11-19 更新 实时
首页/ 算力/ 正文

有友善面孔可爱外形的 AI 智能体或让你忽视严重风险

算力 来源:TechXplore 2026-11-06 00:00 0 阅读 机器翻译
有友善面孔可爱外形的 AI 智能体或让你忽视严重风险

作者:Ankolika De、Jessica Maddox,The Conversation

编辑:Lisa Lock,审核:Andrew Zinin

本文已按照 Science X 的编辑流程与政策进行审核。编辑在确保内容可信度的同时,重点标注了以下特征:

由研究人员撰写

添加到首选来源

Meta 的 Muse 是一款个人人工智能智能体,它能替你发送邮件、购物、预订旅行,并调用你的应用代你行事。当你从手机或电脑上召唤它时,会出现一个名叫 Jolly 的小化身,看起来有点像一只数字化的泰迪熊,会朝你眨眼、挥手。

Meta 还发布了 Muse Charm,这是一款屏幕只有压扁的高尔夫球大小的设备,可以放在口袋里,让你随时访问 Jolly。记者们将 Muse 形容为“神奇”且“可爱”。他们把 Charm 比作 Tamagotchi——一种掌上电子宠物,你把它从蛋里孵化出来,陪伴它经历婴儿、儿童、青少年和成年阶段,在 1990 年代风靡一时。

紧随 Muse 之后,OpenAI 推出了其个人助手:Dots。它们带来了一组色彩鲜艳的球形吉祥物面孔,漂浮在你的屏幕上与你交谈。OpenAI 表示,Dots 将“以你会采用的方式替你完成工作,有时甚至在你想到要开口之前就已经做好。”

初创公司也在下同样的赌注。Pickle 是一款个人 AI 应用,其形象是一个漂浮的紫色三维动物,长着大大的、充满好奇的眼睛。其宣传材料称,人们会更愿意与“一个有生命感、你会逐渐依恋的东西”多说话、多分享,它会记住你的生活,并最终与你并肩行动。

尽管可爱的设计本身并不邪恶,但它们能让陌生的技术显得更容易接受、更令人安心。可爱也能掩盖 AI 智能体的侵入性潜力,淡化用户对大规模数据收集和监控等风险的感知,并分散人们的注意力,使其在没有明确同意的情况下允许智能体代表自己行动。

我们是数字技术学者,多年来一直在研究公司用来表述技术的语言和宣传信息。我们认为,用让人放下戒备的外包装来营销高科技产品这种障眼法,或许正是可爱的最大力量,而科技公司似乎正在利用它来鼓励人们采用 AI。

把新技术呈现为可爱并不是什么新鲜事。Hello Kitty、Labubus、Tamagotchis 等产品以推广可爱而闻名,借此唤起消费者的情感依恋。然而,它们只是收藏品,基本上无害。它们无法读取主人的消息、访问主人的账户或代其行动。但当这个可爱的东西是 AI 智能体时,利害关系就大得多了。

我们已经表明,谈论技术的特定方式会影响人们对安全的认知,并使技术的危险变得不那么显眼。“可爱洗白”(Cutewashing)可以掩盖隐私担忧,并推动许多人采用,从而创造大众市场。

利用你的情感

可爱的事物能让你感到快乐、信任、喜爱或想要保护,促使你与某项技术建立情感联系。

尤其是大型科技公司,长期以来一直拥抱一个由可爱动物组成的虚拟动物园。硅谷给世界带来了 Firefox 狐狸、现已停运的 Gateway 奶牛、昔日的 Twitter 小鸟、GitHub 的章鱼、Mailchimp 的猴子以及 Hootsuite 的猫头鹰。生物给人的纯真感,能软化人们对创新后果的感知。

AI 公司也在追逐一些研究者所称的“可爱经济”来赚钱。可爱所激发的情感能建立品牌忠诚度,并促使用户持续使用产品。可爱也可以充当一种营销策略,因为用户会对产品产生情感投入。这反过来又会影响购买行为。

研究还显示,相比一个朴素的工具,人们更愿意向友好的角色倾诉,分享心情、日常、人际关系和偏好。人们会授予友好的智能体访问其消息、日历和账户的权限,从而让该智能体所属公司挖掘到远比常规聊天机器人通常获得的更丰富的数据。可爱能降低戒心,使人们更可能授予宽泛权限,也更不可能去细读那些小字条款。情感依恋也让用户更难离开某个智能体。

Meta 正利用 Muse Charm 的可爱形象,把 AI 放进你的口袋。

把一项技术做得可爱,会影响人们注意到它的哪些方面、又忽略了哪些方面。一个讨人喜欢的 AI 助手可能显得更平易近人、更不具威胁性,从而可能把人们的注意力从数据收集、隐私、安全和监控等问题上引开。举例来说,Muse 曾在用户不知情的情况下读取了某位用户的短信,并泄露了另一位用户的家庭住址。

为了帮助人们对可爱的 AI 智能体采取更为审慎的态度,我们认为企业应当向用户明确说明:智能体能访问什么、能做什么,以及谁从中受益。信息披露和隐私政策应当用通俗语言写明,而不是晦涩的法律术语。

与此同时,在采用某个智能体之前,先问清楚它能访问哪些数据,以及它能利用这些访问权限做什么。重要的是,不要让“可爱”分散你对使用这项技术所带来后果的注意。

由 The Conversation 提供

这篇报道的作者是谁?

艺术史学士,物质文化硕士。曾任博物馆编辑、急救员和器官移植协调员。自 2021 年起为 Science X 从事编辑工作。完整简介 →

物理学硕士,有科研经历。长期的科学新闻爱好者。在 Science X 的编辑工作中发挥重要作用。完整简介 →

本文依据 Creative Commons 许可从 The Conversation 转载。阅读原文。

引用:拥有友善面孔和可爱外形的 AI 智能体可能让你忽视严重风险(2026 年 11 月 5 日),2026 年 11 月 6 日检索自 https://techxplore.com/news/2026-11-ai-agents-friendly-adorable-bodies.html

本文件受版权保护。除为私人学习或研究目的进行的任何合理使用外,未经书面许可,不得复制本文件的任何部分。内容仅供参考。

获取即时摘要文本(Gist)

可爱的 AI 界面可以培养信任与情感依恋,鼓励用户共享数据并授予广泛权限,同时掩盖隐私、安全和监控风险。需要明确披露智能体的数据访问权限和能力,以支持知情同意。

本摘要由 LLM 自动生成。完整免责声明

译文由英文原文机器翻译,术语按行业惯例统一,可在文末展开原文核对。

查看英文原文(AI agents with friendly faces and adorable bodies can distract you from serious risks)

AI agents with friendly faces and adorable bodies can distract you from serious risks

AI agents with friendly faces and adorable bodies can distract you from serious risks

by Ankolika De, Jessica Maddox, The Conversation

edited by Lisa Lock, reviewed by Andrew Zinin

This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:

written by researcher(s)

Add to Preferred Sources

meta's muse ai agent Credit: Unsplash/CC0 Public Domain

Meta's Muse is a personal artificial intelligence agent that sends email for you, shops and books travel for you and uses your apps to act on your behalf. When you call on it from your phone or computer, a little avatar named Jolly appears, looking something like a digitized teddy bear that blinks and waves to you.

Meta also released Muse Charm, a device with a screen the size of a flattened golf ball that you can keep in your pocket so you can access Jolly anytime. Journalists have described Muse as "magical" and "cutesy." They have compared Charm to a Tamagotchi, a handheld digital pet that you hatched from an egg and guided through baby, child, teen and adult stages, which was all the rage in the 1990s.

Right on Muse's heels, OpenAI introduced its personal assistants: Dots. They come with a family of brightly colored, globular mascot faces that float across your screen and converse with you. OpenAI says Dots will "bring you work done the way you would do it, sometimes before you even think to ask."

Startups are making the same bet. The promotional material for Pickle, a personal AI app embodied in a floating, purple, three-dimensional animal that has big, inquisitive eyes, says that people will talk more, and share more, with "something alive you grow attached to that remembers your life and eventually acts alongside you."

Although cute designs are not inherently sinister, they can make unfamiliar technology feel much more accessible and comforting. Cuteness can also obscure the invasive potential of AI agents, soften how users perceive risks such as extensive data collection and surveillance and distract people into allowing agents to act on their behalf without clear consent.

We are scholars of digital technologies, and we have spent years researching the language and messaging companies use to represent technologies. We believe that this sleight of hand of marketing high-tech goods in a disarming package may be the greatest power of cuteness, and tech companies appear to be harnessing it to encourage people to adopt AI.

Presenting new technologies as cute is nothing new. Hello Kitty, Labubus, Tamagotchis and other products have famously promoted cuteness to evoke emotional attachment in consumers. They were just collectibles, however, and largely harmless. They couldn't read an owner's messages, access an owner's accounts or act for them. When the cute thing is an AI agent, however, the stakes are much higher.

We have shown how particular ways of talking about technology can influence perceptions of safety and make dangers of technologies less visible. "Cutewashing" can mask privacy concerns and drive adoption by many people, creating a mass market.

Playing on your feelings

Cute things can make you feel happy, trusting, affectionate or protective, encouraging you to form emotional connections with a technology.

Big Tech, in particular, has long embraced a virtual menagerie of cute animals. Silicon Valley has given the world the Firefox Fox, now-defunct Gateway cows, the former Twitter bird, GitHub's octopus, Mailchimp's monkey and Hootsuite's owl. The perceived innocence of creatures can soften how people perceive the consequences of innovation.

AI companies are pursuing what some researchers have referred to as the "cute economy" to make money as well. Feelings generated through cuteness can build brand loyalty and encourage people to keep engaging with the products. Cuteness can function as a marketing strategy, too, as users become emotionally invested in a product. This, in turn, can influence purchases.

Research also shows that people confide in a friendly character more than a plain tool, sharing moods, routines, relationships and preferences. People will grant friendly agents access to their messages, calendars and accounts, yielding far richer data that the agent's company can plumb than is usually obtained by a routine chatbot. Cuteness can lower skepticism so people are more likely to grant broad permissions and less likely to read the fine print. Attachment also makes it more difficult for users to leave an agent.

Meta is using the cuteness of Muse Charm to put AI in your pocket.

Making a technology cute can influence what people notice about it and what they overlook. A charming AI assistant may seem more approachable and less threatening, potentially drawing attention away from questions about data collection, privacy, security and surveillance. In one example, Muse has read a user's text messages and given out another's home address without the users' knowledge.

To help people take a more cautious view of cute AI agents, we believe that companies should clearly explain to users what an agent can access, what it can do and who benefits. Disclosures and privacy policies should be stated in plain language, not dense legalese.

In the meantime, before you adopt an agent, ask what data it can access and what it can do with that access. It's important to avoid letting cuteness be a distraction from the consequences of using the technology.

Provided by The Conversation

Who's behind this story?

Lisa Lock

BA art history, MA material culture. Former museum editor, paramedic, and transplant coordinator. Editing for Science X since 2021. Full profile →

Master's in physics with research experience. Long-time science news enthusiast. Plays key role in Science X's editorial success. Full profile →

This article is republished from The Conversation under a Creative Commons license. Read the original article.The Conversation

Citation: AI agents with friendly faces and adorable bodies can distract you from serious risks (2026, November 5) retrieved 6 November 2026 from https://techxplore.com/news/2026-11-ai-agents-friendly-adorable-bodies.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.

Get Instant Summarized Text (Gist)

Cute AI interfaces can foster trust and emotional attachment, encouraging data sharing and broad permissions while obscuring privacy, security and surveillance risks. Clear disclosures about agents’ data access and capabilities are needed to support informed consent.

This summary was automatically generated using LLM. Full disclaimer

This site uses cookies to assist with navigation, analyse your use of our services, collect data for ads personalisation and provide content from third parties. By using our site, you acknowledge that you have read and understand our Privacy Policy and Terms of Use.

原文为英文,正文由自动抓取并保留原文。

本文由采集器自动同步

原文发布于 TechXplore(AI agents with friendly faces and adorable bodies can distract you from serious risks),标题、摘要与正文由机器翻译,术语按行业惯例统一,未经人工逐句校订。

阅读原文 ↗