AssembleAI
Back to home

The Evidence

Quantitative comparisons and the decision science research that underpins our approach.

01Comparative Analysis

Measurably better

Per-response comparison: “How will the appointment of Kevin Warsh as FED Chair change the composition of the FED's balance sheet?”

AssembleAI Opus 4.6 GPT 5.2 Gemini 3-Pro
Reasoning Paths
5
1
3
2
Actionable Insights
12
0
0
0
Lexical Density %
68
45
58
52
Technical Entities
34
9
19
14
Logic Steps Used
12
5
9
6
02Decision Science

Grounded in research

Our approach is built on decades of research into collective intelligence, structured disagreement, and the aggregation of diverse expertise.

Structured aggregation of diverse opinions reduces decision error by ~49% vs individual or crowd averages.

Navajas et al., 2018·Nature Human Behaviour

Aggregating intuitive and analytical judgments improves estimation accuracy beyond homogeneous judgments.

Keck & Tang, 2020·Journal of Experimental Psychology

Cognitively diverse groups consistently outperform homogeneous ‘best-expert’ groups in complex decision spaces.

Hong & Page, 2004·PNAS

Disagreeing perspectives boost aggregate accuracy on difficult tasks.

Inner-crowd Research, 2025·Cognitive Science

~49%

reduction in decision error through structured aggregation of diverse opinions

Navajas et al., 2018 · Nature Human Behaviour