Truace
Subscribe
The Index recomputed live from the record

What the evidence says.What the public feels.

The record holds 939 sourced gains and 772 sourced problems, averaging 68 and 66 on the index score. Readers have logged 22 public signals on the Pulse, which is kept apart and never counted as evidence.

939 gains
772 problems
Every sourced claim in the Index, one square each, shaded by the strength of its evidence. High Moderate Emerging

Ranks distinct AI gain and problem claims from the published record. Scores reward impact, independent source strength, scale, confidence, and recency.

1,711 results
Show filters and sorting

Download every matching row, not just this page:Export CSVExport JSON

AI gains · 939

509
PolicyStableModerate evidence · 1 source

A multi-stage framework integrating YOLOv8 detection, OC-SORT tracking, dynamic CROI filtering, and ST-GAT prediction enables accurate real-time traffic-conflict prediction at signalized intersections to enhance safety and mitigate accident risks.

Researchers built a multi-stage video-based framework for signalized intersections that combines YOLOv8 detection with OC-SORT tracking to extract vehicle trajectories, uses a dynamic scaling Conflict Region of Interest to reduce data volume, and predicts conflicts with a Spatio-Temporal Graph Attention Network followed by causal forest interpretation. Tested on field video from an intersection in Nanning, China, the ST-GAT model outperformed existing deep learning architectures on training and testing sets.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
89

Updated Aug 15, 2026 · TRV-2026-0767

67Index score
510
HealthStableModerate evidence · 1 source

AI simplification of dermatopathology reports for patients was rated by dermatology professionals as mostly factual, complete, and harmless.

A peer-reviewed survey study from January to April 2025 asked 52 US dermatology and dermatopathology professionals to rate AI-simplified versions of six fictitious dermatopathology reports. One version used Basic ChatGPT-4.0 with a simple prompt and the other used a custom DermDecoder GPT with a structured 489-word prompt, evaluated for factualness, completeness, and potential harm.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
89

Updated Aug 15, 2026 · TRV-2026-0766

67Index score
511
HealthStableModerate evidence · 1 source

Automated AI segmentation of macular OCT quantified selective inner retinal thinning during silicone oil tamponade and identified RNFL, GCL+IPL and PR+RPE as strongest predictors of visual acuity change, enabling prognostication after oil removal.

In 76 eyes treated with silicone oil endotamponade for rhegmatogenous retinal detachment, researchers used an automated OCT segmentation tool and a random forest classifier to track retinal layer changes between oil insertion and removal and to predict categorical best-corrected visual acuity change.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
89

Updated Aug 15, 2026 · TRV-2026-0765

67Index score
512
HealthStableModerate evidence · 1 source

AI assistance during routine mammography was associated with higher overall cancer detection and no increase in average interpretation time.

Between August 2023 and July 2024, four radiologists interpreted 4577 screening and diagnostic mammograms in a prospective alternating-month design where a commercial AI system was shown or hidden. Reading times from PACS logs, cancer detection rates, and abnormal interpretation rates were compared between AI-assisted and non-AI-assisted months, with reading time analysis restricted to 2917 cases under five minutes.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
89

Updated Aug 15, 2026 · TRV-2026-0764

67Index score

AI problems · 772

509
PolicyStableModerate evidence · 1 source

Existing AI development remains monocultural and top-down, with gaps in inclusion and persistent power asymmetries and epistemic justice concerns.

On 2026-07-02, a peer-reviewed paper in AI & SOCIETY introduced Value-Sensitive Citizen Science (VSCS), a framework that combines Value-Sensitive Design with citizen science to involve community members as co-researchers in AI development. It uses the Participatory Value-Cognition Taxonomy and extended scenario reasoning to translate local values into technical requirements and embeds governance for lifecycle oversight.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
83

Updated Jul 17, 2026 · TRV-2026-0253

67Index score
510
ScienceStableModerate evidence · 1 source

Machine learning models for biomolecular recognition do not inherently enforce thermodynamic consistency and may produce configurations that are not physically realizable.

Published July 17, 2026, this perspective argues that quantitative prediction of biomolecular recognition requires moving beyond static structures to ensemble-based thermodynamic and kinetic observables. It reviews physics-based sampling under approximate Hamiltonians and modern machine learning models that learn from structural and bioactivity data.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
83

Updated Jul 17, 2026 · TRV-2026-0244

67Index score
511
ScienceStableModerate evidence · 1 source

MLFFs still face challenges in computational efficiency and scale, accuracy and generalization, data requirements, interpretability, and physical constraints when applied to inorganic crystalline materials.

On 2026-07-17, a review in Physical Chemistry Chemical Physics summarized machine learning force fields for inorganic crystalline materials, describing how they combine first-principles accuracy with classical force-field efficiency to enable atomic-level studies across structural prediction, physical properties, defects and interfaces, and phase transitions.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
83

Updated Jul 17, 2026 · TRV-2026-0243

67Index score
512
ScienceStableModerate evidence · 1 source

Existing literature on AI-human decision-making was fragmented and lacked integrative frameworks to explain how AI-human dynamics and decision typologies shape outcomes despite increasingly intricate human-AI interplay.

On 2026-04-03, a peer-reviewed article reported a systematic review and bibliometric analysis of 627 articles on human-AI decision-making. The authors identified two critical dimensions, AI-human dynamics and decision typologies, and proposed a novel conceptual framework comprising four paradigms: adaptive intuitive, programmed algorithmic, interpretive analytical, and integrative hybrid decision-making.

Impact 30%
49
Evidence 25%
95
Scale 20%
35
Confidence 15%
87
Recency 10%
83

Updated Jul 15, 2026 · TRV-2026-0226

67Index score

Recomputed live from the record · Oct 13, 2026, 3:55 PM