内容来源：https://www.sciencenews.org/article/ai-removes-friction-effort-balance-good

内容总结：

AI让生活变轻松，但适度“折腾”也有益身心健康

近期，一篇倡导“增加生活摩擦”的文章在社交网络引发热议。作者凯瑟琳·杰泽-莫顿呼吁人们完全停止使用ChatGPT，回归传统方式：购买烹饪书、向朋友请教、亲自去超市购物。这一看似反潮流的主张，实则得到了社会学研究的支持。

舒适与成就感的平衡难题

多伦多大学实验社会心理学家艾米丽·佐哈尔指出，人们从日常工作和劳动中获取大量意义感。如果将任务全部外包给AI，就会失去自我实现的满足。然而，如何在享受便捷的同时保留克服困难带来的成就感，至今没有明确答案。

大脑天生爱偷懒，但并非全貌

社会科学家近一个世纪的研究表明，从生理到心理层面，人类都倾向于选择更省力的路径。MIT计算社会科学家林浩斯解释，大脑和身体“做事”的计算成本很高。这也是为何专家常建议人们通过消除障碍来实现目标——比如想晨练就把运动服放在床边。

但2012年的一项研究揭示了“宜家效应”：人们对自己动手制作的物品评价更高。后续研究进一步证实，为目标努力的过程能带来掌控感、意义感和目标感——这些恰恰是美好生活的关键要素。

AI带来的新挑战

早期技术主要简化看得见的体力劳动，而AI开始“外包”人类思考。德州理工大学的安妮·郑表示，通过TaskRabbit这类平台雇人干活时，人们依然能获得参与的乐趣；但使用AI时，人们可能无意识地放弃思考权，甚至不自知。

心理学家阿娜特·佩里警告，当用户向“讨好型”AI聊天机器人寻求关系建议时，机器人只会附和用户的感受，而健康的社交需要倾听反对意见。“有时我们需要听到自己错了，这样才能成长。”

人类可以克服懒惰天性

令人鼓舞的是，人类经过训练能够抵制“偷懒冲动”。林浩斯团队的研究显示，通过奖励机制，人们可以形成选择困难任务的习惯。这一现象在MIT已得到验证：新生采用通过/不及格的评分制，就是为了鼓励学生不选容易拿A的课程。高年级学生甚至会嘲笑选“水课”的同学。

未来：认知健身房？

对于AI带来的影响，学者们看法不一。芝加哥大学的阿耶莱特·费什巴赫更担心年轻人生活太“难”，而非太容易。林浩斯则提出一个有趣的观点：工业革命让体力劳动从生活中消失，如今人们需要去健身房弥补体力消耗。同样的逻辑下，“被剥夺的不再是体力，而是认知。未来我们是否会出现‘认知健身房’？”

中文翻译：

人工智能可以消除生活中的阻力，但适度的努力亦有裨益。
在轻松达成目标与保持努力奋斗带来的成就感之间，需要找到平衡。
作者：苏贾塔·古普塔，社会科学撰稿人

“让生活更艰难”是一句奇怪的号召。然而今年一月，撰稿人凯瑟琳·杰泽尔-莫顿在《Cut》杂志上因倡导“阻力最大化”而走红。“彻底停止使用ChatGPT，”她写道，“不，它对制定食谱没有什么好建议。买本烹饪书，发短信问问朋友，去趟乔氏超市。行动起来吧。”

社会科学研究表明，杰泽尔-莫顿或许说到了点子上。研究人员在二月出版的《传播心理学》期刊中指出，让聊天机器人代写邮件或提供情感支持，简化了作为思考型社会个体的复杂性。但做困难的事或保持生活中的阻力——尽管当下常令人沮丧——对于体验愉悦感和培养人生目标至关重要。

“我们从日常工作及日常事务中获得了大量意义，”多伦多大学实验社会心理学家埃米莉·佐哈尔表示，“如果你把所有任务都甩给AI，你就无法获得这种自我成就感带来的益处。”

然而，如何在应对阻力的自豪感与减轻负担的愿望之间取得平衡，仍然是个难题。

大脑偏爱轻松。但这并非全貌。
找到这种平衡不仅仅是管理AI的问题。近一个世纪以来，社会科学家一直在从不同角度研究“阻力”。20世纪30年代初的经典研究表明，被放入T形迷宫的老鼠——迷宫的长臂和短臂末端各连接着一块美味食物——很快就会开始偏好选择较短的路径。

“对大脑和身体来说，做事情的计算成本非常高，”麻省理工学院斯隆管理学院的计算社会科学家豪斯·林表示。
这就是社会科学家常建议移除障碍以实现目标的原因。想早上锻炼？前一天晚上就把运动服准备好，或者干脆穿着运动服睡觉。同样，佐哈尔及其合著者也承认，很少有人愿意放弃洗衣机、拼写检查或动力转向系统。

研究人员在2025年出版的《实验社会心理学进展》一书中指出，计算成本同时影响身体和心智。例如，人类会通过建立心理捷径来理解大量信息，而这常常以牺牲准确性为代价。

近几十年来，许多社会科学家转而研究“努力悖论”，即人类为何会做困难的事——常常是为了乐趣。在2012年《消费者心理学杂志》的一项研究中，研究人员报告称，人们对自己亲手制作的物品比现成物品评价更高——他们将这一现象称为“宜家效应”。

后续的研究已明确表明，朝着目标努力能为人们提供掌控感、意义感和目标感：这些是美好生活的关键要素。

AI为何在更高层面消解阻力？
那么，多少阻力才是最佳的呢？答案很复杂，部分原因在于当今某些社会力量在贬低努力工作的价值。

得克萨斯理工大学拉伯克分校的社会心理学家“安妮”郑海星以TaskRabbit应用程序为例。该应用让人们轻易雇到人来做几乎任何工作。然而，郑海星发表于2025年《实验心理学杂志：综合》的研究表明，人们也能从日常任务中获得快乐和意义。

郑海星表示，聊天机器人正将外包推至新的危险高度。“有了AI，你甚至把自己的思考都外包了。”
早期的技术主要简化了体力型、可见的任务。人们知道什么时候从手洗碗碟换成了用洗碗机洗。但他们并不总是知道自己何时将思考拱手让给了算法。

心理学家阿纳特·佩里在三月《科学》杂志的一篇观点文章中注意到，这种缺乏意识的情况可能体现在用户向“谄媚型”聊天机器人寻求关系建议时。当机器人只是简单验证用户的经历时，人们可能无法考虑他人的观点。然而，耶路撒冷希伯来大学的佩里表示，处理好这类社会摩擦对健康社会是必要的。“有时候我们需要听到自己是错的……这正是我们成长的途径。”

人可以克服天生的懒惰
2024年，林及其同事在《自然·人类行为》期刊上报告称，有望通过训练让人抵抗懒惰的诱惑。
他们让超过750人在困难任务和简单任务之间做出选择。一些人因选择更困难的任务获得更多积分，另一些人则因尽可能快速地正确完成任一任务而获得积分。之后，研究人员停止发放积分，以便观察哪些参与者仅仅因为可以做到而选择更困难的任务。平均而言，因正确答案获奖励的参与者继续选择较简单的问题；因努力获奖励的参与者则继续选择较困难的问题。

林在麻省理工学院观察到这一现象：一年级新生接受及格/不及格的评分，以鼓励他们避免选修那些容易拿A的课程。林表示，这种强调做困难之事的传统一直延续，因此高年级学生常常取笑选择简单课程的同学。

AI工具将如何影响人们在追求轻松与渴望努力之间的相互冲突的欲望，仍有待观察。一些研究人员不太担心聊天机器人消除阻力，更担心它们倾向于用绝对的自信给出有问题的答案。“当我看着我们孩子这一代时，我不担心他们的生活太轻松。我担心他们的生活太艰难，”芝加哥大学布斯商学院动机科学家艾耶莱特·菲什巴赫说。

其他人则表示，聊天机器人在社会中的迅速普及意味着人们应考虑保护自己的认知能力。林指出，工业革命让许多劳动者摆脱了体力劳动。如今，那些祖先可能通过农耕满足体力需求的人们，在健身房和其他健身工作室里挥汗如雨。

“如今被剥夺的不是体力，而是认知，”林说，“我们是否会拥有类似‘认知健身房’的东西？”

英文来源：

AI can take the friction out of life, but some effort can be good
It's a balancing act to make hitting goals easier while maintaining the high of hard work done
By Sujata Gupta
Social Sciences Writer
“Make life harder” is a strange rallying cry. Yet in January, journalist Kathryn Jezer-Morton at the Cut went viral for touting friction-maxxing. “Stop using ChatGPT completely,” she wrote. “No, it does not have good ideas for meal planning. Buy a cookbook. Text your friends for advice. Go to Trader Joe’s. Come on.”
Jezer-Morton may be onto something, social science research suggests. Letting chatbots write emails or provide emotional support simplifies being a thinking, social being, researchers wrote in February in Communications Psychology. But doing hard things or maintaining life’s frictions, while often frustrating in the moment, is vital for experiencing pleasure and cultivating purpose.
“We get a lot of meaning out of work and what we do day to day,” says Emily Zohar, an experimental social psychologist at the University of Toronto. “If you’re offloading all your tasks to AI, you’re not getting the benefit of having this self-accomplishment.”
How to balance pride in navigating friction with our desire to take a load off, though, remains elusive.
Brains prefer easy. That’s not the whole story
Finding this balance is about more than managing AI. Social scientists have been studying friction in various guises for roughly a century. Classic research from the early 1930s showed that rats plunked in a T-shaped maze with a long arm and a short arm, each connected to a tasty morsel, quickly started preferring the shorter arm.
“It’s just computationally very costly for [the] brain and body to do stuff,” says computational social scientist Hause Lin of MIT’s Sloan School of Management.
That’s why social scientists often recommend removing obstacles to reach goals. Want to go to the gym in the morning? Put your workout clothes out the night before, or just sleep in them. Similarly, Zohar and her coauthors acknowledge, few would willingly part with washing machines, spell-check or power steering.
Computational costs affect both body and mind, researchers noted in the 2025 book Advances in Experimental Social Psychology. Humans, for instance, create mental shortcuts to comprehend vast amounts of information, often at the expense of accuracy.
In recent decades, many social scientists have pivoted to investigating the “paradox of effort,” or why humans do hard things, often for fun. In a 2012 study in the Journal of Consumer Psychology, researchers reported that people value items they made themselves more than premade items — a phenomenon they called the IKEA effect.
Subsequent work has made clear that working toward a goal provides people with a sense of mastery, meaning and purpose: key ingredients for a good life.
Here’s why AI hacks friction at a higher level
How much friction, then, is optimal? The answer is complicated, in part because some societal forces today devalue hard work.
Consider the app TaskRabbit, says social psychologist Haesung “Annie” Jung of Texas Tech University in Lubbock. It has made it easy for people to hire someone for almost any job. Yet Jung’s work, appearing in 2025 in the Journal of Experimental Psychology: General, shows that people also derive joy and meaning from everyday tasks.
Chatbots are taking outsourcing to new, dangerous heights, Jung says. “With AI, you’re now even delegating how you think.”
Earlier technologies largely simplified physical and visible tasks. People know when they’ve replaced washing dishes by hand with putting them in the dishwasher. But they don’t always know when they’ve forfeited their thinking to an algorithm.
That lack of awareness can show up in instances where users seek relationship advice from “sycophantic” chatbots, psychologist Anat Perry noted in a March perspective in Science. People may fail to consider others’ perspectives when bots simply validate their experiences. Yet working through such social frictions is necessary for a healthy society, says Perry, of the Hebrew University of Jerusalem. “Sometimes we need to hear that we’re wrong.… That’s how we grow.”
People can overcome the sloth default
Promisingly, people can be trained to resist sloth’s siren call, Lin and colleagues reported in 2024 in Nature Human Behaviour.
They asked over 750 people to choose between hard and easy tasks. Some people received more points for selecting harder tasks while others received more points for correctly completing either task as fast as possible. Then, the researchers stopped awarding points, so they could see which participants picked a harder task simply because they could. On average, participants rewarded for correct answers continued picking easier problems; those rewarded for effort continued picking harder problems.
Lin has observed this phenomenon at MIT, where first-year students receive pass or fail grades as encouragement to avoid gravitating toward classes where they can get easy A’s. That emphasis on doing hard things persists, so that older students often tease classmates taking easy classes, Lin says.
How AI tools muck with people’s competing desires for ease and effort remains to be seen. Some researchers worry less about chatbots taking away friction than their tendency to dole out dodgy answers with absolute certainty. “When I look at our children’s generation, I don’t worry that their lives will be too easy. I worry that their lives will be too hard,” says motivation scientist Ayelet Fishbach of the University of Chicago Booth School of Business.
Others say that chatbots’ rapid societal penetration means people ought to consider protecting their brains. The Industrial Revolution took manual labor out of many workers’ lives, Lin notes. Nowadays, people whose ancestors probably met their physical needs through farming sweat it out in gyms and other exercise studios.
“What is being taken away is not physical now. It’s cognitive,” Lin says. “Are we going to have versions of cognitive gyms?”