AI Wasn't Built For You
这场 AI 革命，富人不再需要你
Only one income cohort rates AI positively in 2026 — Americans making over $200,000 a year. Below that line, sentiment flips. Scott Galloway sits down with Steven Bartlett to argue that the technology of the century has, in 18 months, become a luxury good with a marketing problem — and that the threat people should be afraid of isn't the robot apocalypse. It's a quiet life on a screen.
2026 年只有一个收入群体对 AI 持正面态度——年收入超过 20 万美元的美国人。这条线以下，情绪急转直下。Scott Galloway 做客 Steven Bartlett 的节目，给出一个不太客气的判断：这项被吹捧为"世纪技术"的东西，在 18 个月里已经变成了一件带营销问题的奢侈品。而真正应该让人警惕的，从来不是机器人末日——是被屏幕安抚的那种安静人生。
Two Brands Fell in 18 Months
18 个月，两个品牌一起塌了
Galloway opens with a paired observation that ends up structuring the whole conversation. America abroad, and AI at home, fell at roughly the same time — and they fell for the same reason.
Galloway 一上来抛出一对观察，这对观察后来撑起了整场对话的骨架：美国在海外、AI 在国内，几乎同时崩塌——而且是出于同一个原因。
The two fastest-falling brands of the past 18 months, Galloway argues, are the United States abroad and AI at home. For the first time in modern survey history, more people globally now believe China is a net force for good in the world than the United States. And per Quinnipiac's 2025 polling, the only income cohort with a net-positive view of AI is people making over $200,000 a year.
过去 18 个月里跌得最猛的两个品牌，Galloway 的判断是：海外的美国，国内的 AI。在现代民调史上第一次出现这种情况——全球认为中国是世界正面力量的人，已经多过了认为美国是的。再看 Quinnipiac 2025 年的数据：唯一对 AI 整体持正面态度的收入群体，就是年入 20 万美元以上那一档。
The two collapses share a mechanism. American power was supposed to be a public good. Technological progress was supposed to lift everyone. Both have been re-engineered, in the past 18 months, into instruments that benefit a thin sliver and exclude everyone else. People can feel it.
这两次崩塌共用同一个机制。美国的力量本来该是一种公共品；技术进步本来该把所有人往上拽一拽。但在过去这 18 个月里，它们都被悄悄改造成了一套只服务一小撮人、把其他人排除在外的装置。普通人能感觉得到这种变化。
If you're middle class, your view of AI is shaped by your electricity bill, by stocks you can't afford to buy, and by Sam Altman saying — when challenged on energy use — that you should stop complaining because raising a child uses energy too. That's not a tech debate. That's a class signal, received clearly.
如果你是中产阶级，你对 AI 的看法基本就是这几样东西塑造出来的：上涨的电费账单、你买不起的那些股票、以及 Sam Altman 在被人质疑能耗时那句让人血压升高的回应——别抱怨能耗了，养个孩子也耗能。这哪是技术讨论，这是一记清晰可辨的阶层信号。
"Your view of AI is directly correlated to your wealth."
— Scott Galloway
"你对 AI 的看法，跟你的财富直接挂钩。"
— Scott Galloway
When a CEO Says AI Will End Work, He Is Raising Money
当 CEO 说 AI 会终结工作时，他其实在融资
Every technology in history goes through the same arc, Galloway argues — some catastrophizing, some real job loss, then productivity gains, new business formation, and net employment growth. He sees no reason this cycle is different. The data so far backs him.
历史上每一项技术，Galloway 说，都走的是同一条弧线——一阵末日叙事、一些真实的失业、然后是生产力提升、新业务冒头、就业总量净增长。他看不出这一轮有什么不同，目前的数据也站在他这边。
4.4% Unemployment — Radiologist Jobs Are Up
失业率 4.4%——放射科医生岗位反而在涨
US unemployment sits near 4.4% (BLS, April 2026). Youth unemployment around 8.5% — slightly below the historical average. New business formation per capita has roughly doubled over the last decade. The job listings for radiologists — patient zero of "AI is taking your job" — were up in 2026. Job listings for coders are up year-over-year. Even Meta's layoffs come off a base that grew from 16,000 employees in 2019 to ~80,000 in 2025; trimming to 60,000 returns the company to 2024 staffing.
美国当前失业率约 4.4%（劳工部 2026 年 4 月数据）；青年失业率 8.5% 左右——还低于历史均值。过去十年人均新企业注册数大致翻番。被反复念叨"第一个会被 AI 干掉"的放射科医生岗位，2026 年需求反而是上升的。程序员需求同比也在涨。哪怕 Meta 那波裁员，基数是 2019 年 1.6 万人涨到 2025 年约 8 万人；裁回 6 万，不过是回到 2024 年的人员规模。
Catastrophizing Is the Second Leg of the Pitch
末日叙事是这门生意的第二条腿
The capital committed and the valuations granted to AI companies require one of two things in the next three years: either a trillion-plus dollars of new revenue from AI products (not materializing — there's no "AI moisturizer," no AI-engineered car), or massive cost savings from gutting the labor market. Catastrophizing isn't a warning. It's the investment thesis. CEOs need apocalypse math to justify the multiples already in the stock price.
这些 AI 公司已经融到的钱、已经给出的估值，未来三年必须靠两样东西之一兑现：要么市场冒出一万亿美元以上全新收入（目前看不到——没人见过"AI 面霜"，也没有 AI 设计的汽车），要么靠掏空劳动力市场挤出巨额成本节约。所以"末日叙事"不是警告——它是这门生意的投资逻辑。CEO 们需要末日数学，才能撑住股价里已经定价的那个倍数。
At 20% Youth Unemployment, France Had a Revolution
年轻人失业率到 20%，法国会革命
Bartlett presses: could you be wrong? Galloway: "100% — I'm wrong all the time." What would prove him wrong: a V-shaped jobs collapse so steep that even the eventual recovery comes too late. At 20% unemployment among young men, France had a revolution. Weimar Germany turned ugly. The risk isn't that AI replaces work. It's that the transition is sharp enough to break a society before the new equilibrium arrives.
Bartlett 追问：你会不会错？ Galloway 直接答："百分之百会，我经常错。"什么情况会证明他错：就业出现极陡的 V 字下挫，等回升时已经迟了。年轻男性失业率到 20%，法国会革命，魏玛德国变成那副样子。风险不是 AI 替代了工作——而是这场过渡太急太陡，社会还没等到新均衡就先崩了。
Where Jobs Actually Go
真正会消失的工作在哪儿
The hard cases are real. But the deeper pattern Bartlett surfaces — sharper than Galloway's framing — is that AI doesn't take one job per worker. It takes five.
那些真正会被冲击的岗位是真的。但 Bartlett 在对话里抛出了一个比 Galloway 更狠的角度：AI 不是从每个人手里抢走一个工作——它是抢走五个。
First in Line
第一波被替代的人
Long-haul truckers — the largest US employer of non-high-school-educated men — are gone within a decade in Galloway's view. Trucks can drive between 10pm and 4am with no traffic. Customer service: gone in the same wave. The political problem: this is the single biggest employer of working-class men in America.
长途卡车司机——美国非高中学历男性最大的雇主——按 Galloway 的判断，十年内基本会消失。卡车可以挑晚上十点到凌晨四点没车的时段跑。客服岗位也一起，同一波被冲掉。政治上的麻烦是：这是全美工人阶级男性最大的就业蓄水池之一。
A $20 LLM Did the Redlines
月费 $20 的 AI 改完了合同
Galloway estimates he'll cut his $300K annual legal bill by a third this year alone, just by asking Claude or ChatGPT to "pretend you're a $1,200-an-hour employment lawyer" and running the same redline through a second LLM. The result is somewhere between a senior associate and a name-brand partner — at near-zero cost.
Galloway 估算自己今年的法律开支（年均 30 万美元）可以一下砍掉三分之一，方法很简单：让 Claude 或 ChatGPT "假装你是个时薪 1200 美元的劳动法律师"，把合同改一遍；再换一个 LLM 把同一份再过一遍。最后出来的东西大致介于资深助理和大所合伙人之间——成本却几乎是零。
Molly Replaces Five
一个 Molly 顶五个人
Bartlett's investment-fund analyst, Molly, runs two Mac minis with agents. They screen deal flow proactively, run frameworks, score opportunities, and prep the IC pack. Bartlett: "We were going to hire five analysts. We only need Molly." The same logic compresses ten EAs into three. Same logic in junior law.
Bartlett 投资基金的分析师 Molly，桌上摆着两台 Mac mini，跑着 AI agent。它们主动筛项目流、跑分析框架、给机会打分、把决策会的材料整理好。Bartlett 自己说："本来要招五个分析师的，现在只需要 Molly。"同样的算法把十个高管助理压缩成三个，把初级律师团队也压缩成原来的几分之一。
The defense is straightforward, and Galloway gives it as a one-line rule: AI won't take your job. Someone who understands AI will. His prescription: keep a second screen open at all times — open to an LLM. Anything you receive digitally, port it to that screen. Make AI fluency the default register of how you work. The 60 days Bartlett spent retooling his hiring funnel against AI baselines is the example: someone you would have hired in February no longer clears the bar in May, because the AI alternative quietly raised it.
怎么防？Galloway 给了一句话原则：AI 不会抢走你的工作；懂 AI 的人才会。他的具体处方是：始终多开一个屏幕，专门跑一个 LLM。手上一切数字化的东西——合同、文档、邮件——都顺手丢进去玩一遍。把"用 AI 工作"变成默认状态，而不是额外功夫。Bartlett 自己用了 60 天重塑招聘漏斗，标准对齐 AI 的产出能力——结果是：二月还能进面试的候选人，到了五月已经不够看了。门槛是被 AI 悄悄抬高的。
"AI is not going to take your job. Someone who understands AI is going to take your job."
— Scott Galloway
"抢走你工作的不是 AI，而是会用 AI 的那个人。"
— Scott Galloway
The New Founder Skill Is Overpromise
新一代创始人的核心技能：会画饼
The CEO's job has inverted, Galloway argues. It used to be: underpromise and overdeliver. Now it is: overpromise and underdeliver, and create a vision so compelling that capital becomes cheap, so you can pull the future forward.
Galloway 说，CEO 的工作已经翻转过来了。过去那句老话是：少承诺、多兑现；现在反过来——把饼画得震天响、兑现远远跟不上，但只要愿景足够诱人，资本就便宜，你就能把"未来"提前透支到现在。
SpaceX is heading toward an IPO at a projected 90-110× revenues — analysts peg the targeted ~$1.75T valuation at roughly 95× trailing revenue. When Google IPO'd, it did so at 10× revenues — and was growing ten times faster. Tesla currently trades at hundreds of times earnings, while peer auto companies trade at 10-15×. The gap is what storytelling buys.
SpaceX 即将 IPO，目前传出的目标估值是约 1.75 万亿美元，对应过去 12 个月营收倍数大概 95 倍——也有分析师给出 90 到 110 倍这一区间。对照一下：Google 上市那年用的是 10 倍 PS，而且增速还是 SpaceX 现在的十倍。Tesla 目前市盈率是好几百倍；同行的传统车企不过 10 到 15 倍。中间这块巨大差额，就是"讲故事"换来的钱。
Galloway is careful to credit Musk where credit lands: SpaceX has launched the vast majority of all rockets that have left Earth in recent years. Starlink is, in his view, the best technology product of the past five years — only AirPods rivals it. The autopilot demo, the chopstick-catch of a returning rocket booster — these are real. The question isn't whether the magic exists. It's whether it scales to justify the multiples. For Tesla specifically, Galloway thinks the answer is no: capital that has been parked in TSLA is about to migrate to SpaceX at IPO, and TSLA will compress back to a fair auto-company multiple.
Galloway 该夸 Musk 的地方也夸：近几年地球发射出去的大部分火箭，是 SpaceX 干的；Starlink 在他眼里是过去五年最好的科技产品，能跟它打的只有 AirPods；Tesla 那段无干预的自动驾驶演示、火箭助推器被"筷子塔"接住——这些都是真本事。问题不在魔法存不存在，而在于这些魔法能不能撑起这种估值倍数。对 Tesla 这一只来说，他的判断是不行：原本停在 Tesla 上的那部分钱，IPO 一开就会往 SpaceX 流，Tesla 会被打回一个合理车企该有的倍数。
The broader point is that the narrative is now the product. Robots, Mars, AGI, autonomy — the function of these visions is to keep capital flowing into the parent company before the actual numbers can disappoint.
更深的一层是：现在叙事本身就是产品。人形机器人、火星、AGI、全自动驾驶——这些愿景真正在干的事，是在真实数字让人失望之前，先让资本继续往母公司里灌。
Sam Altman Isn't the Villain — But the Script Always Casts One
Sam Altman 不是反派——但剧本里必须有反派
There's a recurring three-act tech story, Galloway notes, and it's getting shorter each cycle. The same arc that broke Sheryl Sandberg, Mark Zuckerberg, the YouTube leadership. Altman is now midway through Act Two.
科技圈有一个反复上演的三幕剧，Galloway 说，而且每一轮都演得更快。把 Sheryl Sandberg、扎克伯格、YouTube 那一票高管打回原形的同一条弧线，现在套到了 Altman 身上——他正演着第二幕。
Act One: a soft-spoken founder is anointed as the good one — the one who cares, the one who'll build differently. Act Two: as competition intensifies and capital demands escalate, the founder makes the same incremental decisions every predecessor made — bend regulation, weaken oversight, ship the product, take the secondary. Act Three: the public realizes the founder is doing what every founder before them did, and recasts them as the villain. Sandberg ran this arc. Zuckerberg ran this arc. The YouTube leadership ran this arc. Altman is now midway through Act Two. Dario Amodei is currently being cast in Act One.
第一幕：一位说话温和的创始人被抬上神坛——是那个"真的有良心"、"会做得不一样"的人；第二幕：竞争收紧，融资压力堆上来，这位创始人开始做和所有前辈一样的渐进决策——绕过监管、削弱审查、把产品强推上线、套现一笔；第三幕：公众回过神来，发现这位"良心创始人"做的事跟之前每一位都没差，于是把他重新选作反派。Sandberg 走完了这条弧线，扎克伯格走完了，YouTube 那批高管走完了。Altman 现在正卡在第二幕中段。而 Dario Amodei，这一轮的第一幕，正在被分配给他。
Galloway's reframe: this isn't a character flaw. It's the role. Tech CEOs are paid to maximize earnings per share. They are not paid to be your friend, comfort you when you're old, or worry about teen-girl self-harm rates. Expecting them to is a category error. The institution that is paid to do those things is government — and we have been running an experiment in seeing what happens when the regulatory layer is gutted. We are seeing it.
Galloway 的视角调整是这样的：这不是个人品格问题，而是岗位职责。科技 CEO 拿的是"把每股收益拉到最大"的薪酬。他不是被请来做你朋友的，不是被请来在你老的时候陪你说话的，也不是被请来操心青少女自残率的。期待他们做这些事情，是搞错了部门。被付钱来做这些事的机构叫"政府"——而过去这几年，我们在做的实验恰好就是把监管这一层挖空看会发生什么。眼下看到的，就是答案。
"These guys would sleep with their cousin for a nickel. That's their job. Their job is to increase earnings. They're not here to save us."
— Scott Galloway
"为了一毛钱他们能跟自己表妹上床——这就是他们的工作。他们的活儿是把利润做大，不是来拯救你的。"
— Scott Galloway
When Your Boss Had a Slightly Bigger House. Now He Has a Bunker.
老板曾经只是房子大一点，现在他有逃生地堡
The most quietly radical claim of the conversation: a meaningful share of the people running the most consequential companies in the world have dissociated from their own countries.
整场对话里最不动声色、却最尖锐的一个判断：当今全球最具影响力的公司，掌舵的那批人里，有相当一部分已经跟自己的国家脱钩了。
The Go Plan Is Anxiety Theater, Not a Survival Plan
逃生计划是焦虑戏剧，不是真的求生方案
Galloway estimates one in three billionaires has a "go plan" — a worked-out scenario for catastrophic societal failure. Some are modest (a stocked basement). Some involve meeting pilots at Oakland airport and walking into a pre-built New Zealand bunker. His response to one: Boss, let's play this out. Society collapses. You don't think your pilots are going to kill you? You don't think the New Zealanders are coming for the rich guy's bunker? The plans are anxiety theater. They reveal one thing: they are no longer invested in this place.
Galloway 估计，亿万富翁里每三个就有一个备着"逃生计划"——专门为社会大规模崩溃准备的脚本。有的版本朴实（囤好物资的地下室），有的版本极致（在奥克兰机场碰上自己的飞行员，飞到新西兰的私人地堡）。他当时怼那位富豪：老板，把这一幕推演完——社会真崩了，你真觉得你的飞行员不会先把你做了？新西兰人不会找上你的地堡？这套方案不是求生——是焦虑戏剧。它泄露的只有一件事：他们已经不再把自己跟这个国家绑在一起了。
Dad's Boss Had a Cadillac. Today's Boss Flies Fractional Jets.
父亲老板开凯迪拉克。今天的老板坐分时私人飞机。
When Galloway was growing up, his dad's boss had a slightly nicer house and a Cadillac — they belonged to the same country club. Today the gap is qualitative, not quantitative. The 1% don't fly business — they fly fractional jets. They don't have insurance — they have concierge medicine with home visits and quarterly blood shipped to a Norwegian lab. Their kids' schools spend many multiples per child what public schools do. The result: they no longer experience the country they govern through capital. They don't queue at TSA, don't wait for healthcare, don't live near homelessness. They have stopped being invested in America's well-being.
Galloway 小时候，他爸的老板就是房子大一点、开个凯迪拉克，一家人加入同一个乡村俱乐部。今天这个差距是"质"上的差，不是"量"。1% 不坐商务舱——坐分时私人飞机；不买"保险"——用礼宾私人医疗，医生上门看诊，血液每季度寄到挪威实验室；孩子的学校人均花费是低收入社区公立学校的好几倍。结果：1% 不在 TSA 排队，不等病床，街区没有流浪汉。他们已经不再把"美国过得好不好"当成自己的事。
The Real AI Threat Isn't Unemployment. It's Loneliness.
AI 真正的威胁不是失业，是孤独
This is the conversation's central reveal, and Galloway is direct about it: the biggest danger of AI is not job destruction, not weaponization, not income inequality, not election contamination. The biggest danger is loneliness.
这是整场对话最重磅的一句反转，Galloway 没绕弯子：AI 最大的风险，不是岗位被消灭，不是被武器化，不是加剧贫富差距，也不是污染选举。最大的风险是孤独。
Discord for Friends. Robinhood for Income. Synthetic Porn for Romance.
Discord 交友，Robinhood 挣钱，合成色情代替恋爱
AI — combined with social media, gig-economy apps, and synthetic intimacy products — is teaching young people, especially young men, that they can experience a reasonable facsimile of life without leaving the screen. Friendship: Discord. Income: Robinhood. Romance: synthetic porn. Each substitute is just sufficient to discourage the higher-friction real version. None is sufficient to actually carry a life. The S&P 500, in Galloway's view, is an economy-wide effort to sequester you from your relationships and replace them with something subscribable.
AI 加上社交媒体、零工经济 App 和合成型亲密关系产品，正在教会年轻人——尤其是年轻男性——你不离开屏幕也能过出一种像模像样的人生。友情有 Discord，收入靠 Robinhood，恋爱用合成色情凑合。每一个替代品都刚刚够用，足够让你不想去碰那个有摩擦力的真实版本；但没有任何一个真的撑得起一段人生。标普 500 里相当大一块，在 Galloway 眼里干的就是同一件事：把你从人际关系里隔开，换上一个按月订阅的版本。
42–45% of Men 18–24 Have Never Asked a Woman Out in Person
42–45% 的 18-24 岁男性从未当面约过女生
Men aged 20-30 spend less time outside than prison inmates. More than four in ten men aged 18-24 have never asked a woman out in person — recent surveys put the figure between 42% and 45%. The skill collapsing fastest among young men is the ability to endure rejection — the one skill every entrepreneur, every founder, every person punching above their weight class actually runs on. Frictionless platforms produce frictionless people. They will struggle the moment friction returns.
20 到 30 岁男性，户外时间比监狱里的服刑人员还少。18 到 24 岁男性中，超过四成从未面对面约过女生——近期调研数字在 42% 到 45% 之间。年轻男性身上掉得最快的技能，是承受被拒绝的能力——恰恰是每一位创业者、每一位创始人、每一个"超水平发挥"的人赖以生存的底层技能。零摩擦的平台批量生产零摩擦的人，等现实里的摩擦回来那一刻，他们就站不稳了。
The Goal Is No. Train for Rejection Until It Stops Being a Crisis.
目标是被拒绝——练到"被拒"不再是一场危机
Galloway's mentorship prescription is concrete. He gets young men to a place where they're willing to make an overture — a friend invitation, a coffee ask — and tells them: the goal is no. Get the no. He calls them the next morning: Are you OK? Yes. Is the other person OK? Yes. Repeat until rejection stops being a crisis. That's the skill. The secret to his own success, he says, is rejection — accumulated over decades of asking for things, pitching, and not getting a yes.
Galloway 给年轻人开的处方很具体。他帮一个年轻人鼓起勇气开口——发起一次友情邀约，约杯咖啡——然后告诉他：这次的目标，是听到"不"。把那个"不"拿到手。第二天早上他打电话：你 OK？对方 OK？好。再来一次，直到"被拒绝"不再是危机。这就是那项技能。他自己成功的秘诀，他说，就是被拒绝——几十年来开口、推销、拿不到"是"，日积月累出来的。
"The secret to my success is rejection."
— Scott Galloway
"我成功的秘诀，是不停地被拒绝。"
— Scott Galloway
What To Actually Do
那到底该怎么办
Galloway's prescriptive register lands here, and it's surprisingly traditional. The enduring skills, in his view, are the ones AI commoditizes least: storytelling, relationships, and resilience.
到了给建议这一段，Galloway 的语调反而非常老派。在他眼里能扛得住时间的技能，恰恰是 AI 最难商品化的那几样：讲故事、人际关系、抗打击的韧性。
Storytelling Is the Durable Skill
最经得起时间的技能：会讲故事
The greatest CEOs of his generation are storytellers — Bezos's 1997 letter, Jensen Huang's keynotes, Karp's earnings calls. Technology converges; products converge; the differentiator becomes who you'd rather work with. Relationships outperform technical advantage on a long enough timeline.
他这一代最厉害的 CEO，本质上都是讲故事的人——贝佐斯 1997 年那封股东信、黄仁勋的开场主题演讲、Palantir Karp 的电话财报会。技术会趋同、产品会趋同，最后比的是"愿意跟谁合作"。时间拉长来看，关系永远跑赢技术优势。
The Honest Answer Is "Slowly"
变富的真实答案是"慢慢来"
"I'm not trying to get rich. I'm trying not to get poor." Diversify aggressively, never more than 3% of net worth in one position, move some capital out of the US. For young people: invest in yourself first; then automate savings so the money never touches your checking account. Cadence: $25/month as a teenager, $100 in your twenties, $500 then $1,000 as income grows.
"我现在不想变富，我想不要变穷。"分散到极致，单一仓位永远不超过净资产的 3%；把一部分资金分散到美国以外。对年轻人来说：先投资自己，然后把存钱这件事自动化，让那笔钱根本就到不了你活期账户。节奏大概是这样——青少年期每月 25 美元，二十几岁每月 100，收入起来后 500、再到 1000。
Find What Won't Pay You Back
找一件你这辈子收不回本的事
Galloway approached most of life as a transaction. The lesson he wishes he'd learned earlier: find the thing you can never get a real positive return on — that's your purpose. For him, it became kids. The math will never balance, and that's the point. The asymmetry is the feature.
Galloway 大半辈子把人生当生意做。他最希望自己早点学会的一课是：找到一件你永远不会赚回本的事，那才是你的人生意义。对他来说，是养孩子。账永远算不平——而"算不平"本身才是关键。这种不对称，正是它的本质。
The Receipts for Love Are Grief
爱的收据，是悲伤
Galloway's mother died decades ago and the grief hasn't faded — and he has stopped trying to make it. He hopes his sons grieve him the same way. Joy is the goal of a life; grief is its proof. He used to see it as a bug; now he sees it as a feature.
Galloway 的母亲已经去世几十年了，悲伤没退场过，他也不再试图让它退场。他希望自己的儿子将来用同样的方式悲伤他。喜悦是这辈子的目标；悲伤则是这辈子真的活过的证明。他过去把它当 bug，现在把它当 feature。
Galloway's life advice to a 30-year-old: nothing is ever as good or as bad as it appears in the moment. When things are going well, you didn't earn most of it — be humble. When they collapse, you didn't cause most of it — forgive yourself. Both are mostly luck and timing. People at the end of their lives don't regret the setback — they regret how upset they were about the setback. The mourning is the optional part. The setback is going to come either way.
Galloway 给 30 岁的人的人生建议：当下任何事的"好"和"坏"都没你以为的那么夸张。顺风顺水的时候，多数不是你挣来的——保持低头；崩盘的时候，多数也不是你害的——放过自己。运气和时机占了大头。人到了生命尽头，回头看不会后悔那一次挫败——他们会后悔的是"当时为那次挫败失态太久"。挫败本身躲不过，但难过多久这个开关，是在你手里的。
The conversation is, ultimately, less about AI than about what AI is exposing. AI didn't create the wealth gap, the loneliness epidemic, the regulatory vacuum, the storytelling-as-substance economy, or the dissociation of the 1%. It just turned up the contrast on all of them at once. The hopeful version: that's also why we can finally see them.
说到底，这场对话谈的其实不是 AI——谈的是 AI 把哪些早已存在的东西显影出来了：贫富差距、孤独流行病、监管真空、用"讲故事"代替"做事"的经济、还有 1% 与这个国家的脱钩。这些不是 AI 制造的，AI 只是同时把它们的对比度全调到了最高。乐观一点说：也正因为对比度调高了，我们才终于看得见它们。
Find the thing you can never get a real positive return on. That's your purpose.
找到一件你这辈子都收不回本的事——那就是你来这一趟的意义。