arXiv cs.AI· Muhammad Rizky Wellyanto, Hajara-Yasmin Isa, Kexin Xue, Vishal Ravichandran, Wenpeng Lu, Jerry O. Talton, Ranjitha Kumar·· 10 小时前AI 评分38
比图灵测试更难:LLM 能否在对话中冒充特定个人?个人模仿游戏
Who are you? The Personal Imitation Game
AI 导读
研究者提出比图灵测试更难的“个人模仿游戏”,考察机器能否在对话中复现某个具体个体而非泛化人类。基于一位作者十年个人数字通信数据构建的 LLM 系统,在 20 人欺骗实验中于 10 轮里有 7 轮被审讯者误认为作者本人,超过作者真实好友;作者本人则在 20 轮中 13 轮胜出,两项比例在该样本量下均与随机猜测无法区分。研究还归纳了审讯者依赖的身份信号:文体特征、人格特质与生平事实。
正文 · 原文
Abstract:Large Language Models (LLMs) now perform strongly on Alan Turing's 1950 Imitation Game, so we propose a harder challenge, the Personal Imitation Game, which asks whether a machine can replicate a specific individual in conversation rather than a generic human. Testing with 30 participants, we establish a human baseline and report the identity signals interrogators rely on: stylometric features, personality traits, and biographical facts. To demonstrate the framework, we design an LLM-based system that plays the game for one of the paper's authors, drawing on an autoethnographic dataset spanning ten years of personal digital communication. In a 20-person deception study, interrogators choose our system over the author's own friends in 7 of 10 rounds, and the author over our system in 13 of 20; neither rate is distinguishable from chance at this sample size. We consider design improvements, applications of personal AI agents, and implications for identity and privacy.
| Comments: | 19 pages, 6 figures, 5 tables |
| Subjects: | Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.24888 [cs.HC] |
| (or arXiv:2611.24888v1 [cs.HC] for this version) | |
| https://doi.org/10.48550/arXiv.2611.24888 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Muhammad Rizky Wellyanto [view email]
[v1]
Wed, 18 Nov 2026 18:32:14 UTC (801 KB)
来源:arXiv cs.AI · arxiv.org