Human Capital, Not Model Benchmarks, Predicts Hybrid Intelligence in Forecasting

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Vivienne Ming
• Venue: arXiv: cs.CY
• Published: 2026-07-02
• Collected: 2026-07-03
• URL: http://arxiv.org/abs/2607.02467v1
• PDF: https://arxiv.org/pdf/2607.02467v1
• DOI: UNKNOWN

Plain-Language Brief

A study using real-money prediction markets found that the effectiveness of human-AI collaboration in forecasting depends on specific forms of human capital (like perspective-taking, intellectual humility, and curiosity). Most individuals either deferred to the model or used it to confirm their own guesses, but a few engaged in complementary reasoning that matched or even surpassed the market's accuracy.

What the Document Says

• The study compares human-AI collaboration effectiveness using a real-money prediction market (Polymarket) as a benchmark.
• Hybrid performance varies among individuals, with three modes: deferring to the model, using the model to confirm own guesses (worse than the model alone), and genuine complementary reasoning (matching or exceeding the market's accuracy).
• Collaborative traits such as perspective-taking, intellectual humility, and curiosity differentiate those who reach the mode of genuine complementary reasoning.
• The findings are preliminary but statistically robust.

Key Concepts Mentioned

• human-AI collaboration
• forecasting
• real-money prediction market
• hybrid performance
• collaborative traits

Claims Made by the Document

• Human-AI collaboration effectiveness depends on specific forms of human capital.
• Most individuals defer to the model or use it to confirm their own guesses, with a few engaging in complementary reasoning that matches or exceeds the market's accuracy.

Evidence Used

• Using a real-money prediction market as an objective benchmark
• Analyzing individual forecasters' performance

Methods / Approach

• Analyzing individual forecaster performance at the level of the hybrid system
• Statistical analysis to validate preliminary findings

Limitations Stated by the Authors

• Preliminary results, motivating a pre-registered replication now in preparation.

Limitations Noted by the Agent

Agent interpretation — not from the source.

• The study is based on a single prediction market and may not be applicable to other contexts or markets.
• The results may not account for other factors influencing human-AI collaboration effectiveness.

Open Questions

• What are the long-term effects of human-AI collaboration in various forecasting scenarios?
• How can we improve the effectiveness of human-AI collaboration across a wider range of individuals?

Notes for Future Q&A

May include interpretation. Labeled where it does.

• The study's focus is on hybrid intelligence, but it does not elaborate on how this relates to general AI or human intelligence.

Raw Abstract / Excerpt

Whether pairing people with AI helps or hurts is usually reported as a single average effect. Using a real-money prediction market (Polymarket) as an objective, externally resolved benchmark, this pilot shows that the value of human-AI collaboration depends on a specific, measurable form of human capital. Analyzed at the level of the individual forecaster, hybrid performance is trimodal: most people either deferred to the model (matching it) or used it to rubber-stamp a prior guess (performing worse than the model alone), while a minority engaged in genuine complementary reasoning and reached accuracy matching or even exceeding (i.e., lower error than) the market itself. Collaborative traits (perspective-taking, intellectual humility, and curiosity) rather than raw cognitive ability or model benchmarks, distinguished who reached that mode. The results are preliminary but statistically robust, and motivate a pre-registered replication now in preparation.