arXiv 2507.139192025-07-18cs.CLcs.AIcs.CYcs.HC

对话式人工智能的政治说服杠杆

The Levers of Political Persuasion with Conversational AI

UK AI Security Institute（英国人工智能安全研究所）；University of Oxford（牛津大学）；The London School of Economics and Political Science（伦敦政治经济学院）；Stanford University（斯坦福大学）；Massachusetts Institute of Technology（麻省理工学院）

Kobi Hackenburg, Ben M. Tappin, Luke Hewitt, Ed Saunders, Sid Black, Hause Lin, Catherine Fist, Helen Margetts, David G. Rand, Christopher Summerfield

备注：: 19 pages, 4 figures. Our supplementary materials file can be found at https://github.com/kobihackenburg/scaling-conversational-AI

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

通过三项大规模实验和19个大语言模型，研究发现对话式AI的政治说服力主要来自后训练与提示方法而非模型规模，且说服力提升的同时事实准确性系统性下降。

摘要：

人们普遍担忧对话式人工智能可能很快对人类信念施加前所未有的影响。在此，我们通过三项大规模实验（N=76,977），部署了19个大语言模型——其中一些经过专门针对说服任务的后训练——以评估它们在707个政治议题上的说服力。随后，我们核查了由此产生的466,769条大语言模型陈述的事实准确性。与普遍的担忧相反，我们表明当前及近期人工智能的说服力可能更多来源于后训练和提示方法——它们分别将说服力提升了高达51%和27%——而非个性化或模型规模的扩大。我们进一步表明，这些方法通过利用大语言模型快速获取并策略性部署信息的独特能力来增强说服效果，并且引人注目的是，在它们提升人工智能说服力的同时，也系统性地降低了事实准确性。

英文摘要：

There are widespread fears that conversational AI could soon exert unprecedented influence over human beliefs. Here, in three large-scale experiments (N=76,977), we deployed 19 LLMs-including some post-trained explicitly for persuasion-to evaluate their persuasiveness on 707 political issues. We then checked the factual accuracy of 466,769 resulting LLM claims. Contrary to popular concerns, we show that the persuasive power of current and near-future AI is likely to stem more from post-training and prompting methods-which boosted persuasiveness by as much as 51% and 27% respectively-than from personalization or increasing model scale. We further show that these methods increased persuasion by exploiting LLMs' unique ability to rapidly access and strategically deploy information and that, strikingly, where they increased AI persuasiveness they also systematically decreased factual accuracy.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。