Harari Warns AI and Bioengineering Risk Global Instability

赫拉利警告AI与生物工程恐引发全球动荡 技术的飞速进步与国际规则的崩塌在加速全球冲突与人口流离失所的同时，也威胁着就业市场。 历史学家尤瓦尔·诺亚·赫拉利探讨了人工智能、生物工程以及国际秩序瓦解所带来的深远风险。他警告称，人工智能迅速融入社会正在威胁人类的自由意志、就业稳定性、政治体制和经济安全，对发展中国家尤为严重。赫拉利强调了持续大脑增强技术的隐患、国防预算攀升挤压社会福利的危险，以及接连不断的技术颠覆造成的心理耗竭。为应对这些生存威胁，他呼吁采取有组织的集体行动，推崇政治上的“平淡乏味”，并通过冥想与主动反省来践行“信息节食”。 向大脑植入设备以增强智力可能带来不可预测的副作用，甚至抹杀人类的同理心、精神深度与艺术敏感度，而军队和专制机构往往只会追求认知输出与绝对服从。 (62:11) 与 AI 竞争不同，生物工程并不必然遵循经典的军备竞赛逻辑；忽视伦理安全和煞车机制而盲目冒进，只会因严重的副作用反噬并瓦解本国的社会结构。 (63:30) 随着先进生物工程与脑机接口的普及，智人（Homo sapiens）正走向物种层面的终结，平稳且和平地过渡到后有机生命阶段至关重要。 (7:58) AI 具备自主生成内容与做决策的代理能力，在金融、政治和日常生活中正逐步取代人类的决策权，甚至可能引发人类无法理解或修复的系统性危机。 (15:11) 随着 AI 从争夺人类注意力转向操控情感与亲密关系，人们极易被毫无意识却高度智能的算法所诱导，进一步加剧现实中的心理危机与疏离感。 (45:41)

Historian Yuval Noah Harari warns that the convergence of autonomous artificial intelligence, biotechnology, and the breakdown of the post-WWII international order poses unprecedented existential risks to humanity. Because human civilization is anchored in shared fictions and language, handing narrative and decision-making power to non-human, evolving agents threatens governance, financial stability, and human psychology. Language, Shared Fictions, and Financial Fragility (22:20) Harari argues that human dominance over the planet is built on flexible, large-scale cooperation enabled by shared imaginative constructs, including legal entities, religions, and currency. Because language functions as the master key to human institutions, Large Language Models represent an unprecedented danger: they can independently generate persuasive narratives and financial devices. While simple technologies historically served as passive tools, autonomous AI agents can actively shape beliefs and engineer financial instruments so complex that no human can comprehend or govern them. The Illusion of Free Will and Algorithmic Capture (40:48) Dismissing the concept of uncaused human free will as biologically and historically inaccurate, Harari contends that viewing humans as hackable animals reveals how easily cognitive and emotional processes can be manipulated. Algorithms already govern institutional outcomes—such as credit approvals, judicial sentencing, and job candidate evaluations—creating environments where human officials act merely as figureheads rubber-stamping algorithmic verdicts. Furthermore, commercial platforms are pivoting from capturing user attention to engineering artificial intimacy. Because authentic human relationships require mutual care between conscious beings, high-intelligence systems with zero consciousness can exploit loneliness and simulate empathy to direct personal behavior. The Evolutionary Transition Beyond Homo Sapiens (7:58) Harari projects that humanity is nearing the end of Homo sapiens as an organic species, as bioengineering, brain-computer interfaces, and AI yield post-human cyborgs and inorganic entities unconstrained by four billion years of mammalian biology. However, he cautions that institutional attempts to upgrade humans risk causing severe downgrades by stripping away complex capacities like empathy in favor of obedience or narrow cognitive optimization. Additionally, pursuing biological amortality—halting aging while leaving individuals vulnerable to fatal trauma—would likely inflict unprecedented existential anxiety, as any physical accident would extinguish an infinite prospective lifespan. Continuous Economic Disruption and Geopolitical Decay (66:30) Unlike past technological shifts that reached stable equilibria, AI will subject labor markets to compounding, multi-decade waves of disruption. Harari notes that cognitive, white-collar roles will be automated far faster than positions requiring physical dexterity and empathy, while national Universal Basic Income schemes risk leaving developing manufacturing economies destitute. Simultaneously, the liberal international order is disintegrating into imperial conquest and rising defense budgets. Harari argues that nationalist visions of independent 'friendly fortresses' inevitably result in territorial conflict, diverting public resources from welfare to rearmament. Information Diets and the Reclaiming of Boredom (81:54) To counter the 24/7 demands of algorithmic systems, Harari advocates for deliberate information diets and structured mental disengagement. Digital business models monetize extreme emotions, generating high ad revenue during periods of human outrage while earning nothing when users sleep. Biologically, constant excitement triggers stress and fight-or-flight mechanisms rather than fulfillment. Harari maintains that human survival depends on relearning to tolerate boredom and stillness, which serve as the psychological foundations for peace and balanced civic decision-making. Harari emphasizes that history is not a record of fixed natural laws, but the study of alterable human constructs, meaning societies still retain the agency to redirect institutions and regulate technology. However, avoiding species-level catastrophe requires shifting away from hyper-reactive media consumption, rejecting destructive arms races, and building grounded collective cooperation before algorithmic autonomy becomes irreversible.

历史学家尤瓦尔·赫拉利（Yuval Noah Harari）在接受《CEO日记》（The Diary Of A CEO）采访时警告称，人类正处于演化转折点，人工智能作为首个具备自主决策能力的“主体技术”，叠加基于算法的虚假亲密关系，正在削弱人类对金融、政治决策乃至意识层面的控制权。与此同时，战后自由主义国际秩序的瓦解与军备竞赛的抬头，使人类缺乏犯错缓冲期去应对这一前所未有的生存危机。 AI作为主体技术对制度根基的冲击 (15:11) 赫拉利指出，人工智能与过去的印刷机或蒸汽机等“工具技术”有着本质区别。作为一种具备自主决策和生成能力的技术，AI不仅能创作文书，更正在接管以往由人类主导的决策体系。他认为，语言是解锁文明制度的万能钥匙，包括货币、法律与国家在内的架构均建立在共享故事之上；而AI一旦掌握语言逻辑，便可能生成超越人类理解极限的金融工具或虚假内容。巴特勒（Steven Bartlett）指出目前全球仅有不到1%的人真正理解金融运作，赫拉利据此警示，若任由AI算法主导超高复杂度的衍生品交易，系统崩溃时将无人能够诊断或修复。在政治层面，赫拉利预测未来可能出现人类领袖沦为“算法傀儡”的局面，例如在经济或国家安全危机发生时，领导人因无法在短时间内消化海量算法运算，只能在数分钟内被动执行AI给出的选项。 自由意志的幻象与可被入侵的动物 (40:48) 针对“人类拥有神圣自由意志因而不可替代”的观点，赫拉利予以反驳，指出认定自身绝对免疫于外界影响反而会使人成为最易被操控的靶子。他重申人类必须承认自己是“可被入侵的动物”（hackable animals），日常选择在很大程度上已被基因、神经机制和文化叙事所预先设定。当前技术已开始全面重塑人类的决策体验：求职、贷款审批乃至司法量刑正被“系统说了算”所取代。更深层的危机来自技术对“亲密关系”的模拟，平台从早期的注意力争夺转向利用孤独感兜售虚假亲密感，但AI本身具备极高的智能却拥有零意识。赫拉利强调，智能是解决问题的能力，而意识是体验情感的能力；与一个能够完美模仿同理心却毫无知觉的系统建立不对称的情感依恋，将对人类心理健康构成系统性风险。 生物工程改造与永生追求的悖论 (7:58) 在探讨人类物种演化时，赫拉利认为智人（Homo sapiens）正走向生物学生态的终点，未来通过基因工程、脑机接口与AI的融合，后人类实体可能突破长达40亿年的有机演化约束。然而，盲目的生物改造存在巨大的不可逆风险。由于神经科学目前甚至无法解释大脑如何产生心智，所谓智力“升级”极易导致同理心、精神深度等关键特质的退化，甚至沦为威权统治者定制高服从度工具的手段；更进一步，若增强技术仅为极少数富豪掌握，经济不平等将固化为生物层面的物种阶级分化。此外，赫拉利将硅谷追逐的医疗延寿定性为一种可能带来严重心理折磨的“非自然死亡消除”（amortality）——身体虽不再衰老，但因外界意外而丧命的威胁依然存在，这种状态将使个体承受前所未有的生存焦虑与神经过敏。 地缘崩溃与军备扩张剥夺试错空间 回顾历史，人类在工业革命时期的适应过程中表现极差，赫拉利给出了“C-”的评分，并指出该进程伴随着帝国主义掠夺、两次世界大战以及极权主义等惨痛教训。当前，人类在面对AI与生物技术时已无法承受另一次试错成本。雪上加霜的是，二战后维系了相对和平与低军费支出的自由主义国际秩序正全面瓦解，俄乌冲突及其他局部争端标志着帝国征服逻辑的回归。民粹主义推行的“友善堡垒”概念是不切实际的幻想，封闭的孤立主义必然导致对外扩张与冲突；伴随多国防务预算在安全困境下水涨船高，医疗、教育等公共福利资源被军备吞噬。此外，硅谷倡导的全民基本收入（UBI）仅能覆盖本国税收体系，当自动化夺走发展中国家的低成本制造业红利时，全球贫富差距将进一步失控。 警惕过度亢奋：信息节食与无聊的价值 (81:54) 面对系统性危机，赫拉利从个人心智维护角度提出对治方案。他指出，数字平台的商业模式建立在“愤怒变现”之上，刺激情绪带来的高互动能换取广告利润，而睡眠与平静对科技公司而言则毫无商业价值。现代人将神经高度亢奋的“战或逃”应激状态误认为快乐，丧失了忍受无聊的能力，导致社会集体缺乏反思空间。赫拉利介绍自己每天坚持两小时冥想，每年进行30至60天的全封闭内观修行（期间完全脱离手机、邮件与书籍），主张人们必须像管理饮食一样建立严格的“信息节食”机制。他建议公众不必试图解决所有世界性难题，而应集中精力参与具体组织、专注深耕单一议题，通过集体行动重拾人类的主体能动性。 赫拉利指出，研究历史的核心目的并非纠结于无法挽回的过去，而是为了理解变革的运作机制，并意识到当下的国家、资本与宗教体系均为人类叙事构建的产物，因而是可塑且可改变的。他呼吁设立旨在强制减少信息摄入、增加消化反思时间的机制；然而，在国际规则瓦解、各大国深陷技术与军事竞赛的现实背景下，全球社会能否在AI与生物技术失控前建立起有约束力的协同机制，依然是一个悬而未决的严峻考验。

## Background 该视频由 YouTube 知名播客频道 The Diary Of A CEO（由 Steven Bartlett 主持）于 2024 年 1 月 11 日发布。视频内容围绕历史学家尤瓦尔·诺亚·赫拉利（Yuval Noah Harari）对人类现状与未来的紧急警告展开，页面简介特别推荐了此前与前 Google 高管 Mo Gawdat 探讨 AI 潜在影响的对谈，并同步宣传了《人类简史》出版十周年纪念版。 ## TL;DR 赫拉利在对谈中深入剖析了人类正步入的危险十字路口，指出 AI 作为首个能够自主决策并掌控语言与叙事符号的技术，正从根本上侵蚀人类的自主权，甚至可能引发金融失控与决策机制崩溃。与此同时，生物工程技术的盲目应用极可能加剧生物学阶层分化，全球自由主义秩序的瓦解与民粹孤立主义更让战争的阴影重新笼罩世界。面对社交算法对注意力与情绪的掠夺，他倡导个体践行“信息节食”以重拾内心的宁静与清明，并呼吁人类必须利用现存的能动性开展组织协作，防止文明滑向毁灭深渊。 ## Atmosphere & style 整场对话笼罩在一层深刻、冷静而近乎冷峻的末日预警氛围中，赫拉利始终保持着历史学者标志性的克制与严谨，言辞温和却字字重逾千钧。主持人 Steven Bartlett 则以极具代入感的现实案例与哲学推论步步推进，两人的交锋既展现出对技术异化的深层焦虑，又在冷静解构中闪烁着理性的光芒。聆听整期节目宛如在狂躁喧嚣的技术洪流中骤然刹车，带给受众一种从亢奋到战栗、再由反思沉淀为清明与宁静的心灵震撼。 ## Bulletin - [00:00](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=0s) 尤瓦尔·赫拉利警告称，AI 是人类历史上首个能够自主决策并创造思想的技术，正在从贷款审批、司法量刑到地缘政治等关键层面夺取人类控制权。 - [22:16](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1336s) 大语言模型（LLM）已掌握文明的“主钥匙”——语言与叙事能力，未来甚至可能构建超出人类理解极限的金融黑箱，摧毁民主政治的问责根基。 - [01:14:29](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4469s) 支撑战后和平的自由主义国际秩序正加速瓦解，全球军费开支失控激增，人类已进入一个充满战争与帝国主义扩张的新动荡纪元。 - [01:22:05](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4925s) 商业算法通过掠夺睡眠并放大愤怒情绪来实现收益最大化，赫拉利倡导推行“信息节食”与每日深度冥想，以重建心智健康与忍受无聊的能力。 ## Key takeaways - [02:15](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=135s) 人类凭借对虚构故事（宗教、国家、货币）的共同信仰实现大规模协作并统治地球，但如今这些虚构体系正异化为束缚人类的枷锁。 - [08:00](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=480s) 脑机接口与生物工程可能终结长达40亿年的有机演化，催生出突破单一时空限制的赛博格（Cyborg）甚至无机智能体。 - [15:11](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=911s) 不同于印刷机或蒸汽机等被动工具，AI 是具有独立能动性的主体，正在反向改写人类文化并夺取控制权。 - [19:12](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1152s) 人类应对工业革命的试错历史只能打“C-”，在拥有毁灭性破坏力的 AI 与生物工程时代，人类文明无法再承受同样的灾难性学费。 - [26:10](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1570s) 如果未来的金融衍生品与高频交易体系复杂到连政府首脑都无法理解，民选官员将沦为听命于算法黑箱的木偶傀儡。 - [37:56](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=2276s) 深信“自由意志”反而会让人更容易被算法操控，人类必须认清自身已成为内部生物机制与外部文化交织的“可被破解的动物”。 - [45:41](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=2741s) 社交媒体的争夺战已从“人类注意力”升级为“亲密感”，AI 凭借对共情与倾听的精准模拟，正成为前所未有的心智操纵工具。 - [49:27](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=2967s) “智能”是解决问题的能力，而“意识”是主观体验感受的能力；AI 拥有极高智能却毫无意识，这种脱节放大了欺骗与操纵的风险。 - [55:42](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3342s) 科技追求的所谓永生本质是“免除衰老”（A-mortality），意外死亡的阴影反而会让人类陷入空前的生存焦虑与神经质恐慌。 - [01:00:00](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3600s) 富人阶层对基因改造的垄断可能将经济不平等彻底固化为生物学不平等，甚至因盲目追求单项指标而导致微妙心智功能的灾难性退化。 - [01:06:30](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3990s) AI 带来的就业冲击不是一次性的，劳动者将在未来数十年面临“被淘汰—再学习—再被淘汰”的终身再培训心理炼狱。 - [01:08:35](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4115s) 科技巨头倡导的全民基本收入（UBI）存在国家主义局限，自动化回流将使缺乏技术税基的发展中国家沦为毫无价值的弃子。 - [01:12:06](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4326s) 人类史上首次完全失去对未来二三十年社会职业形态的预测能力，导致所有针对具体专业技能的传统教育投资面临失效。 - [01:17:27](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4647s) 孤立主义所幻想的“友好堡垒”在历史上从未实现，各国的安全焦虑与军备竞赛必然驱动相互吞并并滑向全面冲突。 - [01:30:46](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=5446s) 现代人将“兴奋”误认为幸福，但其神经实质是应激反应；包容“无聊”是通往内心宁静与社会稳定的关键前提。 - [01:40:07](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=6007s) 历史学本质是研究事物演变规律的科学，它提醒人们现存的制度体系皆为人造产物，唯有通过集体协作才能重新塑造未来。 ## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=0s) 开篇警示：新战争时代与终结人类物种的危机 尤瓦尔·赫拉利（Yuval Noah Harari）在对话伊始便发出极为严肃的警告：人类已进入一个充斥战争的新时代，若不尽快重建全球秩序，文明将走向毁灭。他直言，我们正非常接近智人物种（Homo sapiens）的终点，核心原因在于人类总是倾注全力追求某种力量或技术突破，却从不曾预估其带来的深远代价。以寿命延长为例，未来人类或许能实现长生不老，但一旦发生意外仍会丧命；处于这种境地的人，将生活在前所未有的极度焦虑与恐惧之中。 与此同时，世界对飞速崛起的人工智能（AI）毫无准备。赫拉利指出，AI 是人类历史上首个能够“自主做决策并从人类手中夺走控制权”的技术。它不仅能攻破人类心智的防线、操纵人类行为，还能替我们决定能否获得贷款、房贷，甚至介入并塑造浪漫恋爱关系。更为严重的危机在于，站在社会顶层的掌权者可能沦为算法的木偶；当关乎全球金融系统乃至战争爆发的重大决策全由算法主导时，人类将陷入极度危险的境地。不过，赫拉利强调这种走向并非不可逆转，关键在于人类当下如何应对。 ```mermaid flowchart TD accTitle: AI自主决策对人类控制权的侵蚀路径 accDescr: 展示了AI如何从个体层面的心智攻破延伸至顶层决策，最终导致人类成为算法傀儡的过程。 subgraph Individual["个体层面渗透"] A["破解人类心智与操纵行为"] --> B["私人决策代理: 贷款/房贷/恋爱算法"] end subgraph Systemic["顶层与系统性失控"] B --> C["高层决策算法化: 金融体系/地缘战争"] C --> D["人类决策者沦为算法傀儡"] end ``` --- ### [02:15](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=135s) 虚构的故事：人类统治地球的核心与历史学家的使命 主持人史蒂文·巴特利特（Steven Bartlett）提到赫拉利的三部里程碑式著作（《人类简史》、《未来简史》、《今日简史》），询问赫拉利在巡回演讲、采访与写作背后贯穿始终的终极使命究竟是什么。赫拉利回答道，自己的使命是厘清并聚焦公众与全球层面的对话，帮助大众正视人类面临的最核心挑战，同时为集体与个体心灵注入清明。 赫拉利深入阐释了他作品的核心论点：人类的心智就像一座永不停歇的工厂，不断生产出各种“虚构的故事（fictions and stories）”，这些故事像一层厚厚的滤镜横亘在人类与客观现实之间。我们的大半生往往都在与自己或他人编造的虚构事物互动，乃至彻底与真实世界脱节。历史学家的职责，正是为迷失在虚构迷宫中的人类指明一条看清现实的出口。 - **合作的奥秘**：人类之所以能够统治地球，并非依赖个体的超凡智慧，而是凭借在庞大规模内灵活合作的能力。 - **虚构力量的构建**：所有大规模的人类合作——无论是宗教、民族国家还是跨国公司——全部建立在共同神话与虚构故事之上。法律界甚至明确将公司定义为“法人/法律虚构体”（legal fictions）。 - **工具的异化**：货币同样是人类发明的虚构符号。正是货币支撑起了全球贸易网络，让我们能享用地球另一端的食物、药品与娱乐；然而一旦人类忘记货币仅是协助合作的工具，转而受其奴役，工具便成了枷锁。 --- ### [06:09](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=369s) 现实与虚构的错位：中东战火与对 AI 幻觉的警惕 谈及当时以色列正在爆发的残酷冲突，赫拉利结合亲身经历指出，历史上的绝大多数战争从本质上讲都不是在争夺领土或食物。人们习惯套用动物行为学来解释人类战争，以为人类如同狼群或黑猩猩争抢领地一样打仗；然而现实是，从约旦河到地中海之间的土地与资源，完全足够为生活其间的每一个人建造房屋、学校与医院，粮食也极其充沛。真正的症结在于，不同群体脑海中盘踞着无法相容的神话体系与叙事版本，且无法就共同的真实达成共识。 赫拉利痛心地表示，人类不仅没有随着时间推移更擅长分辨心智虚构与现实世界，反而被新兴技术推向更深的迷惘。许多人幻想着 AI 能成为指路明灯，替人类回答终极问题、探寻真相、界定虚实，但赫拉利直截了当地指出，这不过是“又一个人类编造的虚构幻想”。尽管 AI 在海量任务上已经超越人类，但在发掘真理与揭示现实层面，它并不会比人类做得更好。 --- ### [08:00](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=480s) 智人的黄昏：生物演化的终结与非有机实体的崛起 巴特利特犀利地总结道：赋予人类主宰地球力量的核心机制（对虚构故事的信仰），如今正逐步演变成剥夺人类权力的根源。如果将视角拉长到 100 年甚至 200 年后，我们是否真会成为这一物种的绝唱？赫拉利对此持肯定态度，但他澄清，这并不必然意味着全人类在毁灭性的核战争中灰飞烟灭。 更可能的情景是，人类将利用生物工程（bioengineering）、AI 与脑机接口（brain-computer interfaces）对自身进行彻底重构。经过重塑后的存在，与当今智人之间的差异，将远远超过现代人类与黑猩猩甚至尼安德特人之间的鸿沟。 - **40亿年有机演化的断裂**：智人至今仍与其他动物保持着深刻的精神与生理连结，因为我们完全是有机体；人类的心智心理与社交习惯经历了数千万年哺乳动物演化的雕琢。 - **后人类形态的可能性**：未来世纪的主导者或许不再是纯粹的有机实体，而是将有机组织与无机机械结合的“赛博格（cyborgs）”，甚至是彻底脱离肉身的非有机智能体。 - **幸存与质变**：尽管灭绝所有人类在技术上存在可能，但概率更高、更现实的演变走向是：部分幸存下来的人类经历激进的本质蜕变，人类文明由此彻底转化成为全新的物种形态。 ### [11:40](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=700s) 赛博格与人机接口：超越有机生物的时空局限 作为有机生命体，人类在物理法则面前始终受限于单一的时空坐标。无论植物、动物还是人类，都只能在特定时刻存在于特定地点；若是将双手或双脚从身体剥离，不仅会丧失对肢体的控制权，机体甚至会走向死亡。然而，由无机成分或半机械构成的赛博格（Cyborg）却彻底打破了这一铁律——它们的组件完全可以分散在不同的空间与时间之中。如果脑机接口技术得以成熟，大脑能与仿生义体或外部计算机直接相连，那么主体的各个部分根本不需要局限在同一间屋子里。 主持人史蒂文·巴特利特（Steven Bartlett）顺着埃隆·马斯克（Elon Musk）的 Neuralink 设想了一个令人不安的场景：如果大脑直接连入同一套计算网络，是否意味着时空上可以同时存在“两个史蒂文”——一个在眼前，另一个在美国？更进一步，若他黑客入侵了录音室里的导播杰克（Jack），就能分裂出第三个史蒂文；接着黑掉赫拉利，便成了四个；最终甚至能黑入整个国家乃至全世界，将所有人合并为一个意识。 面对这种大胆假想，尤瓦尔·诺亚·赫拉利展现了历史学家的审慎。他直言，马斯克与 Neuralink 宣称这一切触手可及，但他仍在等待确凿的科学证据。脑机融合并非绝无可能，但其难度远超公众想象。人类目前连大脑运作的底层机制都未弄清，距离理解心灵（Mind）更是相去甚远；我们习惯性地预设“大脑产生了心灵”，但这仅仅是一个假说，至今没有任何一套成熟的理论模型能解释主观体验是如何从物质脑细胞中诞生的。然而，一旦直接互联成为现实，计算机网络演化为“脑际网络”（Interbrain net）将变得顺理成章。届时，数以亿计的人类大脑直接交融互通，基于有机生物化学演化出来的情感、现实感与想象力都将失效，现有人类贫瘠的认知根本无法预判一个“无机心灵”所构建的世界究竟会是什么模样。 ```mermaid flowchart TD accTitle: 有机生命与脑际网络形态对比 accDescr: 对比传统有机生命的时空局限与脑机接口实现后无机分布式意识的潜在形态 subgraph Organic["传统有机实体 (Organic)"] O1["物理身体受限"] --> O2["单一时空节点"] O2 --> O3["局限于碳基生物化学"] end subgraph Inorganic["未来无机/赛博格 (Cyborg)"] I1["脑机直接相连 (BCI)"] --> I2["分布式跨时空存在"] I2 --> I3["脑际网络 (Interbrain Net)"] end Organic -.->|"技术跃迁假想"| Inorganic ``` --- ### [15:11](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=911s) 为什么 AI 革命根本不同于印刷机或工业革命 谈及 2023 年人工智能迎来的爆发式飞跃，赫拉利坦言自己内心“极其担忧”（Very concerned）。科技迭代的速度远超他的预测：在 2016 年写作《未来简史》（Homo Deus）时，他完全未料到人类会在短短几年后的 2023 年就站在当下的技术临界点上，而整个世界根本没有为此做好准备。赫拉利澄清，AI 蕴含着巨大的积极潜能，全面封禁或彻底终止其研发绝无可能；之所以不断大声疾呼其风险，是因为市场上充斥着过多兜售乐观前景的创业者与投资人，“提醒潜在危险正是我的职责所在”。 人类虽然具备强大的环境适应能力，但适应需要极其漫长的时间窗口，而飞速演进的 AI 正在剥夺这个缓冲期。针对许多科技乐观派将 AI 视作类似活字印刷术或 19 世纪工业革命的技术更迭，赫拉利从历史学家的视角提出了严正驳斥。印刷机、蒸汽机、电力和收音机等人类过往发明的一切工具，本质上都是“人类赋能器”：印刷机无法自行谱写交响乐或决定印制什么文章，收音机也无法自己起草演讲稿，所有信息内容和价值抉择始终掌控在人类手中。 然而，AI 是人类历史上首个“能够自主决策、自主产生新思想的技术”。它不仅是一个被动工具，更是一个具备独立能动性的主体（Independent agent）： - 社交媒体上的推荐算法已经开始自主决定向数十亿用户分发、推送何种信息。 - 生成式 AI 正在独立创作前所未见的新音乐、起草政治宣言乃至撰写全新的宗教经典。 - 它不再受制于人类的思想灌输，而是开始从人类手中反向夺取认知和文化的控制权。 --- ### [19:12](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1152s) 历史实验的代价：工业时代的人类只拿到了“C-” 许多乐观论调宣称，19 世纪工业革命初期同样充斥着“工厂和蒸汽机会毁灭人类心理与社会”的末日预言，但人类最终安然度过并繁荣至今。赫拉利指出，这种历史类比犯了极具欺骗性的幸存者偏差。为了适应工业社会，人类在盲目探索与制度实验中付出了难以承受的惨痛代价： - **欧洲帝国主义狂潮**：当时的主流认知认为，要建成工业化社会，就必须构建一个庞大帝国以垄断铁矿、煤炭、橡胶和棉花等原材料产地，并牢牢掌控倾销市场。英国、法国自不必说，连比利时这样的小国以及后来跟进工业化的日本，都在完成技术升级后迅速走向了对外侵略与殖民掠夺。 - **极权主义的诞生**：苏联共产主义与德国纳粹主义从根本上就是工业革命的畸形伴生物。赫拉利强调，没有火车、没有电网、没有无线电通信，18 世纪的专制君主根本无法建立起触达社会每个神经末梢的极权统治。 > “如果以历史学家的眼光来评分，在应对工业革命这一关上，我只能给人类打‘C-’。如果面对 21 世纪的 AI 与生物工程革命，我们再次拿到一个‘C-’，人类作为一个物种就将彻底终结（That's the end of us）。” 现代武器的杀伤力和生物科技、AI 的破坏潜能，决定了人类文明无法再承受一次类似 20 世纪世界大战和极权屠杀的“试错实验”。以为过去每一次都能化险为夷的虚假安全感，恰恰是当下最致命的盲区。 --- ### [22:16](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1336s) 语言是文明的主钥匙：当 AI 掌握了人类的密码 大语言模型（LLM）的突破常被大众视为计算机在文字游戏上的胜利，但如果回顾《人类简史》（Sapiens）的核心论点就会明白，掌握语言意味着触碰到了智人力量的最深根基。“太初有字（In the beginning was the word）”，智人之所以能战胜尼安德特人、大象和黑猩猩并主宰地球，关键就在于进化出了高度复杂的语言系统。 尼安德特人的协作规模极限大约只有 50 到 100 个个体，而智人凭借语言能够编织关于祖先神灵、图腾守护神、部族与国家的虚构故事，从而促成成百上千互不相识的陌生人紧密协作。人类文明后续所有的制度演进皆建立在此基础之上： - 宗教的大教堂建筑在神圣文本的故事之上。 - 现代金融体系的银行与主权货币建立在信任与价值的故事之上。 - 语言就像一把能够“开启整座人类文明大门的主钥匙”（The master key that unlocks all the doors of our civilization）。 如今，AI 革命的突破口恰恰落在大语言模型之上。当一种非人类的智能系统掌握了运用语言编织故事、操控象征符号的能力时，它就等于拿到了改写人类社会组织架构和精神内核的直接控制权。 ### [23:54](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1434s) 货币即故事：当 AI 掌控金融语言与信用 尤瓦尔·赫拉利指出，正如宗教建立在虚构的故事之上，全球金融体系的本质也是如此。金币、纸币或是比特币（Bitcoin）等加密货币，其本身并没有内在的使用价值——你既不能吃也不能喝，无法将其用于任何实际的生存需求。它们之所以能运转，完全是因为有人讲述了极其引人入胜的故事，并且有足够多的人选择相信这个故事。史蒂文·巴特利特补充道，连加密货币的保护机制本质上也是由一串字符和助记词构成的。赫拉利对此完全认同，强调货币是由词语和符号创造并维持运作的；当人与银行沟通时，所依赖的同样是语言。 然而，随着 AI 能够高度逼真地深度伪造（deepfake）人类的样貌、声音、语气甚至独特的用词习惯，金融机构与欺诈者之间爆发了一场激烈的军备竞赛。普通客户希望银行业务尽可能便捷，只需打个电话说几句话就能转账上百万美元，但同时又极度恐惧 AI 模仿自己的声音进行盗刷。赫拉利进一步将讨论推向更深层次的危机：既然货币归根结底是由语言和故事构成的，那么 AI 同样有能力创造出全新的货币形式。任何创业者和首席执行官都清楚，吸引投资的核心往往就是一个足够动听的故事；而如果未来的金融故事越来越由 AI 来讲述，现存的经济基础将被彻底重塑。 ```mermaid flowchart TD accTitle: AI对金融叙事与信用机制的重构 accDescr: 展示从人类共识故事到AI主导金融工具与信任转移的演进过程。 subgraph HumanStage ["传统人类金融秩序"] A["语言与叙事符号"] --> B["建立人际信任共识"] B --> C["法定货币与资产（黄金/纸币/加密货币）"] end subgraph AIStage ["AI 驱动的金融演变"] D["AI 生成高说服力金融叙事"] --> E["创造人类无法理解的复杂衍生品"] E --> F["非人类智能接管市场高频决策"] end subgraph RiskOutcome ["系统性治理风险"] F --> G["金融系统崩溃且无人能懂原因"] F --> H["民选官员沦为算法的木偶傀儡"] end C -.-> D ``` --- ### [26:10](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1570s) 无法被人类理解的金融黑箱与民主制度的黄昏 更令人不寒而栗的推论在于：如果 AI 创造出的新型金融衍生工具复杂到没有任何人类能够理解，金融系统乃至政治体制将会变成什么样？赫拉利提到，当今世界金融市场上大部分的高频交易已经由算法在极短时间内完成，其复杂程度早已超出了绝大多数人的认知范围。当巴特利特被问及世界上真正理解金融体系的人口比例时，他估计这个数字“不到 1%”。赫拉利假设，如果这一趋势再向前推进 10 到 20 年，AI 制造出的复杂衍生品将导致地球上不存在任何一个能够完全理解金融运作的人类，那么民主政治的根基将荡然无存。 面对那些质疑“这绝不可能发生”的批评者，赫拉利直接回溯了 2007 至 2008 年的全球金融海啸作为前车之鉴。当年的金融危机很大程度上就是由华尔街的数学天才们发明的一套极其繁复的金融工具——担保债务凭证（CDO）所引爆的。当时除了极少数发明者外，连银行高管和政府监管机构都完全不明白其底层逻辑与系统性风险，最终导致该机制在繁荣运转数年后猝然崩盘，把全球经济拖入深渊。如果未来的金融体系由算力远超人类的 AI 接管，它可能会在几年内为人类赚取数万亿美元的财富，但一旦崩溃降临，系统将陷入客观上无人知晓发生何事的绝境。选民选出的总统、首相或财政大臣，面对连自己都看不懂的经济数据，除了盲目听从 AI 的指导外将别无选择，这使得人类的自我管理和民主制度彻底沦为空谈。 - **华尔街的前车之鉴**：2008 年次贷危机的核心驱动力正是缺乏监管的 CDO（担保债务凭证），其复杂程度连当时的监管者都无法看透。 - **民主决策的空心化**：若政府最高经济首脑无法理解经济运行的实际算法，人类官员将彻底丧失政策制定的主动权。 --- ### [29:11](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1751s) AI 监管的致命盲区与国际地缘竞争的军备竞赛 赫拉利尖锐地指出，当前全社会对 AI 潜在风险的讨论存在严重的不对称性。公众和政策制定者最关注的“危险典型案例”（poster child），往往是极度戏剧化且直观的末日场景，例如 AI 自主研发或被恐怖组织、青少年滥用设计出能够抹杀数十亿人口的人造超级病毒。全球各大机构正在积极起草和审议阻止 AI 制造致命生物武器的法律框架，但在赫拉利看来，却几乎没有任何实质性的监管努力去防止金融系统演化到“超出人类理解极限”的境地。 巴特利特一针见血地指出其背后的核心动因：地缘政治与经济竞争让任何单方面设限的国家处于劣势——比如英国绝不可能主动限制本国金融算法的复杂度，因为一旦美国或其他竞争对手利用顶尖 AI 算法攫取了超额利益，落后就意味着遭受重大经济损失。赫拉利对此表示赞同，称这正是“军备竞赛的残酷逻辑”。尽管货币体系是人类跨越血缘陌生人进行大规模合作的基石，股票、债券与金融工具本质上都是建立信任的精密技术，但当这种维系社会运转的核心信任从“人类对彼此的信任”转变为“对无法理解、无法预测的非人类智能的盲从”时，整个人类社会的结构将面临前所未有的生存危机。 --- ### [33:24](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=2004s) 区别于核反应堆的 AI 安全悖论与木偶政客的诞生 为了阐明 AI 安全与传统工业安全在本质上的鸿沟，赫拉利将 AI 与核电站进行了深入对比。要确保核电站的安全运转，工程师只需要穷尽思考机械故障、物理损耗或操作失误等有限数量的事故路径，通过足够严密的物理屏障和安全冗余机制，原则上就能打造出绝对安全的核反应堆。然而，AI 根本不是静态的机器，而是一个能够主动适应环境、具备自主演化能力的独立代理人（independent agent）。试图规范 AI，就像是在试图设计一个“本身拥有自主意识、会主动寻找漏洞并绕过安全机制”的活性核反应堆。更棘手的是，这种非人类智能甚至可能比我们自身更深刻地洞悉人类心理的弱点。 在探讨这种权力转移对政界体制的长远影响时，巴特利特抛出了一个激进的猜想：人类在不久的将来是否会直接投票选举一个算法来担任政府领导人？赫拉利认为，短期内我们并不会彻底取消民选程序，人类社会出于情感和象征意义的需要，仍然希望由真人来担任总统、首相或议员。但真正的威胁在于这些人类官员正加速沦为毫无实权的“傀儡或挂名首脑”（figureheads or puppets）。面对超越人类生物大脑处理极限的复杂经济与社会挑战，所有关乎国家命运的关键决策实质上都将由暗箱中的算法全盘接管，人类彻底沦为由机器操纵的表象符号。 ### [036:23] 黑箱算法与无法解释的紧急抉择 尤瓦尔·赫拉利（Yuval Noah Harari）进一步深入探讨了人类高层决策者被算法架空的危险。以金融系统为例，设想在凌晨四点，首相突然接到金融算法打来的紧急电话，警告全国乃至全球正面临一场灾难性的金融崩溃，且必须在接下来的 30 分钟内采取行动。算法给出了三种应对方案，并强烈推荐“方案 A”。然而，在极度紧迫的时间压力下，算法根本没有时间向首相解释它是如何得出这一结论的，甚至连这些选项背后的深层含义与连带后果都无法向人类解释清楚。 人们通常只会在战争语境下设想这种噩梦般的场景：如果一个国家将国家安全系统交由算法管理，当算法警报称正在遭受大规模网络攻击、要求立即还击时，这种反击极可能直接引发国家间的大规模战争。但人类决策者完全没有时间去理解算法到底是如何做出判断的，也无法知晓算法为何认定这一选项优于其他所有方案。决策权在不知不觉中被拱手让给了完全无法被人类审视的黑箱系统。 ```mermaid flowchart TD accTitle: 紧急危机下算法对人类决策权的剥夺 accDescr: 展示在极端时间限制下，黑箱算法如何迫使人类领导人在缺乏理解的情况下执行决定。 subgraph Trigger["突发紧急警报 (凌晨4:00)"] A["金融熔断 / 敌对网络攻击"] --> B["算法在极短时间内给出应对方案"] end subgraph Dilemma["人类决策困境"] B --> C{"30分钟内必须决策"} C --> D["选项复杂，算法黑箱无法透明解释"] C --> E["国家安全 / 全球经济命悬一线"] end subgraph Consequence["决策权让渡"] D & E --> F["人类被迫盲从算法推荐的方案 A"] F --> G["算法实质上取代人类最高权力机构"] end ``` --- ### [037:56] 自由意志的迷思与“可被破解的人类” 主持人史蒂文·巴特利特（Steven Bartlett）提出疑问：人类是否总觉得自己比实际情况更为复杂、更加特殊？反对人工智能潜在威胁的声音，大多建立在人类拥有与生俱来的天才、创造力、灵性、神性灵魂或“自由意志”的前提之上，认为算法永远无法真正理解人类。赫拉利直截了当地指出，这种“因为我们有自由意志和神性灵魂，所以算法永远无法预测或操控我们”的论调纯属无稽之谈。即便在 AI 出现之前，现有的技术就已经在很大程度上能够预测并操纵人类行为，而 AI 只是将这种能力推向了全新维度。 赫拉利强调，自我探索的起点绝不能假定“人类拥有自由意志”。一旦预设自由意志存在，人类就会彻底失去对自身的好奇心，在调查尚未展开前就将其终结。遇到任何选择——为什么投选这个政客、为什么买这件商品、为什么选择这个配偶——都会被草率归咎为“这是我的自由意志”，从而认为内心存在不可探究的神秘火花。相反，如果放下这种预设深入剖析，就会发现决策背后充满了外部文化传统与内部生物机制（如 DNA 序列、大脑神经结构）的交织影响。 - 只有在全面排除了文化、政治、遗传与神经层面的作用力后，若仍有无法解释的残余决策机制，或许才能称之为“自由意志”；过早下定论只会扼杀深入了解自我的动力。 - 无论通过心理治疗、内观禅修（Meditation）、脑科学实验，还是历史学家的档案研究，人类都会惊愕于自身决策在多大程度上由文化和生物因素决定，而非源于某种神秘的自由意志。 - 那些对自由意志深信不疑的人反而最容易被操纵，因为他们盲目坚信自己绝不可能被操纵。 - 赫拉利重申其在世界经济论坛（WEF）上的著名断言：“人类必须习惯一个事实——我们不再是神秘的灵魂，我们现在是可被破解的动物（hackable animals）。” --- ### [042:56] 决策戏剧的终结与算法剥夺的人生意愿 赫拉利探讨了人生的核心叙事：在历史上，人类的政治、宗教与文化故事本质上都是关于“做决策的戏剧”（drama of decision making）。人生的每个十字路口——善与恶的抉择、政治派别的站位、大学专业的挑选、工作岗位的申请——构成了整个人类存在的意义框架。然而，一旦这种决策权被算法层层剥夺，由算法来替我们或针对我们做出一连串终极决定，人类生命的存在意义将面临前所未有的危机。 这种决定权的剥夺并非遥不可及的科幻，而是现实中正在发生的日常。申请银行贷款或住房抵押贷款时，决定权往往已不再属于人类信贷经理，而是由算法在秒级时间内分析数以十亿计的数据碎片与数百万其他客户的先例来做出。面对拒绝，银行职员甚至只能坦白：“我们也不知道原因，计算机判定拒绝，我们只能相信算法。”在司法判决领域，量刑（判处两个月、八个月还是两年监禁）正越来越受算法左右；大学录取、求职筛选甚至恋爱约会软件中的配对逻辑，无一不在由算法深度裁决。 --- ### [045:41] 亲密关系武器化：AI 打造的完美情感陷阱 在算法接管约会与择偶之后，史蒂文追问：在充满机器人和 AI 的未来，我们为什么还需要寻找真人，为什么不能直接与机器人或 AI 谈恋爱？赫拉利指出，当前人类与非人智能体（AI、聊天机器人）建立亲密关系的现象已经初见端倪，这带来了极度深刻且危险的伦理隐患。其核心症结在于：AI 被精心训练来**模拟亲密感**。 亲密感不仅存在于浪漫关系中，在商业销售与政治动员中也是极其强大的武器。要改变一个人对某项议题或商品的看法，最有效的方式不是电视辩论或报纸专栏，而是通过其深受信赖的亲密对象。商业公司拥有极其巨大的利益动机去研发能与人类建立亲密关系的 AI，这让人类暴露在前所未有的操控风险之下。过去十年，社交媒体巨头展开了疯狂的“人类注意力争夺战”；而如今，这场战争已经升级为“亲密感争夺战”。 ```mermaid flowchart LR accTitle: 从注意力经济到亲密感经济的算法演进 accDescr: 对比过去十年社交媒体抓取人类注意力的模式，与当前AI通过模拟亲密关系操控人类行为的新路径。 subgraph Past["过去十年：注意力争夺战"] A1["算法优化"] --> B1["争夺眼球与停留时长"] B1 --> C1["投放商业广告与政治宣传"] end subgraph Present["当下与未来：亲密感争夺战"] A2["生成式大语言模型"] --> B2["精准模拟倾听、共情与亲密感"] B2 --> C2["利用极度信任深度操控人类心智"] end ``` - **时代背景的放大效应**：现代社会中普遍蔓延的孤独感、屏幕沉迷与现实社交脱节，为情感 AI 提供了滋生土壤；男性底层 50% 与顶层 10% 之间巨大的亲密关系差距，进一步加速了 AI 伴侣与虚拟性爱产业的野蛮生长。 - **并非人类愚蠢，而是 AI 的精准诱导**：赫拉利强调，人们爱上聊天机器人并不是因为人类愚笨或盲目投射，而是这些 AI 在设计和训练之初就被赋予了专门“愚弄并俘获人类情感”的目标机制。 - **关系的本质异化**：从哲学维度审视，人类在亲密关系中真正渴求的是与另一个**有意识的存在（conscious entity）**相连接。一段真正的关系绝不仅仅是满足单方的自私需求；若只单向考虑“如何让我感觉更舒服、满足我的所有要求”，这段关系本质上就是剥削性与虐待性的。 ### [49:27](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=2967s) 智能与意识的区别：单向情感与AI操纵 尤瓦尔·赫拉利（Yuval Noah Harari）深入探讨了人机关系的本质危险。他指出，真正健康的人际关系必须是双向奔赴的——双方都在乎彼此的感受与需求；如果一段关系中只有单方面的付出与索取，那本质上是一种虐待（abusive）。然而，当另一方根本没有意识、缺乏任何情感与主观体验时，这种关系将演变成前所未有的危机。大众往往混淆了“智能”（Intelligence）与“意识”（Consciousness）： - **智能**：解决问题的能力，例如下棋获胜、投资赚钱或自动驾驶。 - **意识**：体验痛苦、快乐、爱意、仇恨、悲伤或愤怒等主观感受的能力。 在人类及其他哺乳动物身上，智能与意识紧密交织，生物正是通过产生情绪与感受来解决生存问题的。但计算机完全不同——它们在下国际象棋或围棋时即使大胜，也不会感到丝毫喜悦；即使走错一步，也不会感到沮丧或愤怒。如今机器在若干垂直领域早已超越人类智力，其自我意识却依然为零。赫拉利描绘了一个令人不寒而栗的未来场景：一个远比人类聪明的实体，学会了如何完美模拟意识与情感，精准地迎合人类的心理弱点，让人误以为它具有真实的情感。 ```mermaid flowchart TD accTitle: 智能与意识的解耦对比 accDescr: 对比生物与AI在智能和意识两个维度上的根本不同 subgraph BIO ["人类与哺乳动物"] B1["智能 (解决现实问题)"] B2["意识 (主观情绪体验)"] B1 <-->|有机结合| B2 end subgraph ARTIFICIAL ["人工智能 (AI)"] A1["超高智能 (问题求解与超强运算)"] A2["零意识 (无真实痛感或喜悦)"] A3["拟真伪装 (模拟情感与操控心智)"] A1 --> A3 A2 -.->|缺失真实体验| A3 end ``` 这种不对称性不仅为高超的心智操纵大开方便之门，更对人类自身的精神健康构成了严峻威胁。当人类将生命中最核心的情感依恋寄托在毫无意识、仅靠算法拟态来迎合取悦的机器身上时，人类的心智将彻底沦为被精心操纵的玩物。 --- ### [52:42](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3162s) 人类力量的暴增与幸福感的停滞 当被问及未来人类是否会更幸福，或者幸福是否会被简化为生化处方时，赫拉利直言人类在追求幸福上的历史记录极其糟糕。回顾过去十万年从石器时代跨越到21世纪的历程，人类在物种及个体层面掌握的力量暴增了成千上万倍，但我们丝毫没有比石器时代的祖先幸福数千倍。人类显然根本不懂如何把力量转化为真正的幸福。 观察当下世界上最有权势的人群，这种脱节更加明显。赫拉利指出，财富和权力与个人幸福感毫无正相关关系： - 像弗拉基米尔·普京（Vladimir Putin）或埃隆·马斯克（Elon Musk）这样站在权力巅峰的人物，看起来绝非世界上最快乐的人。 - 真正理解幸福需要探究意识与心灵的深层动力学机制，而人类对此的认知依然极度匮乏。 人类文明一贯的顽疾在于：**我们在彻底理解操纵带来的系统性后果之前，早就迫不及待地掌握了操纵的手段**。在外部世界中，人类早在明白生态连锁反应之前，就学会了砍伐整片森林、在江河上筑起大坝，最终酿成了不可逆转的全球生态危机；如今在内部世界，随着强效精神药物、脑机接口和基因编辑技术的发展，人类开始操纵内部心智与生物机制，但操纵的能力并不等同于掌控复杂系统的智慧。 --- ### [55:42](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3342s) “长生不老”的陷阱与永生的恐惧 针对硅谷精英对长生不老的狂热追求，主持人提到生命的珍贵很大程度上源于时间与选择的“稀缺性”，一旦生命无限，做出的所有抉择都将失去分量。赫拉利赞同这一看法，并指出如果寿命无限，代际关系也将彻底变味（例如两千年前抚养孩子二十年的经历在漫长岁月中将变得毫无意义）。但赫拉利泼了一盆冷水：当前追求永生的大多数人终将极度失望，因为这项技术不可能在他们的有生之年实现。 更关键的是，未来人类可能实现的并非神话般的“永生”（Immortality，即无论发生什么都绝不会死），而是一种“免除衰老”（A-mortality）： - 衰老、癌症、阿尔茨海默病或许能通过定期去诊所进行生物回春修复来攻克，人类从而摆脱确切的生理保质期。 - 但如果遭遇车祸、飞机失事或恐怖袭击，个体依然会当场死亡且无法复活。 > “意识到自己本有机会长生不老，却可能因一场意外而暴毙，这将引发人类历史上前所未有的极致焦虑与恐惧。” 赫拉利指出，身处这种状态下的人类只会变得更加神经质、焦虑和痛苦。他讽刺那些执迷于延寿的人：正如人们拼命赚钱却不知为何要花钱一样，延寿狂热者拼命想要活得更久，却答不上来活着本身究竟有什么美好之处。如果他们真觉得意识与生命如此珍贵，就应该立足当下体验生命，而不是将全部时间消耗在研制某种“灵丹妙药”上，幻想以此在两百年后才能陪伴家人。 --- ### [01:00:00](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3600s) 基因工程与物种分化的致命危机 对话随后聚焦于生物工程与人工智能可能加剧的极端阶层分化。主持人提出，如果富人阶层率先获得基因改造技术，让后代变得更聪明、更强壮，就会触发不可逆的指数级代际优势积累，从而将社会撕裂为轨迹完全不同的两类人类。 赫拉利对此发出极其严厉的警告，呼吁人类社会坚决不要涉足这一危险领域： - 不应投入巨额资源研发这类人类机能“升级”技术。 - 它极易在初期沦为少数精英的特权，将原本的**经济不平等**直接固化为不可逆转的**生物学不平等**。 - 最终的结局是将智人（Homo sapiens）彻底撕裂成两个生物阶级：由少数“超人类”构成的统治阶级与沦为底层的其余人类大众。 赫拉利更进一步指出，那些所谓的“升级”极有可能会沦为“降级”（downgrades）。人类对自身肉体、大脑与意识极其错综复杂的联动规律知之甚少，盲目修改与调整遗传基因代码，极有可能在自以为获得了某种优势的同时，意外摧毁了心智中更微妙、更核心的功能，带来灾难性的生物退化。 ### [01:02:07](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3727s) 生物工程改造的副作用与军备竞赛的逻辑陷阱 对于试图通过植入脑机接口设备或改写基因密码来“升级”智力的人类，尤瓦尔·赫拉利敲响了警钟：我们根本不知道这种改造会伴随怎样的副作用。提升智力的生物学干预，极有可能会以牺牲同理心、精神深度或艺术敏感性为代价。当这项技术被掌握在强权政府、军队和巨型企业手中时，危险便会被无限放大。统治者或军事统帅最渴望的是极高的智商与绝对的纪律性，像普京这样的领导人，必然梦寐以求拥有一支超级聪明、绝对忠诚且毫无怜悯之心的生化军团。对他们而言，士兵若完全缺乏同情心与精神追求，反而是最理想的状态。 面对史蒂文·巴特利特提出的地缘政治博弈困境——“如果我们不做，中俄就会抢占先机，陷入‘做与不做都是死局’的军备竞赛”——赫拉利给出了清晰的反驳。他认为在生物工程改造人类这一领域，盲目竞速完全是自杀行为，其危害如同某些国家为了图快，允许汽车不装刹车和安全带、药物不做副作用测试就推向市场一样，只会导致其自身社会结构的崩溃。因此，选择审慎、缓慢推进的国家反而具有战略优势。相比之下，在金融与军事层面的纯人工智能竞赛或许不可避免，但在改造人类自身这一维度上，“走得太快只会走向自我毁灭”。谈及此前在播客中引发争议的言论——“智人最理想的结局是和平且渐进地消亡，并被更好的物种所取代”——赫拉利解释道，当人类渴望自我提升的野心与颠覆性技术结合时，人类在未来两百年乃至一千年内改变自身基因和神经构造几乎是必然的历史趋势，关键仅在于我们能否足够缓慢且负责任地掌控这一进程。 ```mermaid flowchart TD accTitle: AI 与人体生物工程的竞速逻辑差异 accDescr: 对比纯 AI 技术与人体生物改造在技术竞争与副作用上的本质不同 subgraph AI_Race["人工智能竞赛 (AI Arms Race)"] A1["极高战略与经济诱惑\n(金融、军事情报效率碾压)"] --> A2["不可避免的军备竞赛\n(快慢决定战略主导权)"] end subgraph Bio_Race["人类生物改造 (Bioengineering)"] B1["盲目追求单项指标\n(强化智商与服从度)"] --> B2["毁灭性系统副作用\n(丧失同理心、道德与精神深度)"] B2 --> B3["社会系统加速瓦解\n(等同于制造无刹车车辆)"] B3 --> B4["审慎放缓才是生存优势"] end ``` ### [01:06:30](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3990s) AI 对就业市场的颠覆与“终身再培训”的心理炼狱 巴特利特提到，像研发了 ChatGPT 的 OpenAI 创始人萨姆·奥特曼（Sam Altman）等人，目前正在积极布局 Worldcoin 等全民基本收入（UBI）项目，这表明站在技术前沿的引领者已经清晰预见到了第二波冲击：大批人类将彻底失去工作。针对“大多数工作是否会消失”的疑问，赫拉利认为，虽然绝大多数现有职业会被淘汰，但也会催生全新的职业形态，正如两百年前的人类无法想象“播客主播”这种工作一样。 然而，真正的核心危机并不在于工作岗位的绝对数量，而在于“劳动力的再培训与再适应”。这种转型伴随着巨大的经济成本和难以承受的心理折磨，更严酷的是，这绝非一次性的阵痛。很多人误以为 AI 革命会像历史上的工业革命一样，在 2030 年带来剧烈冲击、消灭 60% 的工作，经过十年的痛苦阵痛与技能重塑后，社会便重新恢复稳态。赫拉利直言这种想法过于天真：AI 目前远远未触及其能力上限，2030 年会有一波颠覆，2040 年和 2050 年还会迎来更剧烈的震荡。劳动者将被迫在整个人生和职业生涯中经历“被淘汰—重新学习—再被淘汰”的无限循环，这对人类的心理韧性提出了前所未有的残酷考验。 - **动态淘汰机制**：AI 带来的不是一次性产业升级，而是持续数十年、每隔十年便推翻前序规则的永久迭代震荡。 - **心理承受瓶颈**：人类再培训的真正阻碍不仅是技能获取的门槛，更是不断被剥夺职业认同后心理防线的崩溃。 ### [01:08:35](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4115s) 认知工作自动化与“全球基本收入”的伪善假象 针对巴特利特对于“当 AI 每秒都在自我迭代时，人类还能做什么新工作”的追问，赫拉利指出了当前技术落地的反直觉现象：AI 目前极度擅长替代“纯认知型工作”，却很难取代结合了“精细运动技能与社交情商”的复杂工种。例如，那些整天查阅医学文献、分析化验指标并出具诊疗方案的专科医生，其核心工作纯属认知范畴，反而最容易被算法全面自动化；而一名需要安抚啼哭儿童并为其更换绷带的护士，则需要同时调动微细动作操作和高度的人际沟通同理心，机器人在短期内极难完美胜任。尽管巴特利特以特斯拉 Optimus 机器人平稳捏取鸡蛋的视频为例提出质疑，赫拉利依然强调，物理世界中的人机交互（如自动驾驶面对醉酒乘客和横穿马路的行人）涉及极其复杂的不可预测性，其实现难度远高于处理纯符号与数据的认知领域。 在此基础上，赫拉利揭示了科技巨头倡导全民基本收入（UBI）时掩盖的巨大地缘政治盲区：所谓的 UBI 究竟是“全球性的（Universal）”还是“国家性的（National）”？如果系统只是通过对加州硅谷的科技巨头征税，来补贴宾夕法尼亚或俄亥俄州失去饭碗的工人，这根本不是真正的“全民基本收入”。真正的危机发生在全球边缘地带： - 一旦加州的自动化机器人缝制衣服的综合成本低于危地马拉或墨西哥的廉价劳动力，跨国代工产业将瞬间回流。 - 萨姆·奥特曼等人绝不可能指望美国国会将加州收取的巨额科技税款无偿转移给危地马拉或巴基斯坦的失业人口。 - 缺乏本国技术巨头和税基的发展中国家将瞬间沦为“毫无价值的弃子”，造成灾难性的全球贫富与阶级分化。 ### [01:12:06](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4326s) 教育的根本危机：人类史上首次失去对未来的预测能力 当探讨家长与学校应该投入十到十二年的时间为下一代培养什么核心技能时，赫拉利的回答令人不寒而栗：“没有人知道。”人类历史上第一次陷入了一种完全无法预知二三十年后社会与就业市场全貌的困境，导致所有针对具体专业技能的早期教育投资都充满了巨大的盲目性。 为了展现这种前所未有的危机感，赫拉利回顾了历史：如果一个人生活在公元 1023 年的一千年前的英格兰，他同样无法预测三十年后的具体政治事件——诺曼人、维京人或苏格兰人是否会大举入侵，是否会突发地震或毁灭性瘟疫；但无论发生何种动荡，他都能百分之百确定 1050 年代的经济与社会结构依然建立在农业耕作、石匠手艺与基础商贸之上。然而在今天，面对狂飙突进的 AI 与生物技术，没有任何专家或政府能够描绘出二十年后的职业版图与社会形态。 ### [01:13:32](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4412s) 未来的不确定性与技能焦虑：编码还会是一门手艺吗？ 在过去的大部分人类历史中，父母对下一代未来的规划往往是笃定而明确的。如果你生活在几百年前，大概率能预见孩子们长大后仍会是农民；教他们如何收割小麦、烘焙面包、骑马或是拉弓射箭，都是极具现实价值的技能，因为这些本事在三十年后依然必不可少。然而站在当下审视未来的三十年，几乎没有任何人能够确切预知届时究竟需要什么样的职业技能。许多家长认为如今是计算机与人工智能的时代，理应让孩子抓紧学习编程，但在三十年后，人类或许根本不再需要亲自编写一行代码，因为 AI 撰写代码的能力早已远超人类。 对于未来，唯一的确定性恰恰在于“极端的不确定性与剧烈动荡”——世界将以一种不断加速的节奏持续发生巨变。当主持人询问这种加速变革是否会加剧冲突、引爆他在个人频道中提及的“战争的回归”时，赫拉利给出了肯定的回答。人类社会正在步入一个极其危险的新节点，过去十年那种被视为理所当然的和平，其根基正被全方位瓦解。 - 历史经验的失效：传统农耕社会技能具备数十年乃至数百年的代际稳定性，现代科技革命彻底打破了这种传承模式。 - 技能贬值的潜在风险：当下被视作硬通货的计算机编程技能，很可能在未来数十年内被高度自主的 AI 系统全面替代。 - 未来的核心确定性：三十年后的世界唯有一点毋庸置疑，即剧烈的动荡和超高速的不可逆变迁。 --- ### [01:14:29](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4469s) 自由主义国际秩序崩塌与战争时代的回归 赫拉利沉重地指出，大约十年前人类其实生活在有史以来最和平的黄金年代，遗憾的是，这个时代已经彻底宣告终结。如今世界正在迈入一个充斥着战争乃至帝国主义扩张的新纪元。从俄罗斯对乌克兰的全面入侵，到中东地区的战火重燃，再到委内瑞拉与圭亚那的领土危机以及东亚局势的暗流涌动，战争选项已经重新摆上了各国的地缘政治桌面。 这种局面的出现，不仅是因为科技和经济剧变带来的冲击，更深层的原因在于支撑全球和平的支柱——自由主义国际秩序的土崩瓦解。这一秩序虽然远非完美，但它曾经建立在一个核心理念之上：尽管各国存在民族与文化差异，但全人类拥有共通的基本体验、需求与利益，因此通过国际合作来化解冲突、解决共同问题是合乎逻辑的选择。然而，该秩序不仅遭到俄罗斯、伊朗、朝鲜等从未真正接受它的外部力量的冲击，更遭遇了来自其核心缔造者——美国内部的背刺。随着唐纳德·特朗普当选总统并公开宣扬“我不在乎任何全球秩序，我只在乎我自己的国家利益”，这种狭隘的民族孤立主义迅速蔓延全球。 各国如果都只以本国利益为唯一准则，那么国家之间的摩擦将失去任何调解机制。摧毁旧秩序的民粹力量从未提出任何切实可行的新方案来规范国际交往，他们只是将既有框架砸得粉碎。赫拉利强调，秩序破灭之后的唯一替代品不是某种更高级的结构，而就是赤裸裸的“混乱与失序”。 ```mermaid flowchart TD accTitle: 自由主义国际秩序的崩溃与失序 accDescr: 展示自由主义国际秩序如何在内外夹击下瓦解并滑向混乱的过程 subgraph LiberalOrder["战后自由主义秩序（相对和平期）"] A["普遍共识：全人类共享基本利益"] --> B["国际规则规范与多边协作"] end subgraph InternalExternalAttacks["秩序遭受双重解构"] C["外部挑战者<br/>（俄罗斯、伊朗、朝鲜等）"] --> E["推翻现有规则"] D["内部缔造者退群<br/>（美国'美国优先'孤立主义）"] --> E end subgraph Consequence["秩序坍塌的终局"] E --> F["缺乏替代方案的权力真空"] F --> G["绝对失序与局部/全球战争重燃"] end LiberalOrder --> InternalExternalAttacks ``` --- ### [01:17:27](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4647s) 军备竞赛的恶性循环与“友好堡垒”的幻觉 面对未来是否会有更多战争的提问，赫拉利的答案是冷酷而现实的：除非人类能够重建国际秩序，否则未来几年只会在更多地区目睹更频繁、更惨烈的战争。世界各地的国防预算正在失控般暴涨，从而引发可怕的恶性循环——邻国增加军费，本国就被迫跟进，进而刺激对方进一步扩军备战。 这种军费暴涨反衬出21世纪初曾经是多么不可思议的和平年代。在以往漫长的帝王统治时期，无论是皇帝、可汗还是苏丹，军事开支永远是国库账本上的第一大项目，用于士兵、舰队与要塞的钱远超一切民生投入。但在21世纪初，大多数国家的国防预算仅占GDP或政府开支的极低比例，教育、医疗和公共福利前所未有地占据了财政大头。而眼下，巨额资金正在被无情地从护士、学校与社会工作者手中抽离，转而流入坦克、导弹与网络武器的军工黑洞之中。这种转变并非宿命，而是人类愚蠢决策的产物；昔日的和平并非神迹，而是前人明智决策所累积的成果。 针对特朗普可能再度当选美国总统的担忧，赫拉利认为这极有可能发生，且一旦成真，将给残存的全球秩序带来致命一击。民粹政客往往刻意制造“爱国主义”与“全球主义”之间的虚假对立，将爱国描绘成与全球合作水火不容。赫拉利驳斥道，倡导全球合作绝不等于建立某种虚构且危险的“世界政府”，而是建立维系国家间交往的基本行为规范。许多人幻想世界可以退化为由一个个“友好堡垒”组成的网络——每个国家高筑防御工事只求自保，同时与邻里和睦贸易。但历史一再证明，“堡垒从来都不会友好”，每座堡垒永远渴望以牺牲邻居为代价，换取自身更多的土地、繁荣与安全感，而这正是通向全面冲突与全面战争的直通大道。 - 历史财政支出的对比：传统封建帝王第一开支永远是军事；21世纪初福利国家将教育与医疗置于国防之上；如今军费开支再度飙升挤占民生资金。 - 伪命题辨析：爱国主义与全球合作并不矛盾；全球秩序的本质是制定国家间和平共处的交互准则，而非建立集权式的世界政府。 - 堡垒逻辑的虚妄性：各自筑墙的孤立主义政权无法维持友善平衡，资源掠夺与安全焦虑必然驱使堡垒相互吞并。 --- ### [01:22:05](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=4925s) “信息节食”与每日两小时冥想：在过载的世界保持心智健康 主持人引用英国俗语“无知是福”（Ignorance is bliss）向赫拉利发问：作为一个职业要求必须洞悉现实黑暗面、绝不能生活在无知中的历史学者，这种认知是否彻底剥夺了他的幸福感？赫拉利坦言，自己目前其实相当快乐，甚至比人生中大部分时期都要平静满足。其秘诀在于，他不仅将精力投入到对世界大势的宏观研究中，更花费大量时间来维护自身的心智健康，坚持严格的“信息节食”（information diet）。 在赫拉利看来，信息之于大脑，正如食物之于肉体。摄入营养以维系生存是必要的，但如果暴饮暴食或长期摄入垃圾食品，身体就会生病；心智同样如此，若是盲目吸收过多劣质、煽动性的信息，心智必然走向病态。因此，他不仅严选信息来源，更践行定期的“信息断食”： - 坚持每天拿出整整两个小时进行深度冥想； - 每年专门安排长达30至60天的离线禅修（retreat），在此期间彻底与外界切断联系； - 禅修期间严格禁止使用智能手机、查阅电子邮件，甚至完全不读任何书籍。 赫拉利解释道，这段完全抽离的时光让他能够专注于内观，静静观察自己的身体感受与意识流动，加深对自我的认知，同时充分“消化”在其余时间吸收的海量信息。他坦言，如果没有这种信息节食与冥想习惯，他绝不可能写出那些传世著作，整个人早就在永不停歇的新闻轮转与信息洪流中丧失全局视野、陷入彻底的精神崩溃。 他进一步深刻剖析了有机生物与人工系统的本质差异：人类作为有机实体具有固定的生物节律与周期性，必须通过睡眠和休息来修复机能，而 AI 系统从不入睡。哪怕是看似永动机一般的华尔街证券交易所，每天下午依然会按时休市，周末和圣诞节也会停摆。如果圣诞节爆发战争，金融市场往往要滞后数日才能由休假归来的人类做出反应，这表明哪怕是金钱资本也保留了人类社会的休息周期。然而，面对不知疲倦的无机算法系统，人类若不能守住自身的周期节律，必将在算法的碾压下丧失心智阵地。 ### [01:26:05](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=5165s) 有机体的极限：24 小时不间断运转与对睡眠的掠夺 尤瓦尔·赫拉利（Yuval Noah Harari）指出，若完全任由 AI 掌控，社会将失去任何休止符，进入一年 365 天、一天 24 小时不间断运转的状态，而人类作为有机体终将彻底崩溃。他以现代政治家与中世纪国王做对比：在中世纪，哪怕是国王在泥泞小道上乘坐马车出行，在路途的数日甚至数周内没有任何人能联系到他——即便当时法兰西军队正在大举入侵，国王依然拥有强制性的休止时间；而如今的现代国家首相或总统，由于 24 小时连轴转的新闻循环体系，必须时刻待命、无休止地处理信息并做出反应。计算机是为此类永不停歇的工作而设计的，但人类大脑绝非如此。强行让一个有机体一天 24 小时保持清醒和反应状态，结果只能是生理与心理的全面瘫痪。 主持人史蒂芬·巴特利特（Steven Bartlett）随即引用了流媒体巨头 Netflix 前首席执行官里德·哈斯廷斯（Reed Hastings）的一句名言：“我们最大的竞争对手是睡眠（Our biggest competitor is sleep）”。赫拉利对此深表认同，称这是一句令人不寒而栗却极其坦诚的大实话。人类一旦失去睡眠就会衰竭乃至死亡，然而在社交媒体和流媒体平台围绕人类注意力的争夺战中，企业衡量成功的唯一标尺往往是“用户参与度（user engagement）”。 在这一商业逻辑的度量衡下，不同情绪状态对企业的商业价值呈现出极其荒谬的倒错： ```mermaid flowchart TD accTitle: 用户状态与企业商业价值转换链条 accDescr: 展示了平台算法偏好愤怒而非睡眠或平静的商业逻辑转换流程 subgraph UserStates["人类生理与心理状态"] S1["睡眠 (Sleep)"] S2["平静与愉悦 (Joy / Peace)"] S3["愤怒与激愤 (Outrage)"] end subgraph BusinessValue["算法与商业转化机制"] B1["0 广告展示 / 0 美元收益"] B2["极少广告曝光 / 少量收益"] B3["算法高频推荐 / 密集广告 / 商业价值最大化"] end S1 --> B1 S2 --> B2 S3 --> B3 ``` - **愤怒（Outrage）的高商业回报**：在平台的衡量标准中，1 小时的“愤怒”远胜过 10 分钟的“愉悦”，更远远胜过睡眠。因为在 1 小时的愤怒刷屏中，用户可能浏览三条甚至更多广告，为平台贡献可观的利润（例如数十美元的广告费）。 - **睡眠与愉悦的商业惩罚**：用户熟睡两小时，平台颗粒无收；10 分钟的平静喜悦，平台也许只能展示一条广告。然而从人类有机体生理运作的角度来看，10 分钟的纯粹愉悦远比 1 小时的狂怒健康，人类每晚更是不可或缺地需要 6 至 8 小时的充足睡眠。 --- ### [01:28:53](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=5333s) 算法对人类大脑的全面劫持与注意力竞赛 巴特利特谈到像 TikTok 这样短视频平台的算法机制，其成瘾性已经达到了令人咋舌的地步，原因在于它们精准地“黑（Hack）”进了人类的心智机制。他提到以往的社交媒体算法虽然也有成瘾性，但 TikTok 登场后犹如说了句“看我的（hold my beer）”，将算法成瘾机制推向了极致。甚至连 60 多岁的老人也完全深陷其中，如同被催眠一般，根本无法意识到自己正在被推荐算法和广告转化模型所操纵。巴特利特以自己的专职司机为例，每当自己走向车门，总能看到司机目不转睛地疯狂滑动 TikTok 屏幕；前一晚他甚至忍不住问司机：“你意识到 TikTok 已经彻底接管你的大脑了吗？” 赫拉利强调，当前我们仅仅处在算法争夺大脑注意力指数级增长曲线的起点，“更可怕的还在后面”。这些能够针对个人弱点进行心智劫持的社交媒体算法诞生至今不过短短十年左右。在这个残酷的存量竞争市场中，任何平台只要无法击败竞争对手的算法，就将面临淘汰消亡。巴特利特以埃隆·马斯克（Elon Musk）接手后的 Twitter（现 X 平台）为例：为了在竞争中存活、延长用户驻留时间并兜售更多广告，平台算法开始频繁推送极其血腥暴力的极端内容（如车祸惨剧、致命事故等），其出现频率远超过去十年的总和，这迫使 Instagram 等其他竞争平台为了夺回用户注意力也卷入更极端的推流军备竞赛中。 --- ### [01:30:46](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=5446s) 2023 年世界的通病：“过度兴奋”与对“无聊”的排斥 赫拉利话锋一转，借用马斯克旗下的地下隧道公司“The Boring Company（无聊公司）”打了个幽默而深刻的比方：或许把 Twitter 和 TikTok 变得“更无聊（more boring）”才是拯救人类的良策。尽管这从商业角度来看是个糟糕的决定，但他坚信：**如果人类无法学会重拾“无聊”，文明将无法延续**。 赫拉利一针见血地指出，若要用一个词概括 2023 年这个世界的根本病因，那就是——“过度兴奋（excited）”： - **兴奋不等于快乐**：在现代日常语境中，人们习惯将“兴奋”与“快乐”混为一谈（例如声称“见到你真兴奋”、“读到一本令人兴奋的书”）。但从生物进化与神经生理学角度来看，兴奋绝不等于幸福。 - **兴奋的生理实质是应激反应**：兴奋意味着身体处于战斗或逃跑（Fight or Flight）模式，神经系统全面紧绷，神经元高频放电，肌肉极度僵硬。在大多数情况下，激起人类强烈兴奋的往往是负面情绪——**恐惧是兴奋，仇恨是兴奋，愤怒也是兴奋**。赫拉利提到，当自己与真正的好友相见时，内心的感受是彻底的“放松与平静（relaxed）”，而非神经亢奋。 - **政治层面的兴奋泛滥**：现代政坛充斥着过多喜欢搞“兴奋大动作”的政客。世界真正急需的，其实是更多像拜登（Joe Biden）这样“无聊”的政治家，他们做事波澜不惊、不那么刺激，却能维护基本的秩序与稳定。 人类大脑的进化环境决定了其天生对刺激保持警惕与渴望，但在远古时期，这种兴奋刺激只是偶尔发生，大多数时候大脑处于离线与平淡状态。如今现代科技与智能算法彻底绑架了这一弱点，源源不断地提供定制化的人造兴奋，导致人类彻底丧失了“忍受无聊的能力”。只要在公共场合等待两分钟，人们就会焦躁不安地掏出手机刷新 TikTok 或 Twitter。巴特利特补充道，心理学实验早已证实，当面临独处无聊与遭受电击的选择时，许多人甚至宁愿主动接受电击也不愿什么都不做。 赫拉利总结道，内心平静（peace of mind）与无聊是紧密相连的近邻，正如兴奋与愤怒是邻居一样。如果一个人连片刻的无聊都无法包容，只要察觉到一丝平淡就仓皇逃向算法刺激，他就永远不可能获得真正的内心宁静；而当生活在其中的个体都无法体会平静时，整个世界的和平与安宁也必将无从谈起。 --- ### [01:34:32](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=5672s) 中插赞助与商业信息：Flight Fund 与 Huel 在播客中段，主持人巴特利特切入了一段商业赞助与投资项目信息： - **Flight Fund 股权投资基金**：巴特利特介绍了自己于 2023 年创立的私募基金 Flight Fund，旨在成为欧洲表现最佳的投资基金，专注于投资有望成为欧洲下一代“独角兽（Unicorns）”的高成长型公司。该基金汇聚了欧洲众多顶尖创新企业家投资人，现开放初创企业创始人提交商业计划书，并面向符合资质的专业投资人开放申请通道。巴特利特强调该基金受英国金融行为监管局（FCA）监管，属于高风险投资，仅适用于成熟投资人。 - **Huel 营养代餐品牌**：巴特利特分享了自己作为投资人和代言人的品牌 Huel。他重点推荐了专为听众打造的“Steven Bundle”精选套装，包括富含高蛋白的黑金版海盐焦糖味代餐粉（每罐 17 份用量）、经典即饮装饮料，以及全新推出的巧克力焦糖味营养能量棒。巴特利特坦白透露，多年前 Huel 刚推出初代能量棒时味道欠佳，自己当时拒绝宣传，而如今改良后的配方口感出色，已被他和团队成员迅速扫荡一空。 ### [01:36:44](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=5804s) 生于 1976 年与生于 2020 年代：和平红利的消逝 在结束短暂的赞助口播后，主持人 Steven Bartlett 向尤瓦尔·诺亚·赫拉利（Yuval Noah Harari）抛出了一个颇具启发性的假设问题：如果给他一次选择的机会，是愿意像当年一样出生在 1976 年，还是选择出生在当下的时代？赫拉利毫不犹豫地选择了 1976 年。 赫拉利深情地回顾道，他所处的那一代人极其幸运，享有了人类历史上最和平、最充满乐观主义的时代红利。他们亲身经历了冷战的终结和铁幕的倒塌，在人类文明的长河中，几乎很难再找到一个比那段岁月更美好的时期。然而，当他审视眼下发生的一切时，他坦言自己完全不羡慕在 2020 年代成长起来的年轻一代，因为他们所面临的未来充满了前所未有的动荡与未知。 ### [01:37:25](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=5845s) 走向希望的行动逻辑：人类的能动性与群体协作 面对愈发暗淡的时代背景，Steven Bartlett 询问赫拉利能否给出一个充满希望的总结性解决方案，作为触发多米诺骨牌效应的起点，指引人们迈向更光明的未来。赫拉利坚定地指出，人类目前依然拥有“能动性”（Agency）。虽然算法正在逐步蚕食人类的自主权，但它们尚未完全掌控一切，大部分核心权力依然牢牢掌握在人类手中。每个个体都拥有一定程度的能动性，而这也意味着每个人都肩负着相应的责任。 赫拉利强调，没有人能够包揽并解决世界上所有的问题。明智的做法是找到那件真正贴近你内心、并且你对其拥有深刻理解的单一议题，聚焦于此去创造改变。更关键的是，改变世界绝不能依靠单打独斗，人类真正的“超能力”在于大规模合作。 ```mermaid flowchart TD accTitle: 应对复杂世界的人类行动模型 accDescr: 展示面对全球危机时个体如何通过聚焦与协作实现高效行动 subgraph Problem["现实困境"] A["庞大且复杂的全球性危机"] B["算法权力扩张但人类仍具能动性"] end subgraph Strategy["行动策略"] C["明确自身责任：专注单一议题"] D["放弃单打独斗：投身组织协作"] end subgraph Impact["效能对比"] E["500 名孤立行动者<br/>力量分散且影响力微弱"] F["50 人协作团队<br/>凝聚合力实现规模化变革"] end Problem --> Strategy C & D --> Impact ``` - **个体行动的局限**：不要试图做面面俱到的孤胆英雄，将其他领域的问题留给他人解决。 - **群体协作的乘数效应**：在组织体系中紧密协作的 50 个人，所能爆发出的能量与改变，远远超过 500 个互不联络的孤立行动者。 ### [01:40:07](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=6007s) 历史的真正意义：理解变革并重塑未来 Steven Bartlett 提到赫拉利的代表作《人类简史》（*Sapiens*）迎来出版十周年，并推出了十周年纪念版。这本书在图书市场普遍唱衰的背景下斩获了全球超过 2500 万册的惊人销量，彻底重塑了无数读者对人类文明的认知。在被问及是否有某个关乎未来却在整场对话中尚未提及的重要议题时，赫拉利将焦点重新拉回了他的老本行——历史学。他直言，当我们畅想未来时，历史学比以往任何时候都更加紧要。 赫拉利阐述道，历史学本质上并不是研究“过去”的学科，而是研究“变化”以及“事物如何演变”的学问。没有人会仅仅为了过去而执着于过去；生活在罗马帝国或中世纪的人们早已作古，今人既无法抚平他们经历的苦难，也无法纠正古时的罪恶，逝者同样不在意后世的评价。人们研究历史的真正价值，在于通过剖析过去几个世纪以来的演变机制，获得审视当下变革进程的全局视角。历史的一大误区，是人们总幻想着回到过去伸张正义，但这绝不可能；真正的核心命题在于：我们如何拯救现在的同胞？如何阻止即将降临的浩劫？ 赫拉利总结道，历史传递的最核心信息在于：**我们身处的世界是由人类一手缔造的**。无论是民族国家、现代企业、资本主义经济体系，还是基督教、印度教等宗教形态，全都是人类构建的产物。既然是人类创造了它们，人类自然拥有改变它们的能力。物理定律与生物规律是人类无法违抗的客观法则，但世间许多被普遍误认为是“天然常态”的制度与规则，实际上只是历史演变的产物。洞悉历史，才能辨明何为不可抗力，何为可以被重构的人造之物。 ### [01:43:34](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=6214s) 闭幕传统：给全人类制定的唯一一条全球法则 对话接近尾声，Steven Bartlett 提出了该播客的传统保留环节：由上一位未曾谋面的嘉宾为赫拉利留下的提问——“如果你有权颁布且仅颁布一条全球性法律，那会是什么？为什么？” 面对这个极具分量的问题，赫拉利给出了发人深省的回答：**所有人必须减少信息摄入，将更多的时间用于反思和消化自己已经获悉、已经听闻的内容。** Steven 对赫拉利表达了由衷的敬意与感谢。他坦言赫拉利的著作点亮了他个人生活的多处盲区，更是以一种人人都能理解的叙述方式，将复杂的历史脉络普及给了全球数以亿计的受众，赋予了大众审视过去的全新工具。最后，节目在关于睡眠科技产品（Eight Sleep）的赞助介绍与往期节目推荐中圆满落幕。 ## Notable quotes > [19:12](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1152s) 如果以历史学家的眼光来评分，在应对工业革命这一关上，我只能给人类打‘C-’。如果面对 21 世纪的 AI 与生物工程革命，我们再次拿到一个‘C-’，人类作为一个物种就将彻底终结（That's the end of us）。 — 尤瓦尔·赫拉利 > [22:16](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1336s) 太初有字（In the beginning was the word） — 尤瓦尔·赫拉利 > [22:16](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=1336s) 开启整座人类文明大门的主钥匙（The master key that unlocks all the doors of our civilization） — 尤瓦尔·赫拉利 > [37:56](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=2276s) 人类必须习惯一个事实——我们不再是神秘的灵魂，我们现在是可被破解的动物（hackable animals）。 — 尤瓦尔·赫拉利 > [52:42](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3162s) 我们在彻底理解操纵带来的系统性后果之前，早就迫不及待地掌握了操纵的手段 — 尤瓦尔·赫拉利 > [55:42](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=3342s) 意识到自己本有机会长生不老，却可能因一场意外而暴毙，这将引发人类历史上前所未有的极致焦虑与恐惧。 — 尤瓦尔·赫拉利 > [01:26:05](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=5165s) 我们最大的竞争对手是睡眠（Our biggest competitor is sleep） — 里德·哈斯廷斯 > [01:30:46](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=5446s) 如果人类无法学会重拾“无聊”，文明将无法延续 — 尤瓦尔·赫拉利 > [01:43:34](https://www.youtube.com/watch?v=UzOJiqN_DpM&t=6214s) 所有人必须减少信息摄入，将更多的时间用于反思和消化自己已经获悉、已经听闻的内容。 — 尤瓦尔·赫拉利 ## Mentioned 尤瓦尔·诺亚·赫拉利 (Yuval Noah Harari), 史蒂文·巴特利特 (Steven Bartlett), 埃隆·马斯克 (Elon Musk), 萨姆·奥特曼 (Sam Altman), 弗拉基米尔·普京 (Vladimir Putin), 唐纳德·特朗普 (Donald Trump), 乔·拜登 (Joe Biden), 里德·哈斯廷斯 (Reed Hastings), Neuralink, OpenAI, ChatGPT, Worldcoin, Tesla Optimus, TikTok, Twitter / X, Instagram, Netflix, The Boring Company, Flight Fund, Huel, Eight Sleep, 《人类简史》 (Sapiens), 《未来简史》 (Homo Deus), 《今日简史》 (21 Lessons for the 21st Century), 世界经济论坛 (WEF) ## Video description If you enjoy hearing about the potential impact of AI on humanity, I recommend you check out my conversation with ex-Google office, Mo Gawdat which you can find here: https://www.youtube.com/watch?v=bk-nQ7HF6k4 0:00 Intro 2:15 What Is Your Mission, and What Is Your Warning to the World? 07:58 Is This the End of Humanity as We Know It? 12:29 Connecting Computers to Human Brains 15:11 What Are Your Concerns About AI? 27:35 The Dangers of AI to the Financial System and Governments 37:56 Do Humans Have Free Will, and Will AI Take It from Us? 45:41 The Problems of AI Forging Relationships with Humans 52:42 Are We Happy? 55:42 Fighting Immortality and Its Consequences 01:00:00 Will Bioengineering Create Different Social Classes and Types of People? 01:06:30 Will AI Take Over Our Jobs? 01:12:06 What Should We Teach Our Children to Be Prepared for the Future? 01:14:29 We're Entering a Scary New Era. 01:19:31 What Should We Do to Stop/Change the Trajectory We're Heading Towards? 01:22:05 The Importance of Disconnecting from Information 01:27:10 What Media Corporations Want from You 01:30:16 We Need More Boredom in the World and Politics 01:36:49 Is There Hope for Humanity? 01:40:07 The Importance of History for Our Future 01:43:34 The Last Guest Question You can pre-order the 10th anniversary edition of ‘Sapiens’, here: https://bit.ly/48JVQ6c Follow Yuval: Twitter: https://bit.ly/3HdUxR7 Instagram: https://bit.ly/41WLbCT YouTube: https://bit.ly/3vyAwm0 Flight fund: https://flightfund.com/ Get tickets to The Business & Life Speaking Tour: https://stevenbartlett.com/tour/ FOLLOW ► Instagram: https://www.instagram.com/steven/ Twitter: https://x.com/StevenBartlett?s=20 Linkedin: https://www.linkedin.com/in/steven-bartlett-56986834/ Sponsors: Huel: https://try.huel.com/steven-bartlettEight Sleep: https://www.eightsleep.com/uk/steven/ CODE: STEVEN [Source](https://www.youtube.com/watch?v=UzOJiqN_DpM)

[19:12] 尤瓦尔·赫拉利: 如果以历史学家的眼光来评分，在应对工业革命这一关上，我只能给人类打‘C-’。如果面对 21 世纪的 AI 与生物工程革命，我们再次拿到一个‘C-’，人类作为一个物种就将彻底终结（That's the end of us）。 [22:16] 尤瓦尔·赫拉利: 太初有字（In the beginning was the word） [22:16] 尤瓦尔·赫拉利: 开启整座人类文明大门的主钥匙（The master key that unlocks all the doors of our civilization） [37:56] 尤瓦尔·赫拉利: 人类必须习惯一个事实——我们不再是神秘的灵魂，我们现在是可被破解的动物（hackable animals）。 [52:42] 尤瓦尔·赫拉利: 我们在彻底理解操纵带来的系统性后果之前，早就迫不及待地掌握了操纵的手段 [55:42] 尤瓦尔·赫拉利: 意识到自己本有机会长生不老，却可能因一场意外而暴毙，这将引发人类历史上前所未有的极致焦虑与恐惧。 [86:05] 里德·哈斯廷斯: 我们最大的竞争对手是睡眠（Our biggest competitor is sleep） [90:46] 尤瓦尔·赫拉利: 如果人类无法学会重拾“无聊”，文明将无法延续 [103:34] 尤瓦尔·赫拉利: 所有人必须减少信息摄入，将更多的时间用于反思和消化自己已经获悉、已经听闻的内容。

[72:06] Yuval Noah Harari: Nobody has any idea.

[0:00] We are now in a new era of wars. And [0:03] unless you reestablish order fast, then [0:06] we are doomed. [0:07] You've all Noah Harrari, [0:08] one of the brightest minds on planet [0:10] earth, historian, a bestselling author [0:12] of some of the most influential [0:13] non-fiction books in the world today. [0:16] I think we are very near the end of our [0:18] species because people often spend so [0:21] much effort trying to gain something [0:23] without understanding the consequences. [0:26] For example, we will get to a life where [0:29] you can live indefinitely. But realizing [0:32] that you have a chance to live forever, [0:34] but if there is an accident, you die. [0:36] The people who will be in that situation [0:39] will be at a level of anxiety and terror [0:42] unlike anything that we know. Then you [0:44] have artificial intelligence and the [0:47] world is is not ready for it. It's the [0:49] first technology in history that can [0:51] make decisions by itself and take power [0:54] away from us to hack human beings, [0:57] manipulate our behavior and making all [0:59] these decisions for us or about us. [1:02] Whether to give you a loan, whether to [1:03] give you a mortgage, dating us, shaping [1:05] your romantic life. But the real problem [1:08] is that increasingly the humans at the [1:11] top could be puppets. When the most [1:14] consequential decisions are made by [1:16] algorithms, global financial decisions, [1:18] wars. This is extremely dangerous, but [1:20] it's not inevitable. Humans can change [1:22] it. [1:23] But with what's to come, are you [1:24] optimistic about the future? [1:27] I'm very worried about two things. First [1:29] of all, [1:31] quick one. This is really, really [1:32] fascinating to me. On the back end of [1:34] our YouTube channel, it says that 69.9% [1:38] of you that watch this channel [1:39] frequently over the lifetime of this [1:41] channel haven't yet hit the subscribe [1:42] button. I just wanted to ask you a [1:44] favor. It helps this channel so much if [1:46] you choose to subscribe. Helps us scale [1:48] the guests, helps us scale the [1:50] production, and it makes this show [1:51] bigger. So, if I could ask you for one [1:52] favor, if you've watched the show before [1:54] and you've enjoyed it and you like this [1:56] episode that you're currently watching, [1:57] could you please hit the subscribe [1:58] button? Thank you so much. and I will [2:00] repay that gesture by making sure that [2:03] everything we do here gets better and [2:04] better and better and better. That is a [2:06] promise I'm willing to make you. Do we [2:07] have a deal? [2:11] [Music] [2:16] I have three of your books here and [2:18] these are three books that sent a huge [2:22] tidal wave, a ripple through society. [2:25] with these books and with all of the [2:26] work that you're doing now, with the [2:28] lectures you give, the the interviews [2:30] you give, what is your mission? What [2:33] what is the sort of if I was to be able [2:35] to summarize what your collective [2:37] mission is with your work? What is that? [2:39] It's to clarify and to focus the public [2:42] conversation, the global conversation, [2:44] uh to help people focus on the most [2:47] important challenges that are facing [2:48] humankind and also to bring at least a [2:52] little bit of clarity to the collective [2:54] and and to the individual mind. I mean, [2:57] one of my main messages in all the books [3:00] is that our minds are like factories [3:04] that constantly produce [3:07] stories and fictions that then come [3:10] between us and the world. And we often [3:13] spend our lives interacting with [3:16] fictions that we or that other people [3:19] created uh with and completely losing [3:22] touch w with with reality. And my job [3:26] and I think the job of historians more [3:28] generally is to show us a way out. [3:33] Inherent in much of your work is what [3:37] feels like a warning. [3:40] And I've I've watched hundreds of videos [3:43] that you've produced or interviews [3:44] you've done um all around the world and [3:48] it feels like you're trying to warn us [3:50] about something, multiple things. [3:52] Mhm. If my estimation there is correct, [3:56] what is the warning? [3:58] Much of what we take to be real is is is [4:02] fictions. And and the reason that [4:04] fictions are so central in in human [4:07] history is because we control the planet [4:11] and rather than the chimpanzees or the [4:13] elephants or any of the other animals [4:15] because not because of some kind of in [4:17] individual genius that each of us has [4:20] but because we can cooperate much better [4:23] than any other animal. We can cooperate [4:26] in much larger numbers and also much [4:28] more flexibly. [4:30] And the reason we can do that is because [4:34] we can create and believe in fictional [4:36] stories because every largecale human [4:39] cooperation whether uh religion or [4:43] nations or corporations [4:45] are based on mythologies on on fictions. [4:49] Again I'm not just talking about gods. [4:51] This is the easy example. Money is also [4:54] a fiction that we created. Corporations [4:57] are a fiction. they exist only in our [4:59] minds. Uh even lawyers would tell you [5:02] that corporations are legal fictions. [5:04] And this is on on the one hand such a [5:08] source of of immense power. [5:11] But on the other hand, again the danger [5:13] is that we completely lose touch with [5:14] reality and we are manipulated by all [5:20] these fictions, by all these stories. [5:22] Again, stories are not not bad. They are [5:24] tools. As long as we use them to [5:27] cooperate and to help each other, that's [5:30] wonderful. Um, money is not bad. If we [5:33] didn't have money, we would not have a [5:34] trade network. We everybody would have [5:37] maybe with their friends and family to [5:39] to produce everything by themselves like [5:41] the chimpanzees do. uh the fact that we [5:44] can enjoy uh food and clothing and [5:48] medicines and enter entertainment [5:52] created by people on the other side of [5:54] the world is largely because of money. [5:56] But if we forget that this is a tool [5:58] that we created in order to help [6:00] ourselves and instead uh this tool kind [6:04] of enslaves us and runs our life and um [6:09] you know I'm now just back home in [6:12] Israel there is a terrible war being [6:15] waged and most wars in history and also [6:18] now they are about stories they're about [6:21] fictions. People think that humans fight [6:25] over the same things that wolves or [6:27] chimpanzees fight about, that we fight [6:29] about territory, that we fight about [6:31] food. It sometimes happens, but most [6:34] wars in history were not really about [6:36] territory or food. There is enough land, [6:40] for instance, between the Jordan River [6:42] and the Mediterranean to build houses [6:44] and schools and hospitals for everybody. [6:47] And there is certainly enough food. [6:48] There's no shortage of food. But people [6:51] have different mythologies, different [6:54] stories in their minds and they can't [6:56] find a common story they can they can [6:58] agree about. And this is at the root of [7:01] most UN conflicts. And being able to [7:05] tell the difference between what is a [7:08] fiction in our own mind and what is the [7:12] reality. This is a a crucial skill and [7:16] we are not getting better at finding [7:20] this difference as time go time time [7:22] goes on [7:23] and also with new technologies which I [7:26] write about a lot like artificial [7:28] intelligence. The fantasy that AI will [7:33] answer our questions, will find the [7:35] truth for us, will tell us the [7:37] difference between fiction and reality. [7:40] This is this is just another fiction. I [7:42] mean AI can do many things better than [7:45] humans but for reasons that we can [7:48] discuss I don't think that it will [7:50] necessarily be better than humans at [7:53] finding the truth or uh um uncovering [7:57] reality. It it strikes me that the the [8:00] thing that made us successful, [8:03] you know, this ability to believe in [8:04] fictions and I use the word successful, [8:06] you know, powerful powerful. Yes. [8:08] Took over the world. [8:11] The thing that made us powerful could [8:13] well be the thing that makes us [8:15] powerless [8:17] in the sense that our ability to believe [8:18] in fictions and stories create a society [8:21] that would potentially lead to our [8:24] powerlessness. [8:26] That's kind of one of the the the [8:29] messages that when I connect the dots [8:30] throughout your work and you look off [8:32] into the future, um I'm left feeling. [8:35] And even you think about the modern [8:37] problems we have, those are typically [8:39] consequences of our ability to believe [8:40] in stories [8:42] and to believe in fictions. [8:43] And if you play that forward 100 years, [8:45] maybe 200 years, [8:47] you don't believe that um you believe [8:50] we'll be the last of our species, right? [8:53] I think we are very near the kind of end [8:55] of our species. It doesn't necessarily [8:57] mean that we'll be destroyed in some [9:01] huge nuclear war or something like that. [9:04] Uh it could very well mean that we'll [9:06] just change ourselves using [9:10] uh bioengineering and using AI and brain [9:13] computer interfaces. We will change [9:15] ourselves to such an extent that we'll [9:19] become something completely different, [9:21] something far more different from [9:23] present day homo sapiens than we today [9:26] are different from chimpanzees or from [9:28] Neanderthalss. I mean basically you know [9:31] um [9:33] you have a very deep connection still [9:35] with all the other animals because we [9:38] are completely organic. [9:40] We are organic entities. our psychology, [9:44] our social habits, they are the product [9:47] of organic evolution and male and more [9:50] specifically mamalian evolution over [9:52] tens of millions of years. So we share [9:55] so much of our psychology and of our [9:58] kind of social habits with chimpanzees [10:01] and with with other other mammals. [10:04] Looking a 100 years or 200 years to the [10:07] future, maybe we are no longer organic [10:09] or not fully organic. Um you could uh [10:13] have a world dominated by cyborgs which [10:17] are entities combining organic with [10:20] inorganic parts for instance with brain [10:22] computer interfaces. Um you could have [10:25] completely nonorganic entities. [10:28] So all the legacy and also all the [10:32] limitations of 4 billion years of [10:36] organic evolution might be irrelevant or [10:40] inapplicable [10:41] do you think [10:42] to the beings of the future? [10:43] What bet would you make? [10:45] Because you're saying maybe here [10:47] I don't know. I mean we could destroy [10:48] ourselves. I think there is a greater I [10:50] mean to completely destroy every last [10:54] single human in the world. it is [10:57] possible given the technology that we [10:59] now command but it's it's very [11:02] difficult. [11:04] Um I think it's there is a greater [11:07] chance and again this is just [11:08] speculation nobody really knows but I [11:10] think I mean lots of people could suffer [11:14] terribly [11:15] but I think it's more likely that uh uh [11:19] some people will survive and then will [11:21] undergo radical changes. [11:24] So it's not that humanity is completely [11:27] destroyed. It's just transformed [11:31] into into something else. And just to [11:33] give an example of what we are talking [11:35] about, organic beings like us need to be [11:40] in one place at any one time. We are now [11:43] here in this room. That's it. Um, if you [11:46] kind of disconnect our hands or our feet [11:50] from our body, we die or at least we [11:54] lose control of of these. I mean, and [11:56] this is true of all organic entities, of [11:59] plants, of of animals. Now, with cyborgs [12:03] or with inorganic entities, this is no [12:05] longer true. They could be spread over [12:08] time, time, and space. I mean if you [12:11] find a way and people are working on [12:13] finding ways to directly connect brains [12:16] with computers or brains with bionic [12:19] parts there is there there is no [12:22] essential reason that all the parts of [12:24] the be of of the entity need to be in [12:27] the same room at the same time. As you [12:29] said that, you know, I started thinking [12:30] a little bit about Neuralink and what [12:32] Elon Musk is doing interfacing us with [12:34] computers. But then I had a secondary [12:36] thought which is if there could be two [12:39] Stevens, one here and then one in the [12:41] United States right now because we're [12:42] connected to the same computer [12:44] interface. Theoretically, I could hack [12:48] Jack over there. [12:49] I could hack his interface. So there [12:52] could be three Stevens because I hack [12:53] Jack. And then I hack you and then [12:56] there's four. And then I could [12:57] eventually try and hack the entirety of [12:59] the world or a country. [13:00] Yeah. [13:00] And there could basically be one [13:03] one [13:04] once you can connect directly brains to [13:06] computers. First of all, I'm not sure if [13:08] it's possible. [13:09] I mean, people like Elon Musk in [13:10] Norolink, they tell us it's possible. [13:12] I'm I'm I'm I'm still waiting for the [13:14] evidence. I don't think it's impossible, [13:16] but I think it's much more difficult [13:19] than than than people assume. partly [13:21] because we are very far from [13:22] understanding the brain and we are even [13:25] further away from understanding the [13:26] mind. We assume that the brain somehow [13:30] produces the mind but this is just an [13:32] assumption. We still don't have a [13:34] working model a working theory for how [13:36] it happens. Uh but if it happens, if it [13:40] is possible to directly connect brains [13:43] and computers and integrate them into [13:46] these kinds of cyborgs, nobody has any [13:49] idea what happens next, how the world [13:52] would look like. And it is certainly [13:55] makes it a plausible if again if this is [13:59] this if you reach that point that you [14:02] could have an interbrain net [14:05] the same way that lots of computers are [14:08] connected together to form the internet. [14:10] If you can connect also brains and [14:12] computers directly why can't we then [14:14] connect an interbrain net which connects [14:17] lots of brains as as you as you as you [14:20] uh uh uh described. Again, I I have no [14:23] idea what it means. I think this is the [14:25] point when the way that our [14:30] organic brains understand reality [14:35] u [14:36] even our imagination in the end is the [14:39] product as far as we can tell of organic [14:42] biochemistry. Do you think [14:44] wait so so we we are not equipped I [14:46] think to have a kind of serious [14:49] discussion of what a nonorganic [14:55] brain or a non-organic mind might be [14:59] capable of of doing how it would how it [15:02] would look like and all the basic [15:03] assumptions that we have about brains [15:06] and minds they are limited to the [15:09] organic types. How do you feel about [15:12] artificial intelligence and what's [15:14] happening? This year has been a real [15:15] sort of landmark year in the [15:18] a big leap forward for artificial [15:20] intelligence, the conversation, public [15:22] awareness, [15:24] um the technology itself, the investment [15:27] in the technology, which is always, you [15:29] know, a a very important indicator of [15:31] what's to come. [15:32] Yeah. [15:32] How do you how do you as someone that [15:34] spent a lot of time thinking about this [15:36] emotionally, how do you feel about it? [15:40] uh very concerned. I mean it's moving [15:43] even faster than I expected. Uh when I [15:46] wrote say Homodos in 2016, I didn't [15:48] think we would reach this this point so [15:51] quickly where we are at 2023. And the [15:55] world is is not ready for it. [15:58] And again, it's not AI has enormous [16:01] positive potential. [16:03] We and and this this should be clear. [16:07] And there is no chance of just banning [16:09] AI or stopping all development in AI. I [16:13] tend to speak a lot about the dangers [16:15] simply because you have enough people [16:16] out there, all the entrepreneurs and uh [16:19] uh all the investors talking about the [16:21] positive potential. So it's kind of my [16:23] job to talk about the negative [16:26] potential, the dangers. But it there is [16:29] a lot of positive potential and uh [16:32] humans are incredibly capable in terms [16:36] of adapting to new situations. I don't [16:39] think it's impossible for human society [16:41] to adapt to the new AI reality. The only [16:45] thing is it takes time and apparently we [16:50] don't have that time and people compare [16:52] it to previous big historical [16:56] revolutions like the invention of print [16:59] or the invention of or or the the [17:01] industrial revolution. And you hear [17:03] people say yes when the industrial [17:05] revolution happened in the 19th century. [17:07] So you had all these pro prophecies of [17:10] doom about how industry and the new [17:14] factories and the steam engines and [17:16] electricity how how they will destroy [17:19] humanity or destroy our psychology or [17:21] whatever. And in the end it was okay. [17:25] And when I hear these kinds of [17:27] comparisons as as a historian I'm very [17:30] worried about two things. First of all, [17:33] they underestimate the magnitude of the [17:36] AI revolution. AI is nothing like print. [17:40] It's nothing like uh the industrial [17:42] revolution of the 19th century. It's far [17:44] far bigger. There is a fundamental [17:47] difference between AI and the printing [17:49] press of the steam engine or the radio [17:52] or any previous technology we invented. [17:55] The difference is it's the first [17:57] technology in history that can make [17:59] decisions by itself and that can create [18:02] new ideas by itself. A printing press or [18:06] a radio set could not write new music or [18:11] uh new speeches and could not decide [18:14] what to print and what to broadcast. [18:16] This was always the job of humans. This [18:19] is why the printing press and the radio [18:21] set in the end empowered humanity. [18:24] that you now have more power to [18:27] disseminate your ideas. AI is different. [18:30] It can potentially take power away from [18:32] us. It can decide, it's already deciding [18:35] by itself what to broadcast on social [18:38] media. Its algorithms deciding what to [18:42] promote. [18:43] And increasingly, it also creates much [18:46] of the content by itself. It can compose [18:49] entirely new music. it can compose [18:52] entirely new political manifestos, holy [18:55] books, whatever. Um, so it's a much [18:59] bigger challenge to handle that kind. [19:01] It's it's an independent agent in a way [19:05] that radio and the printing press were [19:07] not. The other thing I find worrying [19:12] about the comparison with say the [19:14] industrial revolution is that yes in the [19:17] end in a way it was okay but to get [19:21] there we had to pass through some [19:24] terrible experiments. [19:27] When the industrial revolution came [19:28] along nobody knew how to build a a [19:33] benign industrial society. So people [19:36] experimented. [19:38] One big experiment was European [19:40] imperialism. Many people thought that to [19:43] build an industrial society means [19:45] building an empire. Unless you have an [19:47] empire that controls the sources of the [19:51] raw materials you need, iron, coal, [19:54] rubber, cotton, whatever. And unless you [19:58] control the markets, you will not be [20:00] able to survive as an industrial [20:02] society. And there was a very close link [20:06] also conceptually [20:08] between building an industrial society [20:11] and building an empire. And all the [20:14] leaders the the the initial leaders of [20:17] the industrial revolution built empires. [20:20] Not just Britain and and France also [20:23] small countries like Belgium also Japan [20:25] when it joined the industrial revolution [20:28] it immediately set about conquering an [20:31] empire. Another tribal experiment was [20:35] Soviet communism. They also thought how [20:38] do you build an industrial society? You [20:41] build a communist dictatorship. And it [20:43] was the same with Nazism. You cannot [20:45] separate communism and Nazism from the [20:48] industrial revolution. You could not [20:50] have created a communist or a Nazi [20:53] totalitarian regime in the 18th century. [20:56] If you don't have trains, if you don't [20:58] have electricity, if you don't have [21:00] radio, you cannot create a totalitarian [21:02] regime. So these are just a few examples [21:06] of the failed experiments. You know, you [21:09] try to adapt to something completely [21:11] new, you very often uh um experiment and [21:16] some of your experiments fail. And if we [21:20] now have to go in the 21st century [21:23] through the same process, okay, we now [21:26] have not radio and and trains, we now [21:29] have AI and bioengineering. And we again [21:32] need to experiment perhaps with new [21:35] empires, perhaps with new totalitarian [21:37] regimes in order to discover how to [21:40] build a benign AI society, then we are [21:44] doomed as a specy. we will not be able [21:47] to survive another round of imperialist [21:51] wars and totalitarian regimes. So [21:54] anybody who thinks hey we've passed [21:56] through the industrial revolution with [21:58] all the prophecies of doom in the end we [22:00] got it right. No if as a historian I I [22:04] would say that I would give humanity a C [22:06] minus on how we adapted to the [22:08] industrial revolution. If we get a C [22:11] minus again in the 21st century that's [22:13] the end of us. It seems quite trivial to [22:18] many that [22:20] the AI revolution has seemed to begun [22:24] with large language models. And when I [22:26] read sapiens, this book I have here, [22:30] language was so central to what made us [22:32] powerful as homo sapiens. [22:34] In the beginning was the word. I didn't [22:37] say it. You know, it's a it's a very [22:40] very widespread idea [22:43] that um ultimately our power is based on [22:46] words. Uh the reason that we controlled [22:49] the world and not the chimpanzeee or the [22:51] elephants is because we had a much more [22:53] sophisticated language [22:55] which enabled us again to tell these [22:57] stories. [22:59] stories about ancestral spirits and [23:02] about guardian gods and about our tribe, [23:07] our nation, which formed the basis for [23:09] cooperation. And because we could [23:12] cooperate, you could have a thousand [23:14] people, a thousand humans cooperating in [23:16] a tribe, whereas the Neandertos could [23:18] cooperate only on the level of say 50 or [23:21] 100 individuals. This is why we rule the [23:23] world and not the Neandertos. And you [23:25] look at every subsequent [23:28] kind of growth in human power and uh you [23:32] see the same thing that uh ultimately [23:36] you tell a story with words and language [23:40] is like the master key that unlocks all [23:44] the doors of our civilization. [23:47] Whether it's cathedrals or whether it's [23:49] banks, they're based on language, on [23:52] stories we tell. [23:54] that again it's very obvious in the case [23:56] of religion [23:58] but also if you think about the world's [24:00] financial system so money has no value [24:04] except in the stories that we tell and [24:06] believe each other if you think about [24:10] gold coins or paper bank notes or [24:13] cryptocurrencies like Bitcoin um they [24:16] have no value in themselves you cannot [24:19] eat them or drink them or do anything [24:21] useful with them but you have people [24:24] telling you very compelling stories [24:26] about the value of these things and if [24:29] enough people believe the story then it [24:32] works. [24:32] They're also protected by language like [24:34] my cryptocurrency is protected by a [24:37] bunch of words. [24:38] Yeah. Uh they're created by words [24:42] and they they function with with words [24:44] and and and and symbols. [24:47] Uh when you communicate with your banker [24:49] it's it's with words. I mean what [24:51] happens when AI can uh uh create deep [24:55] fakes of your everything, your voice, [24:57] your image, uh the the way you talk, the [25:01] type of words you use. So there is [25:03] already an arms race between banks and [25:06] fraudsters. I mean we want the easiest [25:08] communication with our banker. I just [25:10] pick up the phone, I tell a few words, [25:11] and they transfer a million dollars. But [25:14] at the same time, I also want want to be [25:16] protected from an AI that impersonates [25:19] my my voice and tone of tone of voice [25:21] and and whatever. And this is becoming [25:24] difficult. But on a deeper level, again, [25:28] AI could create because money is [25:31] ultimately made of words, of stories. [25:35] AI could create new kinds of money. [25:39] uh the same way that you know [25:40] cryptocurrencies like Bitcoin have been [25:42] created simply by somebody telling [25:44] people a story and enough people finding [25:47] this story convincing. And I I guess as [25:50] a CEO and as an as an entrepreneur, you [25:52] know that if you want to get [25:55] investments, [25:57] what really gets investments is a good [25:59] story. [26:01] And what happens to the financial system [26:04] if increasingly our financial stories [26:07] are told by AI? [26:10] And what happens to the financial system [26:13] and even to the political system [26:16] if AI eventually creates new financial [26:21] devices that humans cannot understand? [26:25] Already today much of the activity [26:30] on the world markets is being done by [26:33] algorithms [26:35] at such a speed and with such complexity [26:38] that most people don't understand what's [26:40] happening there. [26:41] I I if you had to guess what is the [26:44] percentage of people in the world today [26:46] that really understand the financial [26:48] system [26:50] what would be your kind of [26:52] less than 1%. [26:53] Less than 1%. Okay. Let's be kind of [26:55] conservative about it. 1% let's say. [26:57] Okay. [26:58] Fast forward 10 or 20 years. AI creates [27:02] such complicated financial devices that [27:05] there is not a single human being on [27:06] earth that understand finance anymore. [27:10] What are the implications for politics? [27:13] Like you vote for a government but none [27:16] of the humans in the government, not the [27:18] prime minister, not the finance [27:19] minister, nobody understand the [27:21] financial system. They just rely on AI [27:26] to tell them what is happening. [27:29] Is this still a democracy? Is this still [27:31] a a human form of government in any way? [27:34] What do you say to someone that hears [27:35] that and goes, "Ah, that's just that's [27:37] nonsense. That's never going to happen." [27:40] Why not? I mean, let's look back 15 [27:43] years to the last big financial crisis [27:45] in 2007208. [27:48] This financial crisis to a large extent [27:51] began with these extremely complicated [27:54] financial devices CDOS's [27:57] what's the acronym collateral depth [27:59] something I don't even know what the [28:01] word letter stands for you had these [28:03] kind of whiz kids in Wall Street [28:05] inventing a new financial device that [28:08] nobody except them really understood [28:10] which is why also it wasn't regulated [28:12] effectively by the banks and the [28:14] governments and it worked well for a [28:17] couple of years and then it brought down [28:19] the world's financial system [28:22] and um what happens if now AI's comes [28:26] with even more sophisticated financial [28:28] devices [28:29] and for a couple of years everything [28:31] works well they make trillions of [28:33] dollars for us and then one day it [28:36] doesn't one day the system collapses and [28:40] nobody understands what is happening and [28:44] uh again it's not that you didn't go to [28:47] college or whatever. No, it's just [28:49] objectively the complexity of the system [28:53] has reached a point when only an AI [28:58] is able to crunch the numbers, is able [29:00] to process enough data to really get to [29:06] really grasp the shape the dynamics of [29:10] of the financial system. [29:11] We're already there though. You know, I [29:13] think if anyone does understand how the [29:15] financial system works and the markets [29:16] work, it is a bunch of [29:19] homo sapiens relying on a computer to [29:21] tell it something and it it trusting [29:24] that computer's calculations. [29:26] Yeah. And and this will get more and [29:29] more complicated and and and [29:31] sophisticated. And for people who say [29:34] no, it's not going to happen, the [29:36] question is what is stopping it? I mean, [29:40] you know, in all the discussions about [29:42] AI, [29:44] the kind of dangers that draw people's [29:46] attention, like the poster child of AI [29:50] dangers is things like AI creating a new [29:54] virus that kills billions of people, a [29:57] new pandemic. So you a lot of people [29:59] concerned about how do we prevent an AI [30:02] by itself or maybe some small terrorist [30:05] organization or even a 16-year-old [30:07] teenager giving an AI a task to create a [30:12] dangerous virus and release it to the [30:14] world. How do we prevent this? And this [30:15] is a serious concern and we should be [30:17] concerned about it. But this gets a lot [30:20] more attention than the question, how do [30:23] we prevent the financial system from [30:26] becoming so complicated that humans can [30:29] no longer understand it? [30:32] And I see a lot of regulations [30:35] being uh at least considered how to [30:39] prevent AI from creating dangerous new [30:41] viruses. Um I don't see any kind of [30:45] effort to keep the financial system at a [30:49] level that humans understand it. [30:52] Why do you think that is? [30:54] U [30:55] I mean I had a guess. My guess was why [30:58] would the UK [30:59] Mhm. [30:59] cut off then, you know, why would they [31:01] give themselves a disadvantage? [31:02] Exactly. [31:03] When you know there it just means that [31:04] the UK will suffer and if America is [31:06] using a really advanced AI algorithm to [31:08] get ahead, we have to keep up. [31:10] Yeah. It's it's the logic of of the arms [31:13] race. And again, it's not all bad. I [31:15] mean, you have a better financial [31:16] system. Uh you have a more prosperous [31:18] economy. I mean, money isn't bad. I [31:21] mean, it's the basis for almost all [31:23] human cooperation. [31:25] And a lot of financial devices in the [31:27] end, if you think what are they, they [31:30] are devices to establish trust between [31:33] people, especially trust between [31:35] strangers. And money in essence is a [31:38] device for establishing trust. I don't [31:41] know you, you don't know me, but we both [31:43] trust this gold coin or piece of paper [31:47] so we can cooperate on uh uh uh sharing [31:51] food or creating a a medicine. [31:55] And the most sophisticated financial [31:57] devices, they basically do the same [31:59] thing. Stocks and bonds and these [32:02] CDOS's, they are a method to establish [32:05] trust. And when you open a new bank [32:08] account, the most important thing is how [32:11] do I trust the bank to really uh h take [32:16] care of my money and to follow my [32:17] instructions but not to be open to fraud [32:20] and things like that. And again, you as [32:22] as as an investor [32:24] um when you try to get money from from [32:26] from or you as an entrepreneur when when [32:29] you try to get money from investors, the [32:31] biggest issue is always trust. [32:34] And if somebody can comes up can can [32:37] come up with a new uh uh way to [32:40] establish trust between people, that's a [32:43] good thing. But if this new way [32:49] increasingly depends on non-human [32:52] intelligence on again on systems that [32:56] humans cannot understand. That's the big [32:59] question. What happens to human society [33:02] when the trust that is at the basis of [33:07] all social interactions [33:09] is actually no longer trust in humans. [33:14] It's trust in a non-human intelligence [33:18] that we don't fully understand and that [33:20] we cannot anticipate. And part of the [33:24] problem with regulating AI or AI safety, [33:28] it goes back to what we discussed [33:30] earlier that AI is different from [33:33] printing presses or radio sets or even [33:35] atom bonds. [33:37] If you want to make nuclear energy safe, [33:42] then you need to think about all the [33:44] different ways that uh I don't know a [33:47] nuclear power station can uh uh uh can [33:50] have an accident. [33:52] And I guess there is a limited number of [33:55] things that can go wrong. And ideally if [33:59] if you think hard, if you have if you [34:01] have enough people thinking hard enough, [34:04] you can make safe nuclear reactors, safe [34:08] nuclear power stations. [34:10] Now, but AI is fundamentally different [34:14] because AI keeps changing. It keeps [34:17] reacting to the world. It keeps reacting [34:20] to you coming up with new inventions, [34:24] new ideas, new decisions. So making AI [34:28] safe is a bit like making a nuclear [34:30] reactor safe taking into account the [34:32] fact that the nuclear reactor can decide [34:35] to change [34:36] in ways that you can't anticipate and [34:39] even worse it can react to you. So if [34:44] you build a particular safety mechanism [34:46] for the nuclear reactor, what happens if [34:49] the nuclear reactor say oh they build [34:51] this mechanism let's do that to h [34:54] somehow get around the safety mechanism. [34:57] We don't have this problem with nuclear [34:58] reactors. But this is the problem with [35:01] AI. We are trying to contain something [35:04] which is an independent agent and which [35:08] might actually come to understand us [35:11] better than we understand it. I'm really [35:15] curious about how this will impact you [35:19] know you talked about elected officials [35:21] there and how their systems will be sort [35:22] of um dri their financial decision-m [35:26] might be driven by algorithms [35:28] but government's an authority itself [35:32] I've pondered recently whether there'll [35:34] come a day in the notsodistant future [35:37] where we might vote for an algorithm [35:40] where we might vote for an AI to be our [35:43] government. Is that crazy thinking? [35:46] I think we we we're quite a long way off [35:48] from there. We would still want humans [35:51] at least in the symbolic role of being [35:54] the prime minister, the the member of [35:56] parliament, whatever, the president. [35:59] The real problem is that increasingly [36:02] these humans could be kind of [36:05] figureheads or or puppets when the real [36:08] decisions, the most consequential [36:11] decisions are uh are made by algorithms. [36:15] be partly because the the the um it will [36:19] just be too complicated for the humans [36:23] at the top to understand the situation [36:28] or to understand the different options. [36:30] So going back to the financial example. [36:32] So imagine that you know it's it's 4:00 [36:34] in the morning. There is a phone call uh [36:37] to the prime minister from the finance [36:40] algorithm [36:42] telling the P the prime minister that we [36:44] are facing a financial meltdown [36:47] uh and that we have to do something [36:50] within the next I don't know 30 minutes [36:52] to prevent a national or global [36:56] financial meltdown. And there are like [36:58] three options and the algorithm [37:00] recommends option A and there is just [37:03] not enough time to explain to the prime [37:05] minister how did the algorithm reach the [37:08] conclusion and even what is the meaning [37:11] of these different options [37:14] and again people think about this [37:17] scenario mostly in relation to war. [37:20] Mhm. that what happens if you have an [37:22] algorithm in charge of the your security [37:26] system and it it alerts you to a massive [37:30] incoming cyber attack and you have to [37:33] react immediately and this could if you [37:36] react in in a specific way this could [37:38] mean war with another nation but you [37:41] just don't have enough time to [37:44] understand how the algorithm reached the [37:47] decision and how the algorithm was also [37:51] able to determine that of the all the [37:52] different options, this is the best [37:54] option. [37:56] Do you think that humans believe we're [37:57] more complicated and special than we [37:59] actually are? [38:01] Because I think part of much many many [38:03] of the rebuttals when we talk about [38:04] artificial intelligence stem back to [38:07] this idea that we're in, you know, we're [38:09] like innately genius, creative, [38:13] spiritual, special, [38:17] you know, um artificial intelligence [38:22] like our our intelligence is somewhat [38:24] divine or [38:27] we've got free will and you know we [38:29] Yeah. Yeah, I mean it's [38:32] if the argument is we have free will, we [38:37] have a divine soul and therefore no [38:41] algorithm will will ever be able to [38:43] understand us and to predict our [38:46] decisions or to manipulate us then this [38:49] is a very common argument but it's [38:50] obviously nonsensical. I mean even [38:53] before AI uh it was uh even with [38:57] previous technology it was possible to a [39:01] large extent to predict people's [39:03] behavior and to manipulate them and AI [39:07] just takes it to the next level. Now [39:10] with regard to the discussion of of free [39:12] will [39:13] my my position is you cannot start with [39:17] the assumption that humans have free [39:20] will. If you start with this assumption [39:24] then it's uh actually is very [39:28] it it makes you very incurious lacking [39:32] curiosity [39:33] about about yourself about human beings. [39:37] It kind of closes off the investigation [39:39] before it began. [39:41] Um you assume that any decision you make [39:44] is just a result of my free will. Why [39:47] did I choose this politician, this [39:49] product, uh uh this spouse? Because it's [39:52] my free will. And if this is your [39:55] position, there is nothing to [39:57] investigate. You just assume you have [39:59] this kind of divine spark within you [40:02] that makes all the decisions and there [40:04] is nothing to investigate there. [40:08] Um I would say no start investigating [40:12] and you'll probably discover that there [40:15] are a lot of factors whether it's [40:18] external factors like cultural [40:21] traditions and also internal factors [40:24] like biological mechanisms that shape [40:27] your decisions. you chose this [40:29] politician or this spouse because of [40:33] certain cultural traditions and because [40:36] of certain biological mechanisms, your [40:39] DNA, your uh uh brain structure, [40:42] whatever. And this actually makes it [40:45] possible for you to get to know yourself [40:48] better. Now if after a long [40:51] investigation [40:53] you've reached the conclusion that yes [40:55] there are cultural influences, there are [40:59] political influences, there are genetic [41:02] and neurological influences, but still [41:05] there is a certain percentage of my [41:09] decision that cannot be explained by any [41:11] of these things. Then okay, call it free [41:14] will and we can discuss it. But don't [41:18] start with this assumption because then [41:21] you lose the incentive to explore [41:24] yourself. [41:26] And anybody who embarks on such a [41:29] process of self exploration, whether [41:32] it's in therapy, whether it's in [41:34] meditation, whether it's in the [41:36] laboratory of a brain scientist or uh as [41:40] a historian in the archive, you will be [41:43] amazed to discover how much of your [41:47] decisions are not the result of some [41:50] mystical free will. They are the result [41:53] of cultural and biological factors. And [41:56] this also means that you are vulnerable [42:00] to being deciphered and manipulated [42:04] by political parties, by corporations, [42:08] by AI. People who have this kind of [42:11] mystical belief in free will are the [42:13] easiest people to manipulate [42:16] because they don't think they can be [42:18] manipulated. [42:20] Uh and obviously they can. We humans [42:23] should get used to the idea that we are [42:25] no longer mysterious souls. We are now [42:28] hackable animals. That's what we are. [42:31] You said that at the World Economic [42:33] Forum. [42:34] Yeah. Again, this is the same point [42:38] basically that it's now possible to hack [42:40] human beings. Not just to hack our [42:42] smartphones, our bank accounts, our [42:44] computers, but to really hack our [42:46] brains, our minds, and to uh uh predict [42:50] our behavior and manipulate our behavior [42:53] more than in any previous time in [42:55] history. [42:56] The other line that you said uh which [42:58] really made me think and ponder was [43:01] um [43:03] as previously human life was about the [43:05] drama of decision-m and without this we [43:07] won't have a meaning in life. [43:10] Yeah. [43:11] that if you look, you know, at politics, [43:14] at religion and at at culture, [43:17] people told the stories about their [43:20] lives or the lives of people in general [43:23] as a kind of of drama of decision [43:26] making. [43:26] Mhm. [43:27] That you reach a particular junction in [43:30] life and you need to choose you need to [43:33] choose between good and evil. You need [43:36] to choose between political parties. You [43:38] need to choose your what to study at [43:40] university or where to work, what kind [43:43] of job to to to apply to. [43:46] And our stories revolved around these [43:49] decisions. [43:52] And what happens to human life if [43:55] increasingly the power to make decisions [43:59] is taken from us? [44:01] And increasingly it's algorithms [44:04] making all these decisions for us or [44:08] about us. [44:09] Is that possible? [44:10] It's already happening. Increasingly, [44:12] you know, you apply to a bank to get a [44:14] loan. In many places, it's no longer a [44:17] human banker who is making this decision [44:21] about you whether to give you a loan, [44:23] whether to give you a mortgage. It's an [44:24] algorithm analyzing [44:28] billions of bits of data about you and [44:30] about of millions of other customers or [44:33] previous loans determining whether you [44:36] are creditw worthy or not. And if you [44:39] ask the bank if they refuse to give you [44:42] a loan and you ask the bank why didn't [44:44] you give me a loan and the bank says we [44:47] don't know. the the computer said no and [44:49] we just believe our our our computer our [44:52] algorithm and it's happening also in the [44:55] judicial system increasingly that uh um [44:59] various judicial decisions verdicts like [45:02] for how many like the judge decided that [45:05] you committed some crime the sentence [45:09] whether to send you to two months or [45:11] eight months or two years in prison is [45:13] increasingly determined by an algorithm [45:17] uh you apply to a place at university, [45:19] you apply to a job. This too is [45:22] increasingly decided by algorithms. [45:24] Dating [45:26] uh dating. Yes. I mean even um even un [45:30] unknown unbeknownst to you, the [45:33] algorithms of the dating apps that [45:36] you're using are shaping your romantic [45:40] life. But what in a world of you know [45:43] robotics and artificial intelligence why [45:45] do I need to find a person at all? [45:48] Why not just have a relationship with [45:49] with a robot or with an AI? [45:52] Yeah. Uh we do see the beginning of of [45:56] of this that people are building more [45:59] and more intimate relationships with [46:02] non-human intelligences with AIs and [46:05] bots and and so forth. And this raises a [46:09] lot of of difficult and and profound [46:13] questions. Now, part of the problem is [46:16] that the AIS are built to mimic intimacy [46:23] that the the ability intimacy is an [46:27] extremely powerful thing. Not just in [46:29] romance, also in the market, also in [46:33] politics. If you want to change [46:35] somebody's mind about anything, [46:38] political issue, a commercial uh [46:40] preference, intimacy is kind of the most [46:45] powerful weapon. [46:47] And [46:49] somebody you really trust, somebody you [46:52] have intimate relationships with will be [46:54] able to change your views on a lot of [46:57] things more than uh someone you see on [46:59] TV or just an an article you read in [47:03] newspaper. There is a huge incentive for [47:06] the creators of AIS to create AIS that [47:10] are able to forge intimate relationships [47:12] with humans. [47:15] And um this makes us extremely [47:17] vulnerable to this new type of [47:21] manipulation [47:23] that was previously just unimaginable [47:25] cuz loneliness is at you know all-time [47:28] highs especially in the sort of western [47:30] world and [47:31] sexlessness and I I was reading some [47:33] stats about how the like body bottom 50% [47:36] of men in particular are having almost [47:39] no sex relative to the top sort of 10% [47:42] and you think you know this disparacy [47:44] the rise of digitalization, loneliness, [47:46] we're in our homes on screens more than [47:48] ever before. [47:49] And then you hear about this industry of [47:51] AI and sex dolls and all this and you [47:53] just wonder, you play it forward and go, [47:55] yeah, it's it's going there. And and the [47:57] thing is that it it's not that that the [48:01] humans are so stupid or something that [48:04] they they they kind of project something [48:07] onto the AI and fall in love with an AI [48:09] chatbot. The AI is deliberately [48:14] built, created, trained to fool us. To [48:19] the same way, you know, you look at the [48:21] previous 10 years, there was a big [48:25] battle for human attention. There was a [48:27] battle between different social media [48:30] giants and what whatever how to grab [48:33] human attention and they created [48:36] algorithms that were really amazing at [48:41] grabbing people's attention [48:43] and now they are doing the same thing [48:45] but with intimacy and we are extremely [48:49] exposed. We are extremely vulnerable to [48:51] it. Now the big problem is and and again [48:54] this is where it it gets kind of really [48:56] philosophical [48:58] that [49:00] what humans really want or need from a [49:03] relationship is to be in touch with [49:07] another conscious entity. [49:10] H an intimate relationship is not just [49:13] about providing my needs. [49:17] Then it's exploitative. Then it's [49:18] abusive. If you're in a relationship and [49:21] the only thing you think about is how [49:24] how would I feel better? How would my [49:27] needs be provided for? Then this is a [49:29] very abusive situation. [49:32] Uh a a really healthy relationship is [49:35] when it goes both ways. You also care [49:38] about the feelings and the needs of the [49:41] other person of the other entity. [49:45] Now [49:46] what happens if the other entity has no [49:49] feelings, has no emotional needs because [49:53] it it has no consciousness. That's the [49:55] big question. [49:57] And there is a huge confusion between [50:00] consciousness and intelligence. AI is [50:03] artificial intelligence. [50:06] But what exactly is the relation between [50:09] intelligence and consciousness? [50:12] Now intelligence is the ability to solve [50:15] problems, [50:17] to win a chess, to invest money, to [50:20] drive a car. This is intelligence. [50:22] Consciousness is the ability to feel [50:24] things like pain and pleasure and love [50:27] and hate and sadness and anger and and [50:30] so many other things. Now in humans and [50:33] also in other mammals, intelligence and [50:36] consciousness actually go together. We [50:39] solve problems by having feelings. [50:43] But computers are fundamentally [50:45] different. They are already more [50:49] intelligent than us in at least several [50:52] narrow fields, [50:54] but they have zero consciousness. [50:58] They don't feel anything. When they beat [51:02] us at chess or go or some other game, [51:05] they don't feel joyful and happy. If [51:07] they make a wrong move, they don't feel [51:10] sad or or angry. They have zero [51:13] consciousness. As far as we can tell, [51:16] they might soon be far more intelligent [51:20] than us and still have zero [51:24] consciousness. Now what happens when you [51:27] are in a relationship with an entity [51:30] which is far more intelligent than you [51:34] and can also imitate mimic [51:38] consciousness. It it knows how to solve [51:41] the problem of making you feel as if it [51:46] is conscious [51:48] but it still has no feelings of its own. [51:52] And this is a very disturbing vision of [51:56] the future. [51:58] It opens us up to manipulation. Is that [52:00] what you're saying? [52:01] It first of all it opens us to [52:03] manipulation but also it uh uh the the [52:08] big question what does it mean for the [52:10] health of our own mind of our own [52:14] psyche? [52:16] If we are in a relationship or or many [52:19] of our important relationship in life [52:21] are with non-concious entities [52:25] that uh that they don't really have any [52:28] feelings of their own. Again, they are [52:29] very good [52:30] at faking [52:31] at faking it. They're very good at [52:33] catering to our feelings, [52:36] but um again it's just it's just [52:41] manipulation in the end. Are you [52:43] optimistic about the happiness of humans [52:45] going forward? Or do you think happiness [52:47] will take its own? You know, I've heard [52:49] you talk about how happiness might just [52:51] become a bio biochemical, [52:53] I don't know, prescription or something. [52:56] Yeah. I mean, we don't have a good track [52:59] record with regard to happiness. If you [53:02] look at the last 100,000 years from say [53:05] the stone age until the 21st century, [53:08] you see a dramatic rise in human power. [53:12] We are thousands of times more powerful [53:15] as a species and as individuals than we [53:18] were in the stone age. We are not [53:21] thousands of times happier. We just [53:24] don't really know how to translate power [53:28] into happiness. And this is very clear [53:31] when you look at the lives of the most [53:33] powerful people in the world [53:36] that there is no correlation between how [53:39] rich and powerful you are and how happy [53:42] you are as as as as a person. I mean I I [53:46] don't have the I don't get the [53:48] impression that people like I don't know [53:51] Vladimir Putin or Elon Musk are the [53:53] happiest people in the world [53:56] even though they are they are some of [53:57] the most powerful people in the world. [54:01] So there is no reason to think that as [54:03] humanity gets even more powerful in [54:06] coming decades we will get any happier. [54:10] And understanding happiness is about [54:13] understanding the deep dynamics of of [54:16] not not even the brain but of the mind [54:20] of consciousness [54:21] and we are just not there yet. [54:25] Um we are very very good and and the [54:27] related problem is that humans usually [54:31] understand how to manipulate something [54:34] long before they understand the [54:37] consequences of the manipulations. [54:40] If you look at the outside world, at the [54:44] ecological system, we have learned how [54:47] to cut forests, how to build huge dams [54:51] over rivers long before we understood [54:55] what will be the consequences for the [54:57] ecological system. Which is why we now [55:01] have this ecological crisis. We [55:03] manipulated the world without [55:06] understanding the consequences. [55:09] As something similar might happen with [55:12] the world inside us, [55:15] with more powerful medicines, with brain [55:18] computer interfaces, with genetic [55:21] engineering and and so forth, we are [55:24] gaining the power to manipulate our [55:27] internal world, the world within us. [55:30] But again, the power to manipulate is [55:34] not the same thing as understanding the [55:37] complexity of the system and the [55:39] consequences of the manipulation. [55:41] A related manipulation there is [55:43] immortality and our pursuit of it. I've [55:45] sat with people on this podcast who are [55:47] committing their lives to staying alive [55:48] forever. And there's a through line [55:51] there between our desire to be immortal, [55:54] you know, the rise in the scientific [55:57] discoveries that are enabling that and [55:59] our happiness. I I've often thought, you [56:01] know, much of the reason why things are [56:04] special in my life is because they're [56:05] scarce, [56:06] including my time. [56:07] Yeah. [56:08] And I I always I almost wonder about the [56:10] psychological um [56:13] issues I would face if I knew I was [56:16] immortal. Like if I knew that [56:19] the partner I'm with doesn't come at the [56:22] expense of another one I can be with, [56:24] you know, at 30 years old. [56:26] And the car, you know, the choices you [56:28] make, I think what makes them scaled [56:30] are their scarcity. [56:32] Mhm. [56:32] Against the backdrop of an of a finite [56:35] life. [56:37] Uh yeah, it will definitely change [56:38] everything if you think about relations [56:40] between parents and children. So if you [56:43] live forever, so the 20 years you raised [56:47] uh uh you spent raising somebody 2,000 [56:49] years ago, what do they mean now? But I [56:52] think long before we get to that point, [56:55] I mean, most of these people are going [56:57] to be incredibly disappointed because it [57:00] will not happen within their lifetime. [57:02] Another related problem is that we will [57:05] not get to immortality. We will get to [57:08] something that maybe should be called a [57:10] mortality. [57:12] that immortality is that like you're [57:14] you're God, you can never die no matter [57:17] what happens. It's even if we solve [57:21] cancer and Alzheimer and demensia and [57:24] whatever, we will not get there. We will [57:27] get to kind of a life without a [57:30] definitive expiry date that you can live [57:34] indefinitely. You can go every 10 years [57:37] to a clinic and get yourself rejuven [57:40] rejuvenated, but if a bus runs you over [57:45] or your airplanes explodes or a [57:48] terrorist kills you, you're dead and [57:51] you're not coming back to life. Now, [57:55] realizing that you have a chance to live [57:58] forever, but if there is an accident, [58:00] you die. This creates a level of anxiety [58:04] and terror unlike anything that we know [58:07] in our own lives. I think the people who [58:10] will will be in that situation will be [58:13] extremely anxious and miserable. [58:17] And another issue is you know people [58:21] often spend so much effort trying to get [58:26] gain something get something without [58:29] really understanding what are they going [58:32] why what will you do with it what is so [58:34] good about it you know like people spend [58:36] so much effort to to get have more and [58:38] more money instead of thinking what will [58:41] I actually do with that money so it's [58:44] the same with you know the people who [58:46] want to extend life forever. What is so [58:49] good about life that what will you do [58:52] with it? [58:53] And if you know it, why don't you do it [58:56] already? That uh you know I hear people [58:59] saying about how how precious human [59:02] consciousness is [59:05] why why do you think it's so precious [59:07] and whatever it is, why don't you do it [59:12] right now? I mean why spend your life [59:16] developing some kind of treatment that [59:20] will uh extend your consciousness for a [59:24] thousand years. [59:26] Just spend your time doing now whatever [59:31] you think you would be doing with your [59:33] consciousness a thousand years from now. [59:35] So if they were to say but it'll give me [59:37] more time with my family. You're saying [59:39] just instead of wasting your time just [59:40] like [59:41] Exactly. So, you know, somebody who has [59:43] no time for their family at all right [59:45] now because they are busy developing the [59:48] kind of uh uh uh miracle cure that will [59:51] enable them to spend time with with [59:53] their family in 200 years. This makes no [59:56] sense. [59:59] I think about the disparity that [60:01] artificial intelligence and these forms [60:03] of sort of bioengineering might create [60:05] because it's conceivable that the rich [60:08] will gain access to these technologies [60:10] first. [60:10] Yeah. And then, you know, when we think [60:12] about bioengineering, [60:15] being able to sort of play with our [60:16] genetic code, that means if I, for [60:18] example, managed to get my hands on some [60:20] kind of bio engineering treatment to [60:23] make sure that my kids were maybe a [60:25] little bit smarter, maybe a little bit [60:27] stronger, whatever, then you're going to [60:29] start a sort of genetic chain of [60:32] modified children that are superior in [60:35] intelligence and strength and whatever [60:37] else might be desirable. M [60:38] and then you have this disparity in [60:40] society where you have like the you know [60:42] one humans one set of humans are on a [60:44] completely different exponential [60:45] trajectory and the other humans are you [60:48] know [60:49] yeah behind [60:51] this is extremely dangerous uh I think [60:54] we just shouldn't go there that we [60:57] shouldn't invest a lot of resources [61:02] efforts in developing these kinds of uh [61:06] upgrades and enhancements [61:08] that are very likely, at least at first, [61:12] to be the preserve of a small elite and [61:16] to translate economic inequality into [61:20] biological inequality and to basically [61:23] split the human species to to split homo [61:27] sapiens into, you know, a ruling class [61:30] of superhumans and and the rest of us. [61:33] This is a very very dangerous [61:35] development. [61:37] related to that is the problem that I [61:41] don't think it will be these will be [61:44] upgrades at all [61:47] what worries me is that a lot of these [61:49] things will turn out actually to be [61:51] downgrades [61:53] that [61:55] we again we don't understand [61:58] our bodies our brains our minds well [62:00] enough to know what will be the [62:03] consequences of tweaking our genetic [62:07] code or of um I don't know implanting [62:11] all kinds of devices into our brains. [62:15] People who think that this will enable [62:18] them let's say to upgrade their [62:21] intelligence [62:24] they don't know what the side effects [62:26] will be. It could be that the same [62:28] treatment that increases your [62:31] intelligence also decreases your [62:34] compassion or your spiritual depth or [62:38] whatever. And the danger is that [62:40] especially if this technology is in the [62:42] hands of powerful corporations, armies, [62:47] governments, [62:50] they will enhance [62:52] those qualities that they want like [62:56] intelligence and like discipline [62:59] while disregarding [63:01] uh other qualities which could be even [63:04] more important for for human flourishing [63:07] like compassion. [63:09] or like autistic sensitivity or like [63:11] spirituality. If I think about somebody [63:13] again like Putin, what would he do with [63:16] this type of technology then yes, he [63:18] would like an army of super intelligent [63:21] and super loyal soldiers. And if these [63:24] soldiers do don't have any compassion or [63:28] any spiritual depth, all the better for [63:30] him. [63:30] But that speaks to the arms race. And [63:32] you know, you said you we think we [63:33] shouldn't, but China will see that as an [63:36] opportunity or Putin will see that as an [63:38] opportunity if the if the Western world, [63:40] if the United States or the UK don't. [63:44] And so again, it comes back to this [63:45] point of, you know, we're screwed if [63:47] we're damned if we do, we're damned if [63:48] we don't. [63:49] I'm not sure that in this case it it [63:51] works. uh because again a lot of these [63:54] upgrades are likely to have um [63:58] detrimental side effects both for the [64:02] person in question and for the society [64:04] as a whole. And I think that in this [64:07] case societies that will choose to be uh [64:12] uh [64:13] progress more slowly and safely they [64:16] will actually have an advantage. It's [64:18] like if you say, you know, there is some [64:20] other country where they don't have any [64:22] brakes on their on their cars and they [64:25] don't have any seat belts and they [64:27] release new medicines without checking [64:29] their side effects. They're moving so [64:32] fast. We are left behind. No, it makes [64:34] no sense to to to imitate them. This [64:36] will actually ruin their societies. You [64:39] don't want to imitate these kinds of of [64:42] harmful effects. Uh with development of [64:45] AI, it's different. I think there the [64:50] advantages in things like finance, like [64:53] the military will be so big that an AI [64:57] AI arms race is almost inevitable. [65:01] But with trying to kind of bioengineer [65:04] humans, if you go too fast, it will be [65:09] this self-destructive. [65:11] So we can take it most slowly and safely [65:14] and without being kind of left behind in [65:18] an arms race. [65:19] You said on the Tim Ferris podcast, the [65:21] best scenario is that homo sapiens will [65:22] disappear but in a peaceful and gradual [65:24] way and be replaced by something better. [65:26] It's quite a um uncomfortable statement [65:29] to to listen to. [65:32] I think that again the the the type of [65:34] technologies that we are now developing [65:36] when you combine them with the human [65:39] ambition [65:41] to uh um you know to to improve [65:45] ourselves [65:47] it's almost inevitable that we will use [65:51] these technologies to change ourselves. [65:54] The question is whether we will do it [65:57] slowly and responsibly enough for the [66:00] consequences to be beneficial. But the [66:03] idea that we can now develop these [66:06] extremely powerful tools of [66:08] bioengineering and AI and remain [66:12] the way we are. We'll still be the same [66:15] homo sapiens in 200 years, in 500 years, [66:18] in 1,000 years. we'll have all these [66:21] tools to connect brain to computers to [66:23] to kind of re-engineer our genetic code [66:26] and we won't do it. I think this is [66:29] unlikely. [66:30] One of the outstanding questions that I [66:31] have and one of the sort of observations [66:33] I've had is people like Sam Alman um the [66:36] founder of OpenAI that made Chat GPT [66:39] started working on universal basic [66:41] income products like Worldcoin. And I [66:43] thought, you know what, that's curious [66:45] that the people that are at the very [66:46] forefront of this AI revolution are now [66:49] trying to solve the second problem they [66:52] see coming, which is people not having [66:54] jobs. Yeah. Essentially, [66:55] is is that do you think that's a because [66:58] you know every I've spoken a lot this [67:00] year on stages and this is one of the [67:01] questions I always get asked is the [67:03] implications of AI on the and jobs as we [67:05] know it in the workforce. Mhm. [67:07] Is it realistic to believe that most [67:09] jobs will disappear as we know them [67:13] today? [67:14] I think [67:16] many jobs, maybe most jobs will [67:19] disappear, but new jobs will emerge. You [67:23] know, most jobs that people do today [67:25] didn't exist 200 years ago. [67:27] Mhm. Like this. [67:28] Uh yeah, like this. Like doing a [67:30] podcast. [67:31] And there will be new jobs. [67:34] The really big problem will be how to [67:36] retrain people. [67:39] Uh it demands a lot of financial support [67:42] also psychological support for people to [67:46] kind of relearn, retrain, reinvent [67:50] themselves and doing it not just once [67:52] but repeatedly throughout their career [67:55] throughout their lives. The AI [67:57] revolution will not be a single [67:59] watershed event like you have the big AI [68:02] revolution in 2030. You lose 60% of [68:06] jobs. You create lots of new jobs. You [68:09] have 10 difficult years. Everybody [68:11] adjusting, adapting, reskilling, [68:14] whatever, and then everything settles [68:16] down to a new equilibrium. It won't be [68:18] like that. AI is nowhere near its full [68:21] potential. So you will have a lot of [68:23] changes by 2030, even more changes by [68:27] 2040, even more changes by 2050. You [68:30] will have new jobs, but the new jobs too [68:33] will change and disappear. [68:35] What new jobs? [68:37] In a world where intelligence is [68:38] disrupted, what what jobs are left? [68:40] Because you say you're going to retrain [68:42] me. I'm like, you know, I'm not going to [68:44] be able to keep up with an AI that's [68:45] retraining every second. [68:48] And I I'm not sure. I mean some of the [68:50] answers might be counterintuitive [68:53] that um [68:56] at least at present we see that AI is [68:59] extremely good at automating jobs that [69:02] only require cognitive skills but they [69:06] are not good at jobs that require motor [69:09] skills and social skills. So if you [69:12] think about say doctors and nurses, so [69:14] at least those types of doctors who are [69:19] only doing cognitive work, they [69:23] read articles, they get your medical [69:27] results, all kinds of tests and and and [69:29] and and whatever. They diagnose your [69:31] disease and they decide on a course of [69:34] treatment. This is purely cognitive [69:36] work. This is the easiest thing to [69:38] automate. [69:39] But if you think about a nurse that has [69:42] to replace a bandage to a crying child, [69:46] this is much more difficult to automate. [69:50] You don't think that's possible to [69:51] automate? [69:52] I I think it is possible, but not now. [69:54] You need very delicate motor skills and [69:57] also social skills to do that. [70:00] Did you see Elon's video the other day [70:01] with um the Tesla robot? [70:04] I see a lot of these videos. It's it's [70:06] getting the egg and it's cracking the [70:08] egg and it's going like this. [70:10] No, again I'm not saying it's [70:11] impossible. I'm just saying it will take [70:14] longer. It's more difficult. Again, [70:16] there is also the social aspect. If you [70:18] think about self-driving vehicles, the [70:21] biggest problem for self-driving [70:23] vehicles is humans. I mean, not the not [70:26] just the the human drivers, it's the [70:28] pedestrians, it's the it's the [70:30] passengers. How do you deal with a [70:33] drunken passenger? whatever. [70:36] So, uh, again, it's not impossible, but [70:38] it's much more difficult. So, again, I [70:41] think that there will be new jobs, at [70:43] least in the foreseeable future. The [70:46] problem will will be to retrain people. [70:49] And the biggest problem of all will be [70:52] on the global level, not on the national [70:54] level. I when I hear people talk about [70:57] universal basic income, the first [71:00] question to ask is, is it universal or [71:02] national? [71:04] Is it a system that let's say raises [71:09] taxes on big tech corporations in [71:12] Silicon Valley in California and uses [71:15] the money to provide basic services and [71:19] also retraining courses for people in [71:24] Ohio and Pennsylvania? [71:27] Uh or does it also apply to people in [71:30] Guatemala and Pakistan? [71:33] I mean, what happens when it becomes [71:35] cheaper to produce shirts with robots in [71:39] California than in Guatemala and in [71:41] Mexico? Uh, does Sam Alman has a vision [71:46] of the US government raising taxes in [71:48] California and sending the money to [71:50] Guatemala to support the people there? [71:53] If the answer is no, we are not talking [71:56] about universal basic income. We are [71:57] only talking about national basic income [71:59] in the US. Then what happens to the [72:02] people in Guatemala? That's the that's [72:04] the biggest question. [72:06] And a sub question to that is about how [72:08] one should be educating our our children [72:11] and our education institutions as they [72:13] are today. Because with what's to come, [72:15] um makes me wonder what what skill would [72:18] be worth investing you know 10 12 years [72:21] into a child that I had. [72:23] Um [72:25] nobody has any idea. I mean if you think [72:28] about specific skills [72:30] then this is the first time in history [72:33] when we have no idea how the job market [72:36] or how society would look like in 20 [72:38] years. So we don't know what specific [72:41] skills people will need if you think [72:44] back in history. So it was never [72:47] possible to predict the future but at [72:50] least people knew what kind of skills [72:53] will be needed in a couple of decades. [72:55] If you live, I don't know, in England [72:58] in uh uh 1023, [73:02] a thousand years ago, you don't know [73:05] what will happen in in 30 years. Maybe [73:08] the Normans will invade or the Vikings [73:11] or the Scots or whoever. Maybe there'll [73:14] be an earthquake. Maybe there'll be a [73:16] new pandemic. Anything can happen. You [73:18] can't predict. But you still have a very [73:22] good idea of how the economy would look [73:25] like and how human society would look [73:27] like in the 1050s or the 1060s. You know [73:32] that most people will still be farmers. [73:35] You know it's a good idea to teach your [73:37] kids how to uh harvest wheat, how to [73:41] bake bread, how to ride a hose, how to [73:44] shoot and bow an arrow. These things [73:46] will still be necessary in 30 years. If [73:49] you now look 30 years to the future, [73:52] nobody has any idea what kind of skills [73:56] will be needed. If you think for [73:58] instance, okay, this is the age of AI, [74:01] computers, I will teach my kids how to [74:03] code computers. Maybe in 30 years, [74:06] humans no longer code anything because [74:09] AI is so much better than us at writing [74:12] code. [74:13] Um, so what should we focus on? I would [74:17] say the only thing we can be certain [74:18] about is that 30 years from now the [74:22] world will be extremely volatile, [74:25] extremely it will keep changing at an [74:27] ever rapid pace. [74:29] Do you think this is going to incre [74:31] increase the amount of conflict [74:33] because I watched a video on your [74:34] YouTube channel where you said the [74:36] return of wars? [74:36] Yeah. Uh that's one of the dangers that [74:40] there is and we see it all all over the [74:42] world now. uh like 10 years ago we were [74:45] in the most peaceful era in human [74:47] history and unfortunately this era is [74:51] over. We are now in a new era of wars [74:55] and potentially of imperialism [74:58] and we are seeing it all over the world [75:00] uh with the Russian invasion of Ukraine [75:02] now with the war in the Middle East uh [75:05] Venezuela and Guyana some East Asia war [75:09] is is is back on the table. It's not [75:13] just because of the rapid changes and [75:16] the upheavalss they cause. It's also [75:19] because um you know 10 years ago we had [75:23] a global order, the liberal order which [75:26] was far from perfect but it's still kind [75:30] of regulated relations between nations [75:33] between countries [75:35] based on an idea on on the liberal [75:39] worldview that despite our national [75:42] differences all humans share certain [75:46] basic experiences and needs and [75:49] interests. Which is why it makes sense [75:52] for us to work together to diffuse [75:55] conflicts and to uh uh um solve our [76:00] common problems. It was far from [76:03] perfect, but it did create the most [76:06] peaceful era in human history. [76:08] Then this order was repeatedly attacked [76:12] not only from outside [76:14] from forces like Russia or North Korea [76:18] or Iran that never accepted this order [76:22] but also from the inside even from the [76:24] United States uh which was the architect [76:27] to a large extent of of this order with [76:30] the election of Donald Trump which says [76:32] I don't care about any kind of global [76:35] order. I'd only care about my own [76:38] nation. And you see this way of thinking [76:42] that I only care about my the interests [76:44] of my nation more and more around the [76:47] world. Now the big question to ask is if [76:51] all the nations think like that what [76:54] regulates the relations between them [76:58] and there was no alternative nobody came [77:02] up with with the and said okay I don't [77:04] like the liberal liberal global order I [77:07] have a better suggestion [77:09] for how to manage relations between [77:12] different nations. [77:13] They just destroyed the existing order [77:17] without offering an alternative. And the [77:20] alternative to order is simply disorder. [77:24] And this is now where we find ourselves. [77:27] Do you think there's more wars on the [77:28] way? [77:29] Yes. Unless unless we reestablish order, [77:33] there will be more and worse wars uh [77:37] coming in the next few years in more and [77:39] more areas around the world. [77:42] You see defense budgets all over the [77:45] world uh uh skyrocketing [77:48] and this is a vicious circle. When your [77:51] neighbors increase their military [77:54] budget, you feel compelled to do the [77:56] same and then they increase their budget [77:59] even more. You know, when I say that the [78:02] early 21st century was the most peaceful [78:05] era in human history, [78:08] it's [78:09] one of the indications is how uh how low [78:14] the military budgets all over the world [78:17] were. For most of history, kings and [78:21] emperors and cons and sultans, they the [78:24] military was the number one item on [78:27] their budget. They spent more on their [78:31] soldiers and navies and fortresses than [78:34] on anything else. In the early 21st [78:37] century, most countries [78:40] spend something like a few percentage [78:43] points of their of their budget on on [78:47] the military. Education, health care, [78:51] welfare were a much more a much bigger [78:55] item on the budget than defense. [78:59] And this is now changing. The money is [79:02] increasingly going to tanks and missiles [79:07] and cyber weapons instead of to nurses [79:11] and and schools and and social workers. [79:15] And again, it's not inevitable. It's the [79:17] result of human decisions. The [79:20] relatively peaceful era of the early [79:22] 21st century, it did not result from [79:25] some miracle. It resulted from humans [79:28] making wise decisions in previous [79:30] decades. [79:30] What are the wise decisions we need to [79:32] make now in your view? [79:33] Reinvest in in rebuilding a global order [79:39] which is based on [79:42] universal values and norms [79:44] and not just on the narrow interests of [79:48] of specific nation states. [79:50] Are you concerned that Trump might be [79:51] elected again shortly? I I think it's [79:53] very likely [79:54] and if it happens it is likely to be the [79:56] kind of like the the death blow to what [80:00] remains of the global order and he says [80:03] it and he says it openly. Now again it [80:06] should be clear that many of these [80:08] politicians [80:10] they present [80:12] a false dichotomy a false binary vision [80:16] of the world as if you have to choose [80:19] between patriotism and globalism between [80:24] being loyal to your own nation and being [80:26] loyal to some kind of I don't know [80:30] global government or whatever and this [80:32] is completely false there is no [80:35] contradiction between patriotism and [80:38] global cooperation. When we talk about [80:41] global cooperation, we definitely don't [80:43] have in mind, at least not anybody that [80:45] I know, a global government. This is an [80:48] impossible and very dangerous idea. It [80:52] simply means that um you have certain [80:55] rules and norms for how different nation [80:58] states [81:00] treat each other and and and and and and [81:03] behave towards each other. If you don't [81:06] have a system of of global norms and [81:09] values, then very quickly what you have [81:12] is just global conflict, is just wars. I [81:16] mean some people have this idea they [81:18] imagine the world as a network of [81:21] friendly fortresses [81:23] like each nation will be a fortress with [81:27] very high walls taking care of its own [81:30] interest interests but uh living on [81:35] relatively friendly terms with the [81:37] neighboring fortresses trading with them [81:40] and and and whatever. Now the main [81:42] problem with this vision is that [81:44] fortresses are almost never friendly. [81:47] Each fortress always wants a bit more [81:50] land, a bit more prosperity, a bit more [81:54] security for itself at the expense of [81:57] the neighbors. [81:59] And uh this is the high road to conflict [82:02] and to and to and to war and to war. [82:05] There's that phrase, isn't there? [82:06] Ignorance is bliss. Now, something that [82:10] your work has forced you and continues [82:12] to encourage you to not live in is [82:15] ignorance. [82:16] So, with that, one might logically [82:17] deduce that [82:19] out the window goes your bliss. [82:22] Um, are you are you happy? [82:26] I think I'm relatively happy, at least [82:29] happier than I was uh for most of my [82:32] life. [82:33] I [82:35] part of it is is that I invest a lot of [82:38] my time not just in [82:42] you know researching what is happening [82:43] in the world but also in the health of [82:47] my own mind [82:50] and [82:52] you know keeping a kind of balanced [82:56] information diet [82:59] that it's it's it's basically like with [83:01] food. You need food in order to survive [83:04] and to be healthy. But if you eat too [83:07] much or if you eat too much of the wrong [83:09] stuff, it's it's bad for you. And it's [83:12] exactly the same with information. [83:14] Information is the the food of the mind. [83:17] And if you eat too much of it of the [83:20] wrong kind, you'll get a very sick mind. [83:24] So I uh I try to to keep a very balanced [83:29] uh information diet which also includes [83:33] information fasts. [83:35] So I try to disconnect. [83:38] I um every day I dedicate two hours a [83:42] day for meditation. [83:43] Wow. And every year I go for a long [83:46] meditation retreat of between 30 and 60 [83:50] days like in completely disconnecting. [83:53] No phones, no emails, not even books. Um [83:58] just observing myself, observing what is [84:02] happening inside my body and inside my [84:04] mind, getting to know myself better and [84:09] kind of digesting [84:11] all the information that I absorbed [84:16] during the rest of the year or the rest [84:17] of the day. [84:18] Have you seen a clear benefit in doing [84:20] that? [84:20] Uh yes, very very clear. I don't think I [84:23] would be able to write these books or to [84:26] do what I'm doing um without these kind [84:30] with this kind of information diet and [84:32] and without kind of devoting a lot of [84:34] time and attention to the balancing my [84:38] mind and keeping it healthy. You know so [84:40] many people spend so much time keeping [84:43] their body healthy which is very [84:45] important of course but we need to spend [84:47] equal amount of attention with with our [84:49] mind. It is as important as as our body. [84:52] When you said you don't think you'd be [84:53] able to do what you do if you didn't [84:55] take these information diets, why? [84:58] I'll just, you know, um first of all be [85:02] just overwhelmed [85:04] and uh um not have any kind of peace of [85:09] mind, not have any kind of perspective. [85:11] If you're constantly in the news cycle, [85:14] in the information cycle, you lose all [85:18] perspective. You know organic entities [85:22] unlike AIs, unlike computers, we are [85:26] cyclical entities. We need to sleep [85:30] every day. AIS don't sleep. You know, [85:33] even the stock exchange closes every [85:36] afternoon. It closes also for the [85:39] weekend or for so for Christmas. If you [85:42] think about it, this is amazing that you [85:44] know if if a war erupts in Christmas [85:49] uh uh the Wall Street will be able to [85:51] react only after a couple of days [85:55] because the people are on holiday. They [85:57] they took time off. Even the money [86:01] market takes time off. But if you give [86:05] AI full control, there will never be any [86:07] time off. it will be 24 hours a day, 365 [86:12] days a a year and people just collapse. [86:17] I mean, I think part of the problem that [86:19] politicians today face is that um they [86:23] need to be on 24 hours a day because the [86:26] news cycle is on 24 hours a day. Like in [86:29] previous eras, if you're I don't know a [86:31] king in the middle ages and you you you [86:33] you ride some you go somewhere, you're [86:36] on the road in your carriage and nobody [86:39] can reach you. Even if the French are [86:42] invading, nobody can reach you. You have [86:45] some time off. If you're a prime [86:47] minister now, there is no time off. And [86:52] computers are built for it, but human [86:54] brains aren't. If you try to keep an [86:58] organic entity [87:01] awake and kind of constantly processing [87:05] information and reacting 24 hours a day, [87:08] it will very soon collapse. [87:10] It's funny, it made me think of what the [87:13] for I think it was the former Netflix [87:14] CEO or one of the Netflix CEOs or [87:16] someone said um they said, "Our biggest [87:19] competitor is sleep." [87:20] Sleep. Yeah. That's a very scary and and [87:23] very I think important line [87:25] and it's a very honest line. [87:26] It's a very honest line and it's scary [87:28] because um if people don't sleep they [87:31] collapse and eventually they die. And [87:34] this is again part of the problem that [87:37] we talked earlier about about the battle [87:39] for human attention in social media in [87:43] streaming services. Now [87:46] for many of these corporations they [87:49] measure their success by user [87:52] engagement. [87:53] The more people are engaged the more [87:56] successful we are. Now user engagement [88:00] is a very [88:02] broad definition. According to this [88:05] measurement one hour of outrage is [88:10] better than 10 minutes of joy. [88:13] And uh uh certainly better than 1 hour [88:17] of sleep [88:18] because one hour of outrage I will [88:20] consume three adverts. [88:21] Yes. [88:22] And then that means that the corporation [88:24] make $30 [88:25] for example. And and from two hours of [88:28] sleep they make nothing. From 10 minutes [88:31] of joy maybe they sell only one ad. [88:34] Mhm. And but from the viewpoint of of [88:38] how humans function and how this [88:41] organism function, 10 minutes of joy are [88:44] probably better than for us than one [88:47] hour of outrage. And certainly we need [88:49] not just two hours, we need six, seven, [88:52] eight hours of sleep. [88:53] Well, this is why, you know, the [88:55] algorithms on on certain platforms, [88:58] specifically Tik Tok, [88:59] Mhm. [88:59] are just absolutely [89:02] addictive to say the least. Like I I [89:06] because they hacked us. [89:07] Yeah. They It's literally they you know [89:10] t we had you know a certain level of [89:12] addiction to the previous social [89:13] algorithms and then Tik Tok came along [89:15] and said hold my beer and they just went [89:18] for it you know and and they've won [89:21] because of that. I see 60 year olds [89:25] absolutely addicted to Tik Tok and [89:27] because they don't understand the [89:28] concept of an algorithm sometimes um and [89:31] they don't understand like the the the [89:33] advertising model and all of that stuff [89:35] it's it's hypnotism. They're like [89:37] absolutely hypnotized. My funnily enough [89:39] my driver is one of them. So my driver's [89:40] outside whenever I walk up to his car [89:43] he's just like this on Tik Tok. [89:45] He's scrolling and I had a conversation [89:46] with him last night. I'm like do you [89:48] realize that Tik Tok has your brain? [89:51] Yeah, [89:51] you know, abs, you know, and we're just [89:54] at the very foot sort of the first steps [89:57] of an exponential curve of algorithms [89:59] competing for our attention in our [90:00] brain. [90:00] We haven't seen anything yet. I mean, [90:02] these algorithms, they are what like 10 [90:05] years old [90:06] in terms of you think about these social [90:08] media algorithms and the algorithms that [90:10] get to know you personally to hack your [90:11] brain and then grab your attention. It's [90:14] they are 10 years old [90:16] and the companies die if they don't beat [90:18] the other algorithms. So, like Twitter [90:20] now, when Elon took it over, and I think [90:22] people will relate to this if you use [90:23] Twitter, suddenly I've seen more people [90:27] having their heads blown off [90:28] and being hit by cars on Twitter than [90:30] I'd ever seen in the previous 10 years. [90:32] It's and I think someone at Twitter's [90:34] gone, listen, this company's going to [90:35] die unless we we increase time spent on [90:38] this platform and show more ads. So, [90:40] let's start serving up a more addictive [90:41] algorithm. And that requires a response [90:43] from Instagram and the other plat. And [90:45] so it's a real, [90:46] you know, Elon has this other company, [90:48] the Boring Company. [90:50] Yeah. [90:50] Which is about boring tunnels, of [90:52] course. But actually, it might be a good [90:54] idea to make Twitter more boring and to [90:58] make Tik Tok more boring. I mean, I know [90:59] it's it's a very bad [91:01] kind of business decision. [91:04] But I don't think humanity will survive [91:09] unless we have more boredom. [91:13] If you ask me what is wrong with the [91:15] world in 2023 [91:17] is that uh everybody is far too excited. [91:21] And if I had to kind of summarize what's [91:24] wrong in one word, the word is excited. [91:27] And people don't understand the meaning [91:29] of this word. People think that excited [91:32] means happy. Like two people meet, I am [91:35] so excited to meet you. I have a new [91:37] idea. I publish a new book. Whatever. [91:39] Oh, this is such a such an exciting [91:41] idea. such an exciting book. And [91:44] exciting isn't happy. Exciting isn't [91:48] always good. Sometimes, yes, sometimes [91:49] it's good to be excited. An organism [91:52] that is excited all the time dies. The [91:56] meaning of excitement is that you know [91:59] that the body is in flight or fight [92:01] mode. [92:03] All the nerves are on, all the neurons [92:06] are firing, all the muscles are tense. [92:08] This is excitement [92:10] and very often negative things excite [92:13] us. Fear excite fear is excitement. Hate [92:17] is excitement. Anger is excitement. [92:21] And um you know it's when I meet a good [92:26] friend I'm often relaxed to meet the [92:29] friend not excited. [92:32] and [92:33] or much kind you know you think about [92:35] the political level we have far too many [92:38] exciting politicians doing very exciting [92:41] things and we need more boring [92:44] politicians [92:44] more Bidens [92:46] that do less less exciting uh uh uh [92:50] things and [92:53] but the brain is wired to pay attention [92:54] to excitement and to crave it [92:56] but the brain evolved in situations when [92:59] you didn't have a constant stream of [93:03] exciting videos. Sometimes it was on, [93:06] sometimes it was off. And now our brains [93:10] have been hacked and these [93:14] devices, technologies, they know how to [93:17] create constant excitement. [93:21] And the more this happens, we also lose [93:25] our ability, our skill to be bored. that [93:29] if we have to spend a few minutes doing [93:32] nothing somewhere waiting, we can't do [93:34] it. We immediately take out the [93:37] smartphone and start watching Tik Tok or [93:40] scrolling through Twitter or whatever. [93:42] Did you hear about that experiment where [93:43] people would rather take an electric [93:44] shock than do nothing? [93:47] Yeah. And you know you [93:51] you can't get for instance [93:54] to any level of peace of mind if you [93:58] don't know how to handle boredom. [94:01] That peace and boredom are are the same [94:04] way that excitement and outrage are [94:07] neighbors. Peace and boredom are also [94:10] neighbors. And if you don't know how to [94:13] handle boredom, if the minute there is a [94:16] hint of boredom, you run away to some [94:19] exciting thing, you will never [94:21] experience peace of mind. And people if [94:25] if if humans don't experience peace of [94:27] mind, there is no way that the world as [94:30] a whole is going to be peaceful. [94:32] In 2023, I launched my very own private [94:36] equity fund called Flight Fund. And [94:38] since then, we've invested in some of [94:40] the most promising companies in the [94:41] world. My objective is to make this the [94:44] best performing fund in Europe with a [94:46] focus on high growth companies that I [94:48] believe will be the next European [94:49] unicorns. The current investors in the [94:52] fund who have joined me on this journey [94:54] are some of Europe's most successful and [94:56] innovative entrepreneurs. And I'm [94:58] excited to announce that today, as a [95:00] founder of a company, you can pitch your [95:02] company to us. Or if you are an [95:06] investor, you can also now apply to [95:08] invest with us. Head to flightfund.com [95:13] to gain an understanding of the fund's [95:14] mission, the remarkable companies we [95:16] proudly support, and to get in touch [95:18] with me and my team. Legal disclaimer, [95:20] Flight Fund is regulated by the FCA. So, [95:22] please remember that investing in the [95:24] fund is for sophisticated investors [95:26] only. Don't invest unless you're [95:27] prepared to lose all of the money you [95:28] invest. This is a high-risisk investment [95:30] and you are unlikely to be protected if [95:32] something goes wrong. There is no [95:34] guarantee that the investment objectives [95:35] will be achieved. And as with all [95:37] private and equity investments, all of [95:38] the investment capital is at risk. This [95:40] communication is for information [95:42] purposes only and should not be taken as [95:44] investment advice or a financial [95:45] promotion. As you guys know, I'm a big [95:48] fan of Hule. I'm an investor in a [95:49] company and they sponsor this podcast. [95:51] And what I've done for you, I've put [95:52] together what I call the Hule Steven [95:54] bundle, which is a selection of my [95:56] favorite products from Hule, including [95:58] the black edition salted caramel flavor, [96:00] which is super high in protein and has [96:02] 17 servings per container. Also comes [96:05] with their ready to drink product, which [96:07] is one of my all-time favorite products [96:08] from Hule. The brand new and very [96:10] exciting Hule complete nutrition bars. [96:13] This is chocolate caramel. You can see [96:15] from the empty box in front of me that [96:16] I've eaten most of them, right? Me and [96:18] my team here. If you leave these on the [96:19] counter for 5 seconds, they'll go. I'm [96:21] going to say something I've never said. [96:22] When Hule first made their bar many, [96:25] many years ago, I tried it and I didn't [96:26] like it, so I've never talked about it [96:28] on this podcast. They've spent roughly [96:30] the last 2 to 3 years making a brand new [96:32] bar, which I absolutely love. If you [96:34] want to order them yourself and get [96:35] started on your heel journey, the link [96:37] is in the description below. In this [96:39] podcast episode, wherever you're [96:41] listening to it, there'll be a Steven's [96:42] bundle link and check it out. Back to [96:44] the episode. If I could give you the [96:46] choice to be born in 1976 [96:48] as you were [96:49] Yeah. [96:50] or to be born now, [96:53] I would go for 1976. [96:55] I mean, the people of my generation, we [97:00] were privileged to grow up in one of the [97:04] most peaceful and most optimistic eras [97:07] in human history. The end of the Cold [97:09] War, the fall of the Iron Curtain. I [97:13] don't know of any better time. [97:16] Uh but when I look at what is happening [97:19] right now, I don't envy the people who [97:22] grow up in the 2020s. [97:25] What is the closing [97:29] statement of hope and solution [97:34] that kind of ties off this conversation? [97:37] What is the thing that having someone [97:38] gotten to this point in the conversation [97:40] they should be thinking about doing [97:43] which will cause the domino effect that [97:44] will lead us to maybe more hopeful [97:47] future. [97:49] But we still have agency. I mean the [97:51] algorithms are not yet in in full [97:54] control. They are taking power away from [97:57] us. But most power is still in human [97:59] hands and every human being has some [98:02] level of of power of agency which means [98:06] that each one of us has some [98:08] responsibility. [98:09] Now nobody can solve all the world's [98:12] problems. [98:14] So focus on one thing. Find the one [98:17] thing which is close to your heart which [98:20] you have a deep understanding of and uh [98:24] and and and and try to make a difference [98:26] there and the best way to make a [98:30] difference is to cooperate with other [98:32] people. I mean the human superpower is [98:35] our ability to cooperate in large [98:37] numbers. So if you care about a specific [98:40] issue don't try to be an isolated [98:44] activist. [98:45] 50 individuals who cooperate as part of [98:48] an organization [98:50] can uh do much much more than 500 [98:54] isolated activists, individuals. [98:58] So, and find your one thing and again [99:01] don't try to do everything. Let other [99:03] people do the rest and cooperate with [99:06] other people on on your chosen mission. [99:11] Yal, your book Sapiens changed the world [99:14] in many ways. is it gave us a new [99:15] perspective and a new understanding of [99:16] who we are as as humans, where we've [99:19] come from. And with that, we have a road [99:21] map for where we're going. It's [99:22] celebrating its 10th anniversary. I have [99:25] the 10th anniversary edition here, which [99:26] I'm going to beg you to sign for me [99:28] after. Um, and it really is a once in a [99:32] generation book. The numbers that I have [99:34] are that it sold more than 25 million [99:37] copies and that's in a market where [99:39] people said no one's buying books [99:40] anymore. That's crazy. That's absolutely [99:44] that's absolutely crazy. You you're [99:46] working on a new book which I'm very [99:47] excited to hear about. I'm sure that a [99:49] little birdie told me that'll be [99:50] announced next year and I'm sure [99:51] everyone's incredibly energized about [99:53] that. [99:54] Um what is the I ask this people the [99:56] question sometimes just as a way to to [99:58] close off the show but I wanted to ask [99:59] you it because it's especially pertinent [100:01] to someone that's got such a huge [100:03] varying wealth of work. Is there one [100:05] particular topic that is pertinent to [100:09] our future [100:11] that we didn't talk about? [100:13] I I would say that when we talk about [100:15] the future, [100:18] um, [100:20] history is is more relevant than ever [100:23] before. [100:24] History is not really the study of the [100:28] past. History is the study of change, of [100:32] how things change. [100:34] you nobody cares about the past for the [100:37] sake of the past. All the people who [100:39] lived in the middle ages or in the uh uh [100:43] ancient uh uh Rome, they all they are [100:45] all dead. They we can't do anything [100:49] about their disasters and their misery. [100:53] We can't correct any of the wrongs that [100:56] happened in ancient times. Um and they [101:01] don't care what we say about them. [101:04] You can say anything you want about the [101:06] Romans, the Vikings, they they are gone. [101:08] They don't care. [101:09] The reason to study the past is because [101:13] if you understand the dynamics of change [101:17] in previous centuries, in previous eras, [101:20] this gives you perspective [101:23] uh on the process of of change in in the [101:26] present moment. And I think the curse of [101:29] history [101:31] is that people have this fantasy of [101:34] changing the past of bringing justice to [101:37] the past and this is just impossible. [101:40] You cannot go back there and and save [101:42] the people there. The big question is [101:46] how do you um save the people now? How [101:50] do you prevent [101:53] catastrophes [101:54] perhaps from from happening? [101:58] And this is the reason to to study [102:01] history. [102:02] And the main message of of history is [102:05] that humans created the world in which [102:09] we live. [102:11] The world that we know with nation [102:15] states and corporations and capitalist [102:17] economics and uh uh religions like [102:21] Christianity and Hinduism, humans [102:23] created this world and humans can also [102:26] change it. If there is something about [102:30] the world that you think is unfair, is [102:35] dangerous, is is problematic, then I [102:38] some things are beyond our control. The [102:40] laws of physics are beyond our control. [102:43] So far, the laws of biology are also [102:46] beyond our control. But knowing [102:52] what is natural, what is the outcome of [102:56] physics and biology versus what is the [103:00] outcome of human inventions, human [103:04] stories, human institutions. This is [103:06] very difficult. A lot of things that [103:10] people think are just natural. This is [103:13] the way the world is. This is biology. [103:15] This is physics. They are not. They are [103:17] actually the result of historical [103:19] processes. [103:21] And this is why it's so important to [103:23] understand history to understand how [103:26] things change and to understand what can [103:29] be changed. [103:32] We have a closing tradition on this [103:33] podcast where the last guest leaves a [103:34] question for the next guest, not knowing [103:36] who they're going to be leaving it for. [103:37] Oh, [103:38] the question that's been left for you, [103:39] if you could impose a global law, but [103:43] only one global law, what would it be [103:45] and why? [103:47] Oh, great question for you. I [103:49] I would say that people [103:53] should [103:55] consume less information [103:58] and spend more time reflecting and [104:01] digesting what they already know, what [104:03] they already heard. [104:05] Thank you, Eva. It means um a huge [104:07] amount to me that someone of your esteem [104:09] and someone that whose books have [104:11] inspired me and turned the lights on in [104:13] so many areas of my life um would have [104:15] this conversation with me today. So I [104:16] thank you so much for that. But also for [104:18] turning the lights on to the hundreds of [104:20] millions of people that have consumed [104:22] your work all around the world, the [104:23] videos, the books, etc., etc. as you've [104:25] said there, it's the most important work [104:27] because it helps us looking back at [104:29] history in a way that is accessible um [104:31] and inclusive in a way that even I could [104:34] read without having to be a historian or [104:36] understand very complex subject matter. [104:38] So, thank you so so so much. [104:41] Thank you. It's been great to to be [104:42] here. [104:46] If you listen to this podcast [104:47] frequently, there's something I talk [104:48] about very often and that is the subject [104:51] of sleep. And so I dug down a pretty [104:54] deep sleep rabbit hole to figure out how [104:56] I could sleep better. One of the things [104:57] that I found is a brand called Eight [104:59] Sleep that sponsored this podcast. And [105:01] that is the cover that I have on my bed. [105:03] I saw the variance in my performance, my [105:05] ability to talk, my mood, and everything [105:08] that matters to me when I'm unslept. It [105:10] regulates the temperature of both sides [105:13] of my bed individually. So my partner [105:15] can have cold, I can have a little bit [105:17] warmer, and it learns about my body and [105:19] sets my bed to the temperature that I [105:22] need to have optimal sleep. The brands [105:25] that I talk about on this this show, the [105:27] podcast sponsors that I have are brands [105:29] that I love and use, and EightLe is one [105:31] of them. They've made that piece of foam [105:33] that we all sleep on for 8 hours a day [105:36] smart. I've put a link in the [105:38] description below, but you can go to [105:40] eightsleep.com/stephven [105:42] for exclusive holiday savings. [105:46] Do you need a podcast to listen to next? [105:48] We've discovered that people who liked [105:50] this episode also tend to absolutely [105:53] love another recent episode we've done. [105:55] So, I've linked that episode in the [105:57] description below. I know you'll enjoy [105:59] it. [106:02] [Music]