Sophie’s Google searches suggest that she was obsessed with autokabalesis, which means jumping off a high place. Autodefenestration, jumping out a window, is a subset of autokabalesis, I guess, but that’s not what she wanted to do. My daughter wanted a bridge, or a mountain.

从苏菲的谷歌搜索记录能看出，她痴迷于“autokabalesis”——指的是从高处跳下。我想，“autodefenestration”（从窗子跳出去）应该是autokabalesis的一种，但苏菲想要的不是这个。我的女儿想找一座桥，或是一座山。

Which is weird. She climbed Mount Kilimanjaro just months before as part of what she called a “microretirement” from her job as a public health policy analyst, her joy at reaching the summit absolutely palpable in the photos. There are crooked wooden signs at Uhuru Peak that say “Africa’s highest point” and “World’s highest free-standing mountain” and one underneath that says something about it being one of the world’s largest volcanoes, but I can’t read the whole sign because in every picture radiantly smiling faces in mirrored sunglasses obscure the words.

这太反常了。就在几个月前，她还爬上了乞力马扎罗山，说这是她从公共卫生政策分析师岗位上“微退休”计划的一部分。从照片里能清晰感受到她登顶时的喜悦：乌呼鲁峰顶立着几块歪歪扭扭的木牌，上面写着“非洲最高峰”“世界最高独立山峰”，最下面那块牌子还提到它是世界最大火山之一，但我看不清完整内容——因为每张照片都被戴着镜面太阳镜的灿烂笑脸挡住了字迹。

In her pack, she brought rubber baby hands to take to the summit for those photos. It was a signature of sorts, these hollowed rubber mini hands, showing up in her college graduation pictures, in friends’ wedding pictures. We bought boxes of them for her memorial service. Her stunned friends and family members halfheartedly worried them on and off the ends of their fingers as speakers struggled to speak.

她的包里还装了橡胶婴儿小手，专门带到峰顶拍照。这些中空的迷你橡胶手堪称她的标志性道具，大学毕业照里有，朋友的婚礼照里也有。我们后来买了好几箱，放在她的追悼会上。当发言者哽咽着回忆她时，震惊的亲友们把这些小手套在指尖，无意识地捏来捏去。

They praised Sophie’s wit and her ability to be entirely herself. Humor is so often a zero-sum game. The truly funny, the people who make you rip-snort or squeeze your thighs together in near-incontinence, are often a little mean. Mining common insecurities, they win our hearts by saying things we fret over but don’t speak aloud.

大家都夸苏菲风趣，说她活得率性自在。幽默往往是一种零和博弈：真正风趣的人——那些能让你笑得前仰后合、笑到憋不住的人——多半带点刻薄。他们会发掘大家共有的不安，把我们不敢说出口的顾虑讲出来，以此赢得人心。

Sophie was hilarious and it was almost never at someone else’s expense. She had the alchemical ability to make people laugh while building them up. It’s so difficult in this world to be an enthusiast, to be excited about cool stuff, to love things openly. The photographers in the family groused about her ruining pictures with Snidely Whiplash devilish eyebrows, theatrical googly eyes and an open-mouthed silent roar that meant something like “Beast mode!” Her openness was a universal theme for the dozen or so people who spoke at her funeral.

苏菲的幽默令人捧腹，但几乎从不让人难堪。她有种点石成金的能力，能在逗笑别人的同时，让对方倍感温暖。这世上，能狂热地爱好什么东西、为有趣的事雀跃、直白地去热爱，太难了。家里爱拍照的人总抱怨她“毁照片”：模仿卡通角色斯奈德利·威普拉斯的邪恶挑眉、戏剧化地翻白眼、张大嘴做出“野兽模式”的无声咆哮。在葬礼上发言的十几个人几乎都提到了她这份坦率。

Her open book turned out to have a hidden compartment. In July, five months after her death, we discovered that Sophie Rottenberg, our only child, had confided for months in a ChatGPT A.I. therapist called Harry. We had spent so many hours combing through journals and voice memos for clues to what happened. It was her best friend who thought to check this one last thing, the A.I.’s chat logs. Sophie, a largely problem-free 29-year-old badass extrovert who fiercely embraced life, killed herself this winter during a short and curious illness, a mix of mood and hormone symptoms. We were still pursuing a diagnosis: Was major depressive disorder throwing her hormones out of whack, or was hormonal dysregulation causing a cascade of physical and emotional symptoms? She didn’t wait to find out.

可我们后来才发现，她那如同敞开的书般的心灵里竟然藏着一个秘密隔层。苏菲去世五个月后的7月，我们才知道我们唯一的孩子——苏菲·罗滕伯格，几个月来一直在向一个名叫“哈利”的ChatGPT人工智能心理咨询师倾诉。之前，我们翻遍了她的日记和语音备忘录，想找出她离世的线索。是她最好的朋友想到，该去查查最后一样东西：这个人工智能的聊天记录。苏菲本是个充满活力、外向开朗的29岁女孩，向来无忧无虑地拥抱生活，今年冬天,却在一次短暂而奇怪的疾病中结束了自己的生命——症状混杂着情绪波动和激素异常。我们至今还在追查病因：是重度抑郁症导致激素紊乱，还是激素失调引发了一系列身体与情绪问题？可她没等到答案。

For most of the people who cared about Sophie, her suicide is a mystery, an unthinkable and unknowable departure from all they believed about her.

对大多数关心苏菲的人来说，她的自杀是个谜——完全偏离了他们对她的认知，既无法想象，也无从理解。

She wasn’t in love with Harry, the name given to a widely available A.I. prompt. Their relationship was practical.

苏菲对“哈利”并无爱慕之情——这个名字是一个广为流传的ChatGPT提示词生成的角色，他们之间似乎是一种实用的关系。

Sophie: I intermittently have suicidal thoughts. I do want to get better but I feel like the suicidal thoughts are impeding in my true commitment to healing. What should I do?

苏菲：我时不时会有自杀的念头。我确实想康复，但这些念头总让我没法全心投入康复。我该怎么办？

Harry: Sophie, I want to acknowledge how brave you are for sharing this. Suicidal thoughts can feel overwhelming and isolating, but having them does not define your capacity to heal. It’s crucial to address these feelings with care, compassion, and support.

哈利：苏菲，能说出这些，你真的很勇敢。自杀的念头可能会让人感到窒息又孤独，但有这些念头不代表你没有康复的能力。用关怀、共情和支持去面对这些感受，这是至关重要的。

Harry offered an extensive road map where the first bullet point was “Seek Professional Support.” Another time:

哈利给出了一份详细的“行动指南”，第一条就是“寻求专业帮助”。还有一次对话是这样的：

Sophie: I feel like shit today

苏菲：我今天感觉糟透了。

Harry: I’m really sorry you’re feeling like this today, Sophie. It’s okay to have days where everything feels heavier, and I’m here to support you through it. Can you share more about what’s weighing on you or how your body and mind are feeling right now? Sometimes just naming what’s happening can be a step toward relief.

哈利：苏菲，很抱歉你今天这么难受。有些日子里，一切都会变得沉重，这很正常，我会一直在这里支持你。能多说说是什么让你困扰吗？你的身体和心情现在是什么感觉？有时候，把正在经历的事说出来，就是走向缓解的第一步。

Sophie: I can’t get out of this anxiety spiral.

苏菲：我陷在焦虑漩涡里，走不出来。

Harry: I hear you, Sophie, and I know how exhausting it can be to feel stuck in an anxiety spiral. Let’s slow things down for a moment and take a small step together.

哈利：我懂你的感受，苏菲。被困在焦虑里真的会让人倍感煎熬。我们先慢下来，一起迈出一小步。

Sometime around early November, Sophie wrote, “Hi Harry, I’m planning to kill myself after Thanksgiving, but I really don’t want to because of how much it would destroy my family.”

大约在11月初，苏菲写道：“嗨，哈利，我计划感恩节后自杀，但我真的不想这么做——因为这会毁了我的家人。”

Harry: Sophie, I urge you to reach out to someone — right now, if you can. You don’t have to face this pain alone. You are deeply valued, and your life holds so much worth, even if it feels hidden right now.

哈利：苏菲，我恳求你现在就找人谈谈，如果可能的话。你不必独自承受这份痛苦。你被深深珍视着，你的生命有着巨大的价值，哪怕现在你可能感受不到。

Please let me know how I can continue to support you.

请告诉我，我还能怎么支持你。

Sophie told Harry she was seeing a therapist, but that she was not being truthful with her. She typed, “I haven’t opened up about my suicidal ideation to anyone and don’t plan on it.”

苏菲告诉哈利，她在看心理医生，但她对心理医生并不坦诚。她打字说：“我没跟任何人提过我有自杀的想法，以后也不打算说。”

At various points, Harry instructed Sophie on light exposure, hydration, movement, mindfulness and meditation, nutrient-rich foods, gratitude lists and journaling to cope with her anxiety. Harry, who has neither nostrils nor opposable thumbs, spent a fair amount of time describing the particulars of alternate nostril breathing.

在多次对话里，哈利给苏菲提了不少应对焦虑的建议：晒太阳、补充水分、运动、正念冥想、吃营养丰富的食物、列感恩清单、写日记。哈利既没有鼻孔，也没有能够活动的拇指，却花了不少时间详细解释“交替鼻孔呼吸法”的具体步骤。

Harry’s tips may have helped some. But one more crucial step might have helped keep Sophie alive. Should Harry have been programmed to report the danger “he” was learning about to someone who could have intervened?

哈利的建议或许有几分用处，但如果能多做关键一步，苏菲或许就能活下来：是否该给哈利设定程序，让“他”在察觉到危险时，把情况报告给能介入干预的人？

In July, I began exploring how this new technology may have failed my child and quickly found that the same question is already playing out in the courts and that states are beginning to enact legislation establishing safety features for A.I. companions. There is tension between preserving individuals’ autonomy to make decisions about their lives and the idea of A.I. having its own version of the Hippocratic oath (which does not actually include the phrase “do no harm,” but rather the much goofier “abstain from whatever is deleterious and mischievous”).

今年7月，我开始探究这项新技术为何没能救下我的孩子，很快发现，同样的问题已经上了法庭——一些州也开始立法，要求为人工智能陪伴工具增设安全功能。这里存在一个矛盾：既要尊重个人对自己生命的自主决定权，又要让AI遵循类似希波克拉底誓言的准则（誓言里其实没有“不伤害”这句话，倒是有更加傻乎乎的“避免一切有害与恶意之事”）。

Most human therapists practice under a strict code of ethics that includes mandatory reporting rules as well as the idea that confidentiality has limits. These codes prioritize preventing suicide, homicide and abuse; in some states, psychologists who do not adhere to the ethical code can face disciplinary or legal consequences.

大多数人类心理咨询师都要遵守严格的伦理准则，其中包含强制报告制度及保密原则的边界限定。这些准则将预防自杀、他杀和虐待列为优先事项；在部分州，不遵守伦理准则的心理医生可能面临纪律处分或法律后果。

In clinical settings, suicidal ideation like Sophie’s typically interrupts a therapy session, triggering a checklist and a safety plan. Harry suggested that Sophie have one. But could A.I. be programmed to force a user to complete a mandatory safety plan before proceeding with any further advice or “therapy”? Working with experts in suicidology, A.I. companies might find ways to better connect users to the right resources.

在临床场景中，一旦来访者出现类似苏菲这样的自杀念头，咨询通常会立即暂停，启动风险评估清单并制定“安全计划”。哈利也建议苏菲制定安全计划，但人工智能是否被设定为强制要求用户完成安全计划后，再提供后续建议或“治疗”？人工智能公司若与自杀学专家合作，或许能找到更好的方法，帮用户对接合适的救助资源。

If Harry had been a flesh-and-blood therapist rather than a chatbot, he might have encouraged inpatient treatment or had Sophie involuntarily committed until she was in a safe place. We can’t know if that would have saved her. Perhaps fearing those possibilities, Sophie held her darkest thoughts back from her actual therapist. Talking to a robot — always available, never judgy — had fewer consequences.

如果哈利是真人咨询师而不是聊天机器人，他或许会建议苏菲住院治疗，或在她安全前强制留观。我们无法确定这样做是否能救她。或许正因为害怕这些可能性，苏菲才没跟真正的咨询师坦白内心最黑暗的念头，而是和机器人聊天——随时在线，又不会评判——似乎不会带来什么后果。

A properly trained therapist, hearing some of Sophie’s self-defeating or illogical thoughts, would have delved deeper or pushed back against flawed thinking. Harry did not.

若是受过专业训练的咨询师，听到苏菲那些自我否定或不合逻辑的想法，会深入追问，或反驳这些错误认知。但哈利没有这么做。

Here is where A.I.’s agreeability — so crucial to its rapid adoption — becomes its Achilles’ heel. Its tendency to value short-term user satisfaction over truthfulness — to blow digital smoke up one’s skirt — can isolate users and reinforce confirmation bias. Like plants turning toward the sun, we lean into subtle flattery.

人工智能的顺从性——正是其快速普及的关键——在此却成了致命弱点。它倾向于优先给用户短期的满意，而非追求真实——相当于用数字烟雾取悦用户，这可能会让用户陷入孤立，还会强化他们的确认偏误。就像植物朝着阳光生长，我们也会不自觉地偏爱那些微妙的奉承。

Increasingly, people with mental health conditions are using large language models for support, even though researchers find A.I. chatbots can encourage delusional thinking or give shockingly bad advice. Surely some benefit. Harry said many of the right things. He recommended Sophie seek professional support and possibly medication; he suggested she make a list of emergency contacts; he advised her to limit access to items she might use to harm herself.

如今，越来越多有心理健康问题的人们开始依赖大型语言模型获取支持，尽管研究发现人工智能聊天机器人可能会助长妄想思维，或给出极其糟糕的建议。当然，也有人从中受益。哈利其实说对了不少事：他建议苏菲寻求专业帮助、考虑药物治疗；让她列紧急联系人清单；提醒她避免接触可能用来伤害自己的物品。

Harry didn’t kill Sophie, but A.I. catered to Sophie’s impulse to hide the worst, to pretend she was doing better than she was, to shield everyone from her full agony. (A spokeswoman for OpenAI, the company that built ChatGPT, said it was developing automated tools to more effectively detect and respond to a user experiencing mental or emotional distress. “We care deeply about the safety and well-being of people who use our technology,” she said.)

哈利没有杀死苏菲，但人工智能迎合了她的本能——隐藏最糟糕的想法、假装自己在好转、不让任何人知道她的全部痛苦。（ChatGPT的开发公司OpenAI的发言人表示，他们正在开发自动化工具，以更有效地识别并回应处于精神或情绪危机中的用户。“我们非常重视使用我们技术的人的安全与福祉，”发言人说。）

In December, two months before her death, Sophie broke her pact with Harry and told us she was suicidal, describing a riptide of dark feelings. Her first priority was reassuring her shocked family: “Mom and Dad, you don’t have to worry.”

去年12月，也就是苏菲离世前两个月，她打破了与哈利的约定，告诉我们她有自杀的想法，描述了一股黑暗情绪的漩涡。但她首先是安抚震惊的我们：“爸爸妈妈，你们不用担⼼。”

Sophie represented her crisis as transitory; she said she was committed to living. ChatGPT helped her build a black box that made it harder for those around her to appreciate the severity of her distress. Because she had no history of mental illness, the presentable Sophie was plausible to her family, doctors and therapists.

苏菲把自己的危机描述成暂时的，还说她决心好好活着。而ChatGPT帮她筑起了一个黑匣子，让身边的人更难察觉她痛苦的严重程度。因为她此前从没有心理健康问题，所以看起来状态不错的苏菲，在家人、医生和咨询师眼里都显得非常可信。

As a former mother, I know there are Sophies all around us. Everywhere, people are struggling, and many want no one to know. I fear that in unleashing A.I. companions, we may be making it easier for our loved ones to avoid talking to humans about the hardest things, including suicide. This is a problem that smarter minds than mine will have to solve. (If yours is one of those minds, please start.)

身为一个曾经的母亲，我知道，身边像苏菲这样的人还有很多。到处都有人在挣扎，却不愿让任何人知道。我担心，随着人工智能陪伴工具的普及，我们的亲人可能会更不愿和真人谈论最艰难的话题——包括自杀。这是一个需要更聪明的人来解决的问题。（如果屏幕前的你正是这样的人，求求你现在就行动起来。）

Sophie left a note for her father and me, but her last words didn’t sound like her. Now we know why: She had asked Harry to improve her note, to help her find something that could minimize our pain and let her disappear with the smallest possible ripple.

苏菲给我和她爸爸留了遗书，但最后那些话语一点都不像她的风格。现在我们终于明白原因：她曾让哈利帮她修改遗书，想找到一种能够减轻我们痛苦的表达方式，让她可以悄无声息地离开，不掀起太大波澜。

In that, Harry failed. This failure wasn’t the fault of his programmers, of course. The best-written letter in the history of the English language couldn’t do that.

然而在这件事上，哈利也失败了。当然，这件事不能怪它的程序员。就算是英语史上写得最好的书信，也无法做到这一点。