跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2605.12646 · v3

Learning to Decide with AI Assistance under Human-Alignment

首次提交 2026-05-13 02:42
版本更新 2026-11-18 17:39
雷达观测 2026-11-19 15:04
arXiv ↗PDF ↗

摘要

It is widely agreed that when AI models assist decision-makers in high-stakes domains by predicting an outcome of interest, they should communicate the confidence of their predictions. However, empirical evidence suggests that decision-makers often struggle to determine when to trust a prediction based solely on this communicated confidence. In this context, recent theoretical and empirical work suggests a positive correlation between the utility of AI-assisted decision-making and the degree of alignment between the AI confidence and the decision-makers' confidence in their own predictions. Crucially, these findings do not yet elucidate the extent to which this alignment influences the complexity of learning to make optimal decisions through repeated interactions. In this paper, we address this question in the canonical case of binary predictions and binary decisions. We first show that this problem is equivalent to a two-armed online contextual learning problem with full feedback, and establish a lower bound of Ω(∣H∣⋅∣B∣⋅T)Ω(\sqrt{|H| \cdot |B| \cdot T} ) on the expected regret any learner can attain, where HH and BB denote the sets of human and AI confidence values. We then demonstrate that, under perfect alignment between AI and human confidence, a learner can attain an expected regret of O(∣H∣⋅Tlog⁡T)O(\sqrt{|H| \cdot T\log T}) and, when ∣H∣=O(log⁡T)\sqrt{|H|} = O(\log T) and BB is countable, a non-trivial generalization of the Dvoretzky-Kiefer-Wolfowitz inequality improves the regret bound to O(Tlog⁡T)O(\sqrt{T\log T}). Taken together, these results reveal that alignment can reduce the complexity of learning to make decisions with AI assistance. Experiments on real data from two different human-subject studies where participants solve simple decision-making tasks assisted by AI models show that our theoretical results are robust to violations of perfect alignment.

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

本站首次收录为 v3,暂无更早版本可对照。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v3 · 2026-11-18 17:39 · 最新观测版本

Learning to Decide with AI Assistance under Human-Alignment

It is widely agreed that when AI models assist decision-makers in high-stakes domains by predicting an outcome of interest, they should communicate the confidence of their predictions. However, empirical evidence suggests that decision-makers often struggle to determine when to trust a prediction based solely on this communicated confidence. In this context, recent theoretical and empirical work suggests a positive correlation between the utility of AI-assisted decision-making and the degree of alignment between the AI confidence and the decision-makers' confidence in their own predictions. Crucially, these findings do not yet elucidate the extent to which this alignment influences the complexity of learning to make optimal decisions through repeated interactions. In this paper, we address this question in the canonical case of binary predictions and binary decisions. We first show that this problem is equivalent to a two-armed online contextual learning problem with full feedback, and establish a lower bound of Ω(∣H∣⋅∣B∣⋅T)Ω(\sqrt{|H| \cdot |B| \cdot T} ) on the expected regret any learner can attain, where HH and BB denote the sets of human and AI confidence values. We then demonstrate that, under perfect alignment between AI and human confidence, a learner can attain an expected regret of O(∣H∣⋅Tlog⁡T)O(\sqrt{|H| \cdot T\log T}) and, when ∣H∣=O(log⁡T)\sqrt{|H|} = O(\log T) and BB is countable, a non-trivial generalization of the Dvoretzky-Kiefer-Wolfowitz inequality improves the regret bound to O(Tlog⁡T)O(\sqrt{T\log T}). Taken together, these results reveal that alignment can reduce the complexity of learning to make decisions with AI assistance. Experiments on real data from two different human-subject studies where participants solve simple decision-making tasks assisted by AI models show that our theoretical results are robust to violations of perfect alignment.

查看此版本解读 →查看此版本原文 ↗