 74.1%.
Fast AI induced instant, blind compliance; human accuracy under deception collapsed to 50.2%.
Reported experimental result comparing Fast/Less-Accurate AI condition to baseline conditions; numeric accuracy reported as 50.2% for humans under deception.