Daniel Dennett Rejects Neuroscience Claims Debunking Free Will

丹尼尔·丹尼特驳斥神经科学否定自由意志的说法 人类智能与道德责任通过文化传播的认知工具得以升级，使得物理决定论与伦理能动性无关。 丹尼特探讨了人类智能是如何通过文化演进的“思维工具”而非快速的生物学演进实现自我提升的，并以弗林效应和陆生生物量的变迁作为认知增强的证据。他审视了常见的诊断工具与批判性习惯——例如识别反问句、“想必”（surely）一词以及非此即彼的虚假二分法（“rathering”）——同时通过寄存器机揭开了计算的神秘面纱。丹尼特还运用诸如两种彩票和虚假神经外科医生等直觉泵来说明道德能力，从而为相容论自由意志辩护，反驳神经科学怀疑论。在观众提问环节的总结中，他探讨了意识作为已安装文化软件的涌现本质、制度性惩罚的结果主义必要性，以及围绕软件可专利性的争论。 自20世纪30年代以来弗林效应带来的20到30点长期增长表明，人类智能代际间的增强并非源于基因演化，而是通过认知假体和文化思维工具实现的。 (3:40) 陆生脊椎动物的生物量平衡已彻底颠倒：一万年前人类及其驯养的牲畜仅占不足1%的微小比例，如今却占陆生脊椎动物总生物量的约98%。 (11:57) 丹尼特驳斥了通俗神经科学家关于物理决定论否定自由意志的观点，认为道德责任依赖于后天文化习得的认知能力，而非量子不确定性。 (30:40) 结果主义惩罚对于民事契约和承诺至关重要；完全用精神病学治疗取代法律惩罚会威胁个人权利，呼应了将异见病理化的极权主义做法。 (55:22)

Philosopher Daniel Dennett argues that human intelligence, consciousness, and moral agency are not mysterious biological gifts but functional outcomes of cultural thinking tools and installed cognitive software. Using thought experiments or "intuition pumps," Dennett contends that complex cognition and responsibility operate through standard causal mechanisms rather than supernatural phenomena. Cognitive Scaffolding and the Evolution of Thought (3:40) Dennett maintains that unassisted human brains are limited without external tools, citing Bo Dahlbom's observation that thinking with a bare brain is as ineffective as doing carpentry with bare hands. Human biology remains essentially identical to that of ancestors 100,000 years ago, yet human power has expanded drastically. Citing Paul MacCready's terrestrial vertebrate biomass metric, Dennett notes that while humans and livestock once accounted for a fraction of 1% of biomass 10,000 years ago, they now comprise roughly 98%. Dennett attributes this shift to a bootstrapping coevolution where culturally transmitted tools enhance cognitive capacity, which in turn leads to deliberate tool design. Similarly, the Flynn effect's secular rise of 20 to 30 raw IQ points since the 1930s demonstrates how internalized conceptual tools—such as percentages—function like installed cognitive software across generations. Intuition Pumps and Necktop Critical Alarms (6:55) Rather than relying strictly on formal proofs, philosophy routinely utilizes intuition pumps—informal narrative thought experiments designed to persuade interlocutors, frame provisional assumptions, and test intuitions. Dennett warns that thinkers must test these pumps by "turning all the knobs," varying their conditions using John Stuart Mill's method of differences to ensure conclusions do not rest on cosmetic narrative flaws. To safeguard against manipulative rhetoric, Dennett outlines "necktop alarms" that serve as cognitive diagnostic checks: probing rhetorical questions directly instead of accepting implied conclusions, inspecting the word "surely" for unargued premises that often signal an argument's weakest point, and treating the word "rather" with caution to identify concealed false dichotomies. From Bottom-Up Emergence to Deliberate Design (20:36) Bridging bottom-up evolutionary processes with top-down intelligent design, Dennett contrasts the Australian termite mound with Antoni Gaudí's Sagrada Família. While clueless termites generate complex structures without a central blueprint or architect, human designers like Gaudí and Alan Turing create top-down blueprints in advance because their brains are scaffolded by culturally accumulated concepts. Dennett emphasizes that computer programming reveals that intricate operations arise strictly through ordinary causal mechanisms. Using Hao Wang's minimalist register machine model—which relies merely on increment, decrement, and branch instructions—Dennett demystifies mental processes, asserting that complex emergence requires no magical forces, psionic powers, or ectoplasm. Free Will, Moral Competence, and Legal Accountability (30:40) Dennett rejects claims from popular neuroscientists that physical determinism renders free will an illusion. Distinguishing moral competence from fundamental physics, Dennett uses a two-lottery thought experiment to show that whether outcomes are fixed in advance or decided at the moment of choice, an agent's practical situation is functionally the same. Moral responsibility belongs to the practical "manifest image" of social life and requires only the installed cognitive competence to understand rules and act reliably. From a consequentialist perspective, Dennett defends legal punishment as essential for upholding promises and social contracts, warning that medicalizing wrongdoing under the guise of therapy erodes human dignity and civic accountability. Installed Consciousness and Practical Substitutes (50:26) Dennett describes rich human consciousness as culturally installed software rooted in language and social interaction, distinguishing it from basic biological sentience like feeling pain. Addressing critics who claim physical explanations fail to explain subjective experience, Dennett cites magician Lee Siegel's insight that skeptics often dismiss real-world mechanisms as "cheesy substitutes" because they only accept impossible miracles as genuine. Dennett accepts these practical mechanisms, arguing that science can fully explain the varieties of consciousness and agency that are actually worth wanting. While Dennett champions computational and consequentialist frameworks for understanding mind and agency, he acknowledges lingering open problems. He notes an ongoing ambivalence regarding Newcomb's paradox, suspecting structural flaws in its premises, and points to the unresolved legal challenge of patenting software based on functional, cognitive information rather than traditional physical constructs.

哲学家Daniel Dennett在Google演讲中指出，人类认知能力的飞跃并非源自基因演化，而是文化演进所打造的“思考工具”与大脑相互协同进化的结果；以此为基底，他倡导通过拆解直觉泵、建立批判性预警，来消解哲学与认知科学对决定论、意识和自由意志的误导性预设。 认知工具的协同演化与思维自举 (3:40) Dennett借用Bo Dahlbom的格言指出，赤手空拳做不好木工，仅凭未经辅助的裸脑也无法处理深度思考。现代人类的生物学构造和大脑机制与10万年前的祖先基本相同，但文明与能力的爆发式增长源自“思维软件”的安装。例如弗林效应（Flynn effect）显示，自1930年代以来人类原始智商测试得分显著上升约20到30点，这在短时间内不可能源自基因演变，而是百分比等概念工具由科学界逐步下沉、内化至公众心智的结果。在生态规模上，Paul MacCready的估算表明，人类与牲畜占陆地脊椎动物生物量的比例从万年前农业文明初期的不足1%暴增至当下的98%。这种力量的膨胀建立在思考工具的自举（bootstrapping）之上：生物智能的微小提升促成了工具使用，而工具文化反过来重构了认知环境，最终使人类由无意识的演化产物转变为有意识的设计者。 直觉泵辨析与批判性认知警报 (6:55) Dennett认为哲学缺乏算法工具，因而严重依赖直觉泵（intuition pumps）这一思考装置。直觉泵旨在说服受众并设立探究的临时锚点，但容易夹杂误导性假象。他主张借用Douglas Hofstadter的“调节所有旋钮”方法，即通过逐一改变思想实验的情境变量，诊断驱动结论的核心究竟是严谨逻辑还是表层修辞。与此同时，Dennett建议读者在脑中建立若干“颈上警报”（necktop alarms）：当阅读中出现反问句时，应意识到作者往往在隐瞒归谬法的前提，读者可尝试直接作答；看到“想必/无疑”（surely）时需警惕作者在无论据地滑过薄弱环节（文本分析显示其约三分之一标注了论证漏洞）；遇到“与其说是X不如说是Y”（rathering）时，则要审视这是否掩盖了一个虚假二分法。 设计倒置：从白蚁盲目试验到图灵式自顶向下 (25:00) 通过对比阿舍利手斧（经历百万年停滞）与现代计算机鼠标（数十年内迅速更迭），以及澳大利亚白蚁城堡与高迪设计的圣家堂，Dennett展示了设计范式的演化断层。白蚁筑巢代表了自底向上的达尔文式研发——不存在建筑师或蓝图，盲目的分散试错即可造就复杂结构；而高迪和艾伦·图灵则代表了自顶向下的智能设计，图灵在实体制造前已在纸面上完整证实了计算机的蓝图。文化演进使得思维工具得以积累，最终将达尔文的盲目试错过程倒置为具备预先概念验证的真正智能设计。大脑内部也是这种机制的具象体现：它由1000亿至2000亿个类似独立白蚁的神经元组成，并非中央集权的超级电脑，而是通过底层竞争结盟涌现出统一心智，这与健康经济体依靠分散代理人运作且辅以适当监管的逻辑一致。 决定论下的道德能力与惩罚机制 (30:40) Dennett强烈反驳部分神经科学家“神经学已证明自由意志是幻觉”的断言。他以“虚构神经外科医生通过谎称植入芯片完全控制患者行为”的思想实验指责这类普及者正在以语言解除公众的道德主体性。Dennett通过两种彩票的直觉泵说明，即使宇宙在物理层面上是决定论的（即结果在大爆炸时已注定，如同彩票在售出前已开出），它在功能上与量子非确定性世界毫无差异，坏运气在两者中同样存在。在现实世界中，自由意志属于法律与契约构成的日常表象，其核心是后天培养的道德胜任力（moral competence），而非违反因果律的“物理奇迹”。基于后果主义立场，Dennett坚持惩罚制度是维系契约与承诺不可或缺的负面后果保障，彻底用医疗治疗替代法律惩罚反而会滑向任意剥夺公民身份的父权主义危险。 主观意识的去神秘化与功能性信息 (50:26) 面对意识是否是宇宙基本维度的提问，Dennett将其定性为“虚假二分”的滥用，并坚决反对将意识与感受性（qualia）神秘化。他引用魔术师Lee Siegel对印度街头魔术的观察指出，哲学家往往将实际可运作的精妙机制贬为“廉价替代品”，却去追逐一种根本不存在的“真魔法”。在Dennett看来，人类丰富且具备反思性的意识本质上是通过语言和社会化安装的文化软件；动物具备知觉与疼痛感知，但不具备这种依赖思考工具构建的意识。在知识产权与技术领域，这种机械与软件观促使他倾向于支持软件专利：法律必须从香农信息走向关注效用与适应性传递的“语义/认知信息”，将精心设计的软件工具视作具备实际功能的真实发明。 Dennett总结认为，《思考的直觉泵和其他工具》旨在为读者提供模块化、可随时检验的便携认知构件。但他坦言，面对诸如纽康姆悖论（Newcomb's paradox）等决策难题，自己仍未能在“选一个盒子”还是“选两个盒子”间达成彻底的理论定见，暗示部分思想实验的底层框架本身仍可能存在尚未厘清的结构性缺陷。

## Background 本视频由 YouTube 频道 Talks at Google 于 2013 年 5 月 22 日发布。内容记录了著名哲学家、塔夫茨大学教授丹尼尔·丹尼特（Daniel Dennett）在 Google 围绕其新书《直觉泵和其他思考工具》（Intuition Pumps and Other Tools for Thinking）所作的演讲。页面简介指出，丹尼特通过生动且接地气的思维工具，旨在引导读者和听众学会可靠且从容地思考心理学、生物学、计算机科学等领域的棘手难题。 ## TL;DR 讲座中，丹尼特与 Google 研究总监彼得·诺维格共同探讨了计算思维与哲学直觉泵的交汇。丹尼特阐述了思维工具如何像“安装在颈上电脑的大脑 App”一样驱动人类智力爆发，并通过阿舍利手斧与鼠标、白蚁丘与圣家堂的对照，解析自下而上与自上而下演化设计的鸿沟。他还借助“神经外科医生的恶作剧”与“两张彩票”等直觉泵，有力辩护了决定论与自由意志、道德责任的完全相容性，并深入剖析了意识的软件本质与惩罚制度的合理性。 ## Atmosphere & style 整场演讲洋溢着理性的幽默、知性交锋的愉悦与深入浅出的哲学魅力。丹尼特白发长髯、风度儒雅，与深谙技术语境的诺维格及 Google 工程师听众迅速产生共鸣，现场笑声与会心的点头此起彼伏。他举重若轻地将晦涩的形而上学思辨拆解为接地气的段子和思想实验，在揭示严密逻辑漏洞的同时展现出极富感染力的思维洞察力。 ## Bulletin - [03:40](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=220s) 丹尼尔·丹尼特强调人类心智的飞跃源于文化演化积累的“思维工具”，正是这些工具的普及驱动了近一个世纪全球智商测验得分显著上升的“弗林效应”。 - [11:57](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=717s) 保罗·麦克克莱迪的测算显示，人类及其家畜在陆生脊椎动物生物总量中的占比已从一万年前的不足 1% 激增至当今的 98%，人类技术与思考工具已全面重塑地球生态。 - [30:40](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1840s) 丹尼特指出因果决定论与自由意志完全相容，神经科学所谓“否定自由意志”的主张混淆了形而上学奇迹与实际的道德契约胜任力，可能削弱社会的道德责任基石。 - [50:26](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=3026s) 人类意识在本质上是后天“安装”的文化软件，必须依赖语言和叙事工具的塑造；若脱离文化输入，人类心智将退化至仅具基本生物感知的状态。 ## Key takeaways - [00:04](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=4s) 计算机科学家习以为常的算法迭代与涌现特性，是哲学家和大众常感陌生却极其强大的计算型思维工具。 - [03:40](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=220s) 思维工具如同做木工的器械，人类智力近百年来的“弗林效应”主要得益于思考工具下沉至日常思维，而非基因进化。 - [06:55](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=415s) 检验“直觉泵”优劣的标准在于逆向工程并微调所有变量旋钮；若对叙述细节的细微改动异常敏感，则该直觉泵通常存在逻辑缺陷。 - [11:57](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=717s) 现代人类与十万年前祖先在基因和生理上基本相同，文明飞跃的根本驱动力在于思考工具与大脑演化的“自举”协同演化。 - [14:57](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=897s) 反问句往往是隐蔽的归谬法论证，读者应像设定警报一样保持警惕，主动尝试正面回答该反问以破解修辞操纵。 - [17:15](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1035s) 学术论文中频现的“想必（surely）”常标记着作者缺乏证据的逻辑脆弱点，而“与其说（rathering）”则常用于强行构造虚假二分法。 - [19:41](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1181s) 理解底层寄存器机的计算模拟能够带来认识论上的去魅效果：一旦过程能被完整模拟，即可排除任何神秘主义或超自然力量的介入。 - [22:45](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1365s) 阿舍利手斧长达百万年的形制停滞与现代鼠标的极速更迭，鲜明对照出自然基因演化的缓慢与文化演化的爆炸式速度。 - [25:00](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1500s) 白蚁丘与高迪圣家堂的对比揭示了演化的倒转：大自然始于自下而上的盲目试错，而文化工具的积累最终孕育了自上而下的真正智能设计。 - [30:40](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1840s) 神经科学家轻率宣称“自由意志不存在”具有极大的社会破坏性，他们混淆了微观物理的因果决定与宏观主体的道德胜任力。 - [33:54](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2034s) “预先决定的彩票”直觉泵表明，即使未来被物理因果提前注定，行动主体所面对的概率、机会与自由选择空间并未因此消失。 - [38:20](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2300s) 大脑由数以亿计各具局部利益的神经元竞争协作构成，这一自下而上的机制启示了健康经济体系需在无序放任与过度集权之间寻求制度平衡。 - [44:51](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2691s) 自由意志的日常核心是签署合同并承担责任的“胜任能力”，而非物理层面上毫无先决原因的形而上学奇迹。 - [52:22](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=3142s) 许多人对意识与感受质的困惑源于对“真正魔术”（超自然奇迹）的执念，实际上基于科学规律的因果解释足以支撑人类有意义的精神生活。 - [55:22](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=3322s) 从结果主义出发，惩罚并非内在之善，但它是维系社会承诺和信任契约的必要威慑机制，不可用无休止的“医疗重塑”取而代之。 - [58:38](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=3518s) 软件本质上是人类精心设计出来的效用机制与适应性结构，在法律层面理应具备可专利性。 ## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=4s) 彼得·诺维格开场：计算思维工具与“泰迪熊调试法” Google 研究总监彼得·诺维格（Peter Norvig）主持了本次讲座并介绍丹尼尔·丹尼特（Daniel Dennett）。诺维格幽默地表示丹尼特的履历漫长而杰出，在此无需赘述，“大家可以直接 Google 一下”（引起全场笑声）。诺维格重点推介了丹尼特的新书《直觉泵和其他思考工具》（*Intuition Pumps and Other Tools for Thinking*）。他认为这本书极为出色，因为其中包含的许多思维工具正是计算机科学界习以为常、却往往未被大众熟知的计算型思维模型。例如，丹尼特早在 1995 年关于演化的著作（《达尔文的危险思想》）中就曾写道： > “Life on Earth has been generated over billions of years by an algorithmic process.”（地球上的生命是通过算法过程在数十亿年间生成的。） 诺维格指出，计算机科学家很自然地理解这种“通过大量微小步骤的反复迭代与组合来产生惊人结果”的逻辑。但某些未具名的哲学家却会说：如果我模拟三四步根本看不到终点，那么这一定是某种奇迹或魔法；他们无法理解几步与上千亿步之间存在着质的飞跃。同样，程序员对“涌现特性”（emergent properties）屡见不鲜——程序经常出现未曾预料的涌现行为，让工程师抓破头皮去排查——但外部世界对此却缺乏认知。 诺维格还分享了书中提到的一个思考工具：“拿外行听众当诱饵”（using lay audience as a decoy）。如果一位哲学教授讲不清楚细节，与其直接找他辩论，不如请他去给入门班讲课；为了让初学者听懂，教授就必须费尽心思把问题梳理得一清二楚。诺维格提到，Google 员工罗布·派克（Rob Pike）在书中也记录过类似方法，称为“泰迪熊法”（teddy bear approach）：某大学计算机入门课的助教办公室门外总是排着长队，学生们都在焦急地等待助教帮忙排查代码故障。于是助教在门外的一张椅子上放了一只泰迪熊，并规定任何人在向助教求助前，必须先对着泰迪熊把问题讲一遍。 ```mermaid flowchart TD accTitle: 泰迪熊调试法流程 accDescr: 展示学生在寻求助教帮助前先向泰迪熊阐述代码问题的排查流程 subgraph Office["办公室门外排查机制"] A["遇到 Bug 的学生排队求助"] --> B["坐在椅子前面对泰迪熊"] B --> C["向泰迪熊逐行解释代码逻辑"] C --> D{"在自述中突然发现漏洞？"} end D -->|"是（约 50% 概率）"| E["自行顿悟并回去修复代码"] D -->|"否"| F["进入办公室向助教求助"] ``` - 学生常常对着泰迪熊说：“熊啊，我的程序跑不通。这循环绝不可能大于 1……咦，等等，好像不是这样，算了我懂了！” - 仅靠这只泰迪熊，就成功解决了大约一半的代码故障，许多学生根本不必再找助教。 --- ### [03:40](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=220s) 丹尼尔·丹尼特开讲：思考工具与“弗林效应” 丹尼特登台致谢，打趣说自己一定要把“泰迪熊法”收入工具箱，甚至笑称这大概直接解释了心理分析师的成功率——如果换成一只泰迪熊，不仅效果一样，而且“便宜得多”。随后丹尼特切入正题，引用他的一位瑞典优秀学生博·达尔伯姆（Bo Dahlbom）的警句展开论述： > “You can't do much carpentry with your bare hands, and you can't do much thinking with your bare brain.”（你无法赤手空拳做木工，你也无法单凭光秃秃的大脑做多少思考。） 丹尼特指出，即使像“词汇”这样简单的东西，本质上也是一种思考工具。人类利用这些工具已有漫长历史，而近年来的研究显示，这些思维工具正在产生惊人的现实效果。丹尼特询问现场听众有多少人知道“弗林效应”（Flynn effect）。该现象虽以研究者吉姆·弗林（Jim Flynn）命名，但他主要是引起了公众对该现象的广泛关注：全球范围内的智商测验得分在过去近一个世纪里出现了显著上升。如果把今天的智商测试题目拿给 20 世纪 30 年代拿到 100 分（标准平均值）的人做，他们大概只能拿到 80 分左右，在今天几乎会被判定为智力迟钝；其净增长幅度高达 20 到 30 个 IQ 点。 丹尼特强调，这种巨大的提升与基因改良无关（时间跨度太短，不可能发生根本性基因进化），也并非单纯由学校补习或课业量驱动。弗林认为——丹尼特也倾向于认同这一观点——其根本原因在于：由科学家、工程师和专业思想家率先研发、测试并调试（debug）的思考工具，正逐步下沉并渗透到普通人的日常思维之中。比如“百分比”（percentage）概念，如今儿童能很快掌握，但在五十年或一百年前，对大多数人来说还是个相当复杂的抽象概念。人类在不知不觉中将这些思维工具“安装”进了自己与他人的大脑中，这构成了人类思维能力跃升的核心动力。 --- ### [06:55](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=415s) 什么是“直觉泵”及其检验方式 作为一名哲学家，丹尼特观察到许多同行对自身所使用的思维工具出奇地缺乏自觉。哲学家向来以高度自省、沉思和“内省自视”（navel-gazing）著称，却很少有人停下来反问：我究竟是怎么推导的？为什么我认为这是正确的论证路径？我正在运用什么工具？丹尼特希望促使大家对日常使用的工具有更清醒的认知。 所谓思考工具，包括词汇、数字、图表、地图、各类系统化方法，以及丹尼特在三十多年前创造的术语——“直觉泵”（intuition pump）。丹尼特最初提出这个词是为了反驳约翰·塞尔（John Searle）著名的“汉字屋”（Chinese Room）思想实验。由于当时丹尼特是在猛烈抨击该实验，导致很多人误以为他全盘否定直觉泵。丹尼特澄清道，直觉泵本身是极好的工具，关键在于区分“完好的泵”与“有缺陷的泵”。 检验直觉泵优劣的标准方法是逆向工程。丹尼特借用道格拉斯·霍夫施塔特（Douglas Hofstadter）的精妙比喻：必须去“转动直觉泵上的所有旋钮”，观察在各种变量调整下其推导机制如何运作。这在本质上就是约翰·斯图尔特·密尔（John Stuart Mill）的“差异法”（Method of Differences），即通过实证性的微调测试，检验直觉泵在模型形变下的鲁棒性（robustness）： > “And sometimes they're remarkably sensitive to tiny variations in the way you tell a story. Then you know it's probably a defective pump.”（有时它们对讲故事方式中极微小的变动异常敏感，这时你便知道这很可能是一个有缺陷的泵。） - **说服工具（Persuaders）**：直觉泵不仅是思考工具，更是辩论说服的利器。丹尼特认为许多优秀的思考都诞生于良性竞争的环境中——“我试图说服对方命题 P 为真，而对方试图说服我非 P”，这种对手博弈过程（opponent process）是探寻真理的高效途径。 - **锚定与假定固定点**：在哲学探讨中没有绝对的既定前提。丹尼特说：“Philosophy is what you're doing when you're not quite sure what the right questions to ask are.”（哲学就是当你不确定该问什么正确问题时所做的事情。）因此研究者只能临时钉下几个支点作为假想锚点，观察由此能推导出什么；若逻辑不通，甚至会同时推翻这些支点。在此阶段尚无现成算法可用，直觉泵便成为在迷茫情境中探索的脚手架。 --- ### [10:28](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=628s) 思想实验：保罗·麦克克莱迪的大爆炸 为了展示直觉泵如何在宏观视野下重构认知，丹尼特举了一个未收录在本书中、但他近期反复引用的案例：“麦克克莱迪大爆炸”（The MacCready explosion）。保罗·麦克克莱迪（Paul MacCready）是人造动力飞行器“信天翁号”（Gossamer Albatross）的发明者，也是一位极具远见的环保先锋工程师，于数年前逝世。麦克克莱迪在一次 TED 演讲中提出了一个震撼的生物量测算构想。 麦克克莱迪提议将时间拨回距今 10,000 年前的农业文明黎明期。当时全球各地的解剖学现代智人（Homo sapiens）刚刚开始驯化动植物。如果制作一个巨型天平： 1. **天平一端**：放上所有现代智人，加上所有家禽、家畜及宠物（即人类及其驯化的全部脊椎动物）。 2. **天平另一端**：放上地球上所有其余的陆生脊椎动物（不含昆虫、蠕虫和鱼类）。 麦克克莱迪计算了这万年节点上双方生物量的对比，这一度量直观揭示了人类技术与扩张对地球生态载荷带来的根本性剧变。 ### [11:57](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=717s) 人类生物量的惊人扩张与思考工具的协同演化 丹尼尔·丹尼特（Daniel Dennett）首先向台下的观众抛出了一个惊人的生物学统计数据。在人类历史早期，人类及其饲养的牲畜仅占陆生脊椎动物生物总量的百分之一的极小一部分。丹尼特询问听众如今这个比例是多少，观众们纷纷猜测“10%”、“超过50%”，甚至有人喊出了“99%”。丹尼特打趣道：“99%，你开玩笑吗？不，是98%。”人类与其家畜如今已经彻底吞没并主导了整个地球的脊椎动物生物量，成为地球演化史上最惊人的生物学现象之一。 随后，丹尼特引用了保罗（Paul）的一段动人论述：“在独一无二的球体上历经数十亿年，机遇绘就了一层薄薄的生命外衣——复杂、不可思议、美妙且脆弱。突然之间，我们人类在人口、技术和智能上飞速增长，攀升到了拥有可怕力量的地位：现在，正是我们在挥舞这支画笔。”丹尼特特别强调了“技术”和“智能”，指出它们本质上就是思考工具。从生物学角度来看，现代人类与十万年前的祖先拥有几乎相同的基因、骨骼肌肉、大脑结构以及司法与道德萌芽，唯一的根本区别就在于我们所掌握的思考工具。这引发了一个类似“先有鸡还是先有蛋”的问题：究竟是演化出的工具让我们变聪明，还是我们演化得足够聪明之后才创造了工具？丹尼特的回答是肯定的——两者兼而有之。这是一场“自举”（bootstrapping）的协同演化过程：微小的演化让我们变聪明，从而能够更具智慧地打造工具。早期的祖先并不是自觉的“智能设计者”，他们没有发明词汇的概念，只是自然而然地开始使用词汇并因而变得更加敏锐；而今天的人类则已经能够完全有意识地、深思熟虑地对思考工具进行智能设计。 - 人类与牲畜在陆生脊椎动物生物总量中的占比：过去不足 1%，如今高达 98%。 - 人类与十万年前祖先的核心区别：不是神经元、基因或解剖生理结构的变化，而是文化与思考工具的叠加。 - 智能与工具的演化机制：文化演化与大脑演化的协同推进，从无意识的使用者演变至有自觉意识的设计者。 ### [14:57](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=897s) 归谬法与“反问句警报” 在梳理各种思考工具时，丹尼特首先提到了大众耳熟能详的“归谬法”（reductio ad absurdum），并将其形象地比喻为逻辑论证的“撬棍”（crowbar）。归谬法的操作机制是：为了论证的需要，你先完全接纳对手的前提或论点，随后借助逻辑撬棍顺理成章地推演，直到推导出一个荒谬的自我矛盾；一旦出现矛盾，对手的论点便不攻自破。丹尼特指出，我们在日常生活中经常不自觉地运用归谬法，例如脱口而出“如果那是只熊，那熊头上就长着鹿角了”。 丹尼特之所以重提这个经典工具，是为了揭示日常语言中“反问句”（rhetorical question）背后隐藏的逻辑陷阱。许多反问句本质上就是经过伪装的归谬法论证，作者或演讲者预设了荒谬的后果是显而易见的，因此懒得给出实际证明，只是用反问轻轻推你一把，让你心照不宣地顺着他的思路走。为此，丹尼特建议大家在脑中设定一个像“Google 快讯”一样的思维警报系统：一旦在论辩中遇到反问句，脑子里的铃声就要立刻响起（“叮！”）。这时你不妨停下来，尝试去认真回答那个理应不言自明的反问。他幽默地回忆起《花生漫画》（Peanuts）中的一幕：查理·布朗在倒数第二格反问道：“有谁能说清对与错的区别呢？”而在最后一格中，露西毫不客气地接茬说：“我能。” ```mermaid flowchart TD accTitle: 归谬法与反问句的逻辑运作机制 accDescr: 展示归谬法的推导流程以及反问句如何通过隐去矛盾推导来迫使受众接受结论。 subgraph Reductio ["经典归谬法 (Logical Crowbar)"] A["假定对方论点成立"] --> B["合乎逻辑地推演"] B --> C["导出显而易见的荒谬或矛盾"] C --> D["证明初始论点不成立"] end subgraph Rhetorical ["被伪装的反问句"] E["抛出无需回答的反问"] -.->|"隐藏中间推导步骤"| C E --> F["直接诱导受众跳过论证并默认结论"] end ``` ### [17:15](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1035s) 思维警报：“想必”与“与其说……不如说……” 顺着反问句警报的思路，丹尼特指出了写作与论辩中另外两个极其危险但极具隐蔽性的修辞陷阱。第一个是“想必”（surely）警报。每当论述中出现“想必/无疑/当然”这类词汇时，受众脑海中的警报就应该鸣响。例如：“诚然，你可以制造一台能够按颜色分拣水果的计算机，但想必（叮！）没有任何计算机能够拥有自己最喜欢的颜色，或者真正懂得欣赏颜色。”作者之所以在这个位置使用“想必”，是因为他深知该命题既未显然到“不言自明”的程度，自己又拿不出实质性的论据，因此只能借助修辞暗度陈仓，硬生生地将其塞入讨论中。 丹尼特在哲学领域亲自验证了这一直觉。他利用 Google 字符串搜索检索了数十篇在线的心灵哲学论文，统计其中出现“surely”的频次。尽管大多数严谨的论文根本不用该词，但在出现的数十个案例中，大约有三分之一正是整篇论文逻辑最脆弱的破绽所在，那是读者最应该集中审视的命题。 另一个需要高度警惕的修辞是“与其说……不如说……”（rathering / rather）。丹尼特指出，这种句式往往在暗中树立一道“虚假二分法”（false dichotomy）：“情况并不是 A，反而/不如说是 B。”读者遇到“rather”时，应立即启动警铃反问：为什么不能既是 A 又是 B 呢？这种非此即彼的对立通常并未经过严格论证，仅仅是通过一个连词被强行预设并草草带过。 - “想必”警报（Surely alarm）：标记作者缺乏论证却希望受众默认同意的脆弱逻辑节点。 - 文本搜索实测：哲学论文中约 1/3 的“surely”精准定位了全文论证链条中最致命的薄弱处。 - “与其说”警报（Rathering alarm）：识别预设对立的虚假二元对立，提示读者探寻兼容两者的第三种可能。 ### [19:41](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1181s) 在颈上电脑中安装 App 与去魅的计算力量 在总结了上述几种识别逻辑漏洞的习惯后，丹尼特形象地比喻道，自己刚刚在听众的“颈上电脑”（necktop，即大脑）里安装了一组应用程序，并直言：“这绝非戏言。你们双耳之间的那个东西就是一种计算机。”丹尼特向身为技术从业者的 Google 员工指出，计算机的力量不在于单个裸露的 CPU，而在于软件的层层抽象与应用程序的不断堆叠；人类的大脑正是通过一代代思维工具这类“软件”的积累，才从单纯的生物计算单元蜕变为具备高度抽象思考能力的智能体。 丹尼特在书中专门插入了一段关于计算机本质的探讨。为了让普通读者彻底理解计算力量的源泉，丹尼特甚至做了一件在通俗读物中堪称“反常”的事情：在名为《揭秘计算机力量的七大秘密》一章中，他详细指导读者如何为王浩（Hao Wang）提出的极简寄存器机（Register Machine）编写程序。读者利用这台只有自增、自减和条件分支跳转基本指令的假想机器，一步步构建出算术运算及更复杂的功能。丹尼特甚至在书中布置了课后习题，让读者去切身体会循环（loop）、计数器（counter）的实际运作。 丹尼特强调，理解这种底层计算逻辑具有无可替代的认识论价值：一旦你能够在计算机中完整模拟出某种功能，你就确凿无疑地证明了该过程背后没有任何神秘力量。“一旦你能做出一个能够运作的计算机模拟，你就能在道德层面上百分之百确定：这其中没有任何魔法，没有形态共振，没有超心理学能量，也没有灵体。”丹尼特总结道，这背后全都是平平无奇、实实在在的因果逻辑，计算模拟本身就是对超自然迷思最有力的祛魅。 > “一旦你能做出一个能够运作的计算机模拟，你就能在道德层面上百分之百确定：这其中没有任何魔法，没有形态共振，没有超心理学能量，也没有灵体。有的只是老式的、日常的、确凿而诚实的因果关系。” —— 丹尼尔·丹尼特 ### [22:45](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1365s) 图像直觉泵：从阿舍利手斧到圣家堂 丹尼特指出，直觉泵不一定全由文字构成，有时单凭两张并置的图像就能产生强烈的思维冲击。他展示了进化生物学家兼作家马特·里德利（Matt Ridley）提供的一组对比图：左边是一把远古的阿舍利手斧（Acheulean hand-axe），右边是一只现代计算机鼠标。 这组图像形成了极其尖锐的演化对照：阿舍利手斧由人类祖先制造并传承了上百万年，在这长达百万年的时间里，其形制与设计几乎处于绝对的停滞状态（stasis），没有发生任何明显的风格演变或功能改良。丹尼特风趣地提到，有学者甚至推测它可能根本不是实用工具，而是类似孔雀尾巴的性选择展示品——好比在向同伴炫耀：“瞧我的手斧收藏，我整天打磨手斧却依然能吃饱穿暖照顾好自己，这就像开着一辆兰博基尼。”相比之下，现代鼠标虽然人人熟悉，但它的设计寿命极其短暂，台下几乎没人相信一百年后人类还会使用物理鼠标。这生动展现了生物基因演化时代的超长停滞与文化技术爆炸时代的极速迭代。 随后，丹尼特给出了第二组震撼的视觉对比：左侧是澳大利亚白蚁筑起的高耸白蚁丘城堡（termite castle），右侧则是建筑大师安东尼·高迪（Antoni Gaudí）在巴塞罗那设计的圣家族大教堂（Sagrada Família）。两座建筑在外观乃至内部通风结构的剖面设计上都有着惊人相似的有机形态，且都是由动物创造出的非凡人造物。然而，在相似的表象之下，两者的设计与涌现机制却有着云泥之别，由此引出了自然选择的“无意识盲目设计”与人类智能的“自觉规划”之间的深层哲思。 - 阿舍利手斧与鼠标的对比：自然演化史上的百万年技术停滞 vs 现代文化技术以数十年计的超高速更迭。 - 白蚁丘与圣家族大教堂的对比：同为动物构造的复杂杰作，外观与内部机能高度同构，但其背后的生成机制——盲目的群体自组织与具高度自觉的智能设计——截然相反。 ### [25:00](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1500s) 从白蚁筑巢到高迪设计：自下而上与自上而下的演化飞跃 Daniel Dennett 通过对比两张截然不同的建造与设计图像，引出生物与文化演化需要解答的核心问题。画面左边展示的是自下而上的达尔文式设计与建造过程——白蚁堆。单个白蚁愚昧无知，对自己正在建造宏伟巢穴的整体目标一无所知；蚁群中没有蚁王发号施令，没有建筑大师，没有图纸，没有领导层，也没有层级体系，纯粹是纯粹自下而上的盲目劳作造就了复杂的结构。画面右边则是安东尼·高迪（Antoni Gaudí）的作品。高迪宛如漫画中天才艺术家的夸张写照：专制霸道、独断专行、制定规矩，手中握有蓝图、宣言和翔实的方案规划，他指挥着工程负责人，再指挥底层工人，是典型的层级化自上而下设计。 演化与文化演化面临的关键问题是：地球是如何从左边的白蚁模式跨越到右边的高迪模式的？人类如何从白蚁式的研发跨越到高迪式的研发？Dennett 指出，高迪绝非拥有一个“裸脑”（bare brain），他掌握着前人积累下来的丰富思考工具。正是这些思维工具解锁了顶层设计与“智能设计”的可能性。达尔文向我们证明，生物圈的万千奇迹不需要智能设计者，自下而上、盲目摸索的试错便足以为之；但当文化演化接棒后，演化过程开始自我反转，产出越来越多的真正意义上的智能设计。 以阿兰·图灵（Alan Turing）为例，他比高迪更为纯粹：图灵在造计算机前就已经在理论上证明了其概念可行性，画出了蓝图和完备设计，若没有事先的数学证明，绝不会有人为他造计算机提供资助。想法和概念真正走在了物理建造的前面，这才是货真价实的智能设计。Dennett 强调，智能设计是演化结出的硕果，绝非其源头。正如他幽默点评道：“《圣经》上说‘太初有道’（In the beginning was the Word）。不，不，‘词语’是非常晚近的产物。但‘词语’确实是智能设计的开端，在这一点上《圣经》说得没错。” ```mermaid flowchart TD accTitle: 从自下而上盲目演化到自上而下智能设计的发展路径 accDescr: 展示从白蚁试错到高迪、图灵式顶层智能设计的演化逻辑 subgraph Darwinian["达尔文式底层演化"] A["白蚁式试错研发"] -->|"盲目摸索、无图纸、无首脑"| B["生物圈奇迹"] end subgraph Cultural["文化演化与工具积累"] B -->|"思维工具介入"| C["文化演化加速"] C -->|"逆转设计流程"| D["顶层思考工具赋能裸脑"] end subgraph TopDown["自上而下智能设计"] D --> E["高迪式宏伟建筑与图纸"] D --> F["图灵式先证明、后建造的计算原型"] end ``` - 白蚁巢穴：缺乏中枢协调与意识，完全由局部的无知个体在简单规则驱动下堆砌而成。 - 高迪与图灵：拥有丰富文化工具的典型代表，设计在执行前已在头脑或纸面上完成了验证。 - 演化倒转：生物学依靠自下而上的无目的筛选，而人类文化演化最终孕育出了真正自上而下的智能先导设计。 --- ### [27:43](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1663s) 生命之树图谱：真核革命与文化认知的爆发 为了形象化这一宏大的时间跨度，Dennett 展示了伦纳德·艾森伯格（Leonard Eisenberg）绘制的“生命之树”（Tree of Life）图表。Dennett 演讲时激光笔恰好没电，他索性举起食指打趣：“瞧，这儿有个思考工具！就在我指尖上，多奇妙。”在这张放射状图谱上，时间从中心向外辐射，依次展现地球诞生、生命起源，以及最先出现的单细胞细菌与古菌。 随后生命历程迎来了极其关键的转折点——真核生物革命（eukaryotic revolution）。内共生事件让两种演化了十亿年的独立谱系合二为一，形成了更强大的复合体。Dennett 将其生动比作演化上的“技术转移”（technology transfer），从此它们不必各自重新发明轮子，而能够共享彼此的技术优势，地球上之后所有的复杂生命皆脱胎于真核生物。紧随其后的是寒武纪大爆发（Cambrian explosion），极短时间内爆发式涌现出无穷无尽的生命构型，是极其丰饶的研发狂潮。 沿着时间线延伸到边缘，最近的一小截分叉约代表700万年前人类与近亲黑猩猩分道扬镳的时刻。在这700万年里，真正拥有语言的时光至多不过100万年（阿舍利手斧甚至早于语言出现，这在学界仍存争议）。Dennett 进而把现代人类文明对地球生态的席卷称作“麦克雷迪大爆发”（MacCready explosion），而彻底重构这个星球面貌的根本力量，正是人类发明的各种思维工具。 - 伦纳德·艾森伯格的生命之树设计有海报与文化衫，以直观的放射轴展示各物种分支的深远历程。 - 内共生事件使两个截然不同的谱系合并，终结了彼此孤立试错的历史，是生物史上顶级的技术整合。 - 从距今约700万年的人猿共祖，到距今不超过100万年的语言出现，思维工具的飞速累积最终引发了文明大爆发。 --- ### [30:40](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1840s) 神经外科医生的恶作剧：直面自由意志与神经科学的伪命题 转入新书《直觉泵和其他思考工具》的具体案例，Dennett 带来了一个他专门用来让部分认知科学家和神经科学家坐立难安的直觉泵，以反击那些断言“神经科学证明人类没有自由意志”的轻率言论。故事的科学基础取自阿姆斯特丹极具才华的学者达米安·德尼斯（Damiaan Denys）所研发的真实技术：通过脑内植入微芯片来有效控制强迫症（OCD）。 基于此事实，Dennett 构建了一个科幻情境：一名强迫症患者找到女神经外科医生植入了该芯片。缝合手术完毕、麻醉消退之际，医生神情严肃地对他交代：“顺便告诉你，我的团队将在控制室里全天候 24/7 监控你的一切行动。你会以为自己拥有自由意志，以为所有决定都是自己做的；但实际上，一切决定都是我们替你做出的。对你来说，自由意志只是幻觉。祝你生活愉快，再见。”患者面对窗明几净的现代化实验室和白大褂权威，对此深信不疑。由于坚信自己不再能控制自己，他变得放任自流、生活邋遢、充满攻击性，最终身犯重罪被押上法庭。 面对法官指控，男子辩解自己毫无自由意志，一切全由神经外科医生幕后操控。医生被传唤出庭作证，在宣誓后坦白：“没错，我是这么跟他说的。但那只是个玩笑，我只是在跟他闹着玩。”现场听众无不承认，这位医生的所作所为不仅刻薄且严重失职，更是对患者造成了实质性的不可逆伤害。Dennett 总结道：“她用语言，非手术地达成了她声称通过手术达成的效果。仅仅通过说服他放弃自由意志，她就把他作为一个道德主体给废掉了。” Dennett 借此猛烈抨击那些动辄宣称“脑科学证明自由意志不存在”的科学家同行。他们给出的理由往往极其苍白——例如“大脑中当下的每个事件都是由先前的大脑事件所决定的”。Dennett 认为因果决定本身根本无法否定自由意志的存在，科学家使用的是一种极其幼稚、狭隘的自由意志定义，但他们瓦解的却是维系整个社会运转的道德责任基石，这种科普言论是极度不负责任的。 --- ### [33:54](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2034s) 两张彩票与宇宙不公：决定论完全相容于自由意志 许多人凭直觉认为，只要大脑中的每一个事件都是由物理定律决定的，自由意志就必然荡然无存。针对这种普遍错觉，Dennett 抛出了“两种彩票”的直觉泵： - 彩票 A：在所有彩票售出**之后**，现场摇号抽出中奖号码。这是绝大多数现实彩票的运作机制。 - 彩票 B：在彩票售出**之前**，主办方就已经提前选好了中奖号码，并将存根秘密锁进保险箱，直到所有人买完彩票后再开箱对奖。 Dennett 问台下听众：买彩票 B 的人是傻子吗？是不是只有彩票 A 才有中奖机会？全场观众很快认同两种彩票的中奖几率完全均等。出版商清算所（Publishers Clearing House）就常用“您可能已经中奖了”作为营销噱头，大众完全理解即便结果提前注定，参与者依然享有对等的“机会”。 ```mermaid flowchart LR accTitle: 两种彩票直觉泵对比 accDescr: 说明提前确定结果与事后随机产生结果在概率与自由机会上具有一致性 subgraph LotteryA["彩票 A（非决定论类比）"] A1["彩票售出"] --> A2["现场摇号生成中奖号码"] A2 --> A3["中奖概率均等，公平"] end subgraph LotteryB["彩票 B（决定论类比）"] B1["中奖号码预先封存入保险箱"] --> B2["彩票售出"] B2 --> B3["开箱揭晓，中奖概率同样均等"] end ``` Dennett 进而将此映射至决定论的世界：假若决定论为真，无非是意味着所有的“彩票号码”在大爆炸之初就已经被提前生成，并封存在你体内的一只隐形小保险箱里。在人生道路上，每当你需要一次掷硬币式的随机决断或遭遇混乱时，就从信封里抽出一张——有输有赢。这与现场发生量子层面的随机奇迹相比，所拥有的自由度没有任何区别。有人或许抗辩说，如果结果一早注定，某些人注定走运，某些人注定倒霉；但 Dennett 反驳道，即使非决定论成立，现实依然是有人注定获胜有人注定落败，你同样事先不知晓结果。 公平并不等同于永远获胜，运气本身就是存在的，而且运气与自由意志完全相容。著名哲学家大卫·威金斯（David Wiggins）曾撰文探讨“决定论的宇宙级不公”（the cosmic unfairness of determinism），但 Dennett 质问道：非决定论的宇宙级不公又当如何？生活中本就充满了随机的不公，但这丝毫不影响我们成为能够承担责任的行为主体。Dennett 斩钉截铁地下了定论：“成为一个负责任的行为主体，拥有真正重要的那种自由意志，与底层物理学究竟是决定论还是非决定论毫无干系。” - 彩票直觉泵证明：“结果已提前定下”并不剥夺当事人的客观概率与现实选择空间。 - 将决定论等同于剥夺机会，是混淆了“预先决定”与“操纵干预”之间的哲学差异。 - 无论物理宇宙是决定论的齿轮还是随机涨落的量子池，人类在宏观社会与伦理层面构建的道德主体地位都坚实有效。 ### [37:11](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2231s) 道德能力与《直觉泵》的收尾总结 Daniel Dennett 继续阐述关于大脑与责任的观点：一个人的大脑是决定论的还是非决定论的，根本无关紧要。真正起作用的是大脑中安装的“软件”，即从小到大逐步建立起来的行为能力。如果一个人接受过良好的道德教养，在大多数情况下就会在正确的时间做出正确的决定；这种普遍的胜任能力使得当其犯错时，必须承担责任。Dennett 总结道：“Why? Because you're good enough to be responsible.”（为什么？因为你的能力足以承担责任。）由于这背后是一套庞大而复杂的理论，他决定在此结束演讲的主体部分。 在谈到新书《直觉泵和其他思考工具》（*Intuition Pumps and Other Tools for Thinking*）时，Dennett 形容它就像一本“西班牙小吃（tapas）式的书”。整本书包含 77 个章节，长短不一，有的只有 1 页，有的长达 10 页左右。每一章都介绍了一种具备独立自主性的思考工具，作者致力于让它们便于携带且易于理解与使用。他向观众建议：拿去尝试、带去兜兜风，最重要的是“转动所有的旋钮，看看它们到底是如何运作的”。演讲在现场掌声中进入问答环节。 ```mermaid flowchart TD accTitle: 道德能力与责任的关系架构 accDescr: 展示从儿童时期的道德教养到大脑安装软件能力，进而形成胜任力并承担道德责任的过程。 A["儿童时期的道德教养与输入"] --> B["大脑中安装的软件与行为胜任力"] B --> C["在多数情况下做出正确决定"] C --> D["具备完全的行为能力与道德责任"] B -.->|"在发生过失时"| D ``` --- ### [38:20](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2300s) 智能设计、自然选择与大脑及经济的组织形式 观众席的第一位提问者提出：智能设计在某些领域（如设计精良的手机或电脑）表现优异，但在经济体系等复杂系统中，人类至今尚未证明自己能够通过智能设计来驾驭它；对此，提问者询问对不适宜智能设计的事物有何见解，以及未来人类是否终将拥有设计一切事物的恰当工具。 Dennett 回应称，这涉及一个已被深入探讨的权衡取舍：自然选择是一台强大的引擎，以混沌和所谓的随机性为养料，但其工作原理是试错（trial and error），伴随着海量的错误与极其巨大的浪费。经济体系也是如此，因此某种陈旧且粗鄙的社会达尔文主义观点——经济中的适者生存——确实包含一部分事实。然而，Dennett 强调：“But that doesn't mean that we should simply honor that and make that a moral principle. Because I think we have very good reasons to think that an unfettered, completely chaotic, laissez-faire, dog-eat-dog economy is not a good way to go.”（但这并不意味着我们应该尊崇这一点并将其立为道德准则。因为我认为我们有非常充分的理由相信，一种不受约束、彻底混乱、放任自流、弱肉强食的经济并非良策。）放任自流会导致赢家通吃，造成极大的伤害。 Dennett 借此将话题延伸至大脑研究。他认为大脑确实是计算机，但与常规的笔记本电脑截然不同： - 大脑拥有 1 亿或 2 亿（注：此处为演讲口误或泛指局部回路，全脑神经元约为 860 亿）各具独立“议程”的神经元，如同微小的白蚁，彼此各不相同。 - 大脑中的“软件”通过形成神经元联盟并促成协同工作来运作，心理上的统一性（unity）是一种后天达成的成就，绝非自上而下预先内置的。 - 经济体系的最佳架构可能与大脑类似：既不能依赖过度的自上而下中央集权控制，但也必须配备适度的监管与秩序维护（policing），以防范系统失控带来的严重弊端。 --- ### [41:32](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2492s) 纽康姆悖论的抉择难题 第二位观众询问 Dennett 对纽康姆悖论（Newcomb's Paradox）的看法。Dennett 随即向全场拆解了这个经典的决策论谜题。他将情境比作电视答题节目：参与者面临“选一个箱子（one-boxer）”还是“选两个箱子（two-boxer）”的抉择。A 箱中装有确定的 1,000 美元；B 箱中可能装有 100 万美元，也可能是空箱（0 美元）。如果仅仅是从 A 和 B 中二选一，那只是一场在确定的 1,000 美元与可能赢取 100 万美元之间的博弈；但纽康姆悖论的特殊规则是：你可以选择只拿 B 箱，或者把 A 箱和 B 箱同时带走。 从纯粹的支配性决策原则来看，似乎理所当然应该选两个箱子，因为无论 B 箱里有什么，选两个箱子都能额外确保拿到 A 箱里的 1,000 美元。但节目的主持人随后补充了一个关键条件： - 你是节目历史上的第 5,000 位选手。 - 节目主办方拥有极其精准的预测模型，能够预判选手会选一个箱子还是两个箱子。 - 如果主办方预测你会选两个箱子，他们就不会在 B 箱里放钱，此时 A 箱只有 1,000 美元；只有当主办方预测你只拿一个箱子时，才会将 100 万美元放入 B 箱。 - 主办方的预测战绩近乎完美：虽然偶有少数意外（例如几百个人中只有一人拿了两个箱子还侥幸拿到全部奖金，或是选了 B 箱却发现是空箱），但在极高统计概率上高度相关。 Dennett 现场询问观众有多少人选一个箱子、多少人选两个箱子，并指出该悖论的迷人之处恰恰在于其绝不显而易见，学界为此倾注了大量心血。针对自己的立场，他坦言：“And I have, myself, not spilled any ink because I still don't have a firm view on it.”（而我自己没有在这个问题上费过笔墨，因为我至今仍未形成坚定的观点。）他倾向于认为该思想实验的设定本身可能存在某些问题，但极难证实；他承认自己立场摇摆不定，“有些日子我是单箱派，有些日子我是双箱派”。 --- ### [44:51](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2691s) 自由意志、决定论与剪刀石头布 第三位观众提及 Douglas Hofstadter（侯世达）关于决定论与自由意志属于不同层面问题的讨论，并询问 Dennett 的视角。Dennett 透露，这恰好是他与 Hofstadter 多年来少数存在严重分歧的议题之一。Hofstadter 对此持悲观态度，认为那种“真正关键的自由意志”在决定论框架下是无法拥有的；而 Dennett 虽然写了多本书、耗费大量时间试图说服 Hofstadter，但至今未获成功。不过 Dennett 幽默地补充说，Hofstadter 并没有像思想实验里那个崩溃的主人公一样做出不负责任的行为。 Dennett 分析指出，“自由意志”并非一个科学术语，而是源自哲学的“显象世界/日常意象”（manifest image），存在于颜色、美元和棒球的日常常识与法律传统之中，而非神经元、蛋白质、原子或物理定律的世界。在日常领域中，自由意志承担了两种定义： 1. **形而上学定义**：事件无前因（uncaused），不存在充分的因果先决条件。 2. **道德能力定义**：成为具备道德行为能力的合格主体（例如有资格贷款、买车，并出于“自身自由意志”签署具备法律效力的合同）。 Dennett 坚信后者才是唯一真正重要的自由意志概念，而这一概念与决定论是否成立毫无干系：“For instance, I don't think your competence to sign a contract would be disparaged in the slightest if determinism were true.”（例如，如果决定论是真的，我认为你签署合同的胜任能力丝毫不会因此贬损。） ```mermaid flowchart TD accTitle: 自由意志的双重定义对照 accDescr: 对比形而上学定义的无原因自由意志与法律道德上的行为能力自由意志。 subgraph Manifest["日常显象意象中的自由意志"] A["自由意志词源：法律、传统与日常经验"] A --> B["定义一：因果断裂 / 无前因（Uncaused）"] A --> C["定义二：道德与法律胜任能力（Competence）"] end B -->|"形而上学诉求"| D["要求非决定论（物理层面）"] C -->|"实践诉求"| E["能够签署契约、承担民事与道德责任"] E --> F["与决定论完全相容，无需依赖物理随机性"] ``` 在讨论脑机制是否偏好非决定论时，Dennett 承认自己曾做出微调：一位坚定的非决定论支持者曾说服他接受一种唯独需要非决定论的特殊情境——“与上帝进行高赌注的‘剪刀石头布’比赛”。在玩剪刀石头布时，最佳策略是彻底随机化以防对手识破模式；如果与全知全能的上帝对决，事先背诵伪随机数表会被洞悉，唯有真随机（非决定论）才能防止被读心。但 Dennett 打趣道，自己未来并不指望会和上帝玩猜拳，因此很乐意生活在一个决定论的世界中，并对自己的行为承担责任，哪怕这些行为早已由自己出生前后的无数前因所注定。 最后，针对观众关于意识是高级生物系统的“副现象”（epiphenomenon）还是如同时间空间一般的“基本物理维度”的提问，Dennett 立即指出这个“二选一”的问法带有某种预设立场的偏颇（rathering），并明确表态：他绝不认为意识是一个将宇宙一分为二的根本物理范畴。 - 关键观点总结： - 道德责任建立在主体的认知与执行胜任能力（competence）之上，而非物理层面的非决定论。 - 大脑结构是由大量具有独立属性的神经元以自下而上的竞争与联盟机制组成的，这一模式同样启发了非绝对中央集权但具备监管规则的经济体系。 - 自由意志的核心价值在于社会、法律层面的理性承诺与道德契约能力，该能力在因果决定论世界中依旧完整有效。 ### [50:26](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=3026s) 意识作为“安装的软件”与野生人类思想实验 Daniel Dennett 深入探讨了意识的本质。他提出，意识与文化深度交织，并将其比作繁殖、新陈代谢、自我修复、愈合以及血液凝固等生理过程——它是一种极为复杂且深度递归的生物学现象。Dennett 强调，意识存在多种含义，但对我们而言最关键的“基准点”是我们在日常生活与文学阅读（如小说）中所体验到的意识生活。他直言不讳地指出，人类是唯一拥有这种意识生活的生物：狗或海豚虽然有感知力（sentient）、清醒且能感知痛苦，但它们并不具备构成人类意识重要性的绝大部分内容。 在 Dennett 看来，这种独特的意识现象在本质上相当于“软件”，而且必须后天“安装”。为了说明这一点，他提出了一个假设的思想实验（并幽默地提醒观众“千万别在家里尝试”）：如果让人类婴儿在完全脱离人类互动、隔绝语言的环境中长大，在成年过程中从未获得任何在社会共同体中自然“安装”的思维工具，那么这个人的心智将会与我们产生极其深刻的差异，以至于我们大家都会一致认为他或她根本不算真正具有意识。关于人类独特的意识形态，Daniel Dennett 总结道：“And I think that phenomena is, in the end, software. And I think it has to be installed.” ```mermaid flowchart TD accTitle: Dennett's Model of Human Consciousness as Installed Software accDescr: A flowchart contrasting basic biological sentience with human consciousness acquired through cultural software. subgraph Biological["生物学生理基底"] A["基本感知能力 (Sentience)"] B["清醒与痛觉感知"] end subgraph Cultural["社会文化安装包"] C["语言系统"] D["思维工具与叙事理解"] end subgraph Outcome["心智状态区分"] E["非人动物/野生人: 仅具备生物感知"] F["完全人类意识: 具备递归与反思能力的心智生活"] end Biological --> A & B A & B --> E Biological --> F Cultural --> C & D C & D --> F ``` ### [52:22](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=3142s) 感受质之迷与“真正魔术”的陷阱 针对哲学界关于色彩体验、味觉体验以及所谓感受质（qualia）的讨论，Dennett 认为这些概念充满了严重的混乱。他调侃道，许多人之所以感到困惑，是因为他们非要让意识成为一个用行话来说“奇妙不可思议”（supercalifragilisticexpialidocious）的东西；如果讨论的对象不够神乎其神，他们就会抗议说这根本不是在谈意识。 为了揭示这种认知误区，Dennett 引用了哲学家兼优秀舞台魔术师 Lee Siegel 在其关于印度街头魔术的著作《Net of Magic》中的经典段落：Siegel 告诉朋友自己在写一本关于魔术的书，朋友问他写的究竟是不是“真正的魔术”（意指通灵、奇迹等超自然法术），Siegel 回答不是，他写的是手法戏法和舞台魔术。Siegel 在书中写道：“换句话说，现实中你能真正做到的魔术，不是真正的魔术；所谓的‘真正魔术’，是那种根本不存在的魔术。” Dennett 感慨，自己的整个学术生涯都在与持这种态度的人打交道。无论是在自由意志还是意识议题上，许多人都执念于“真正的魔术”：有人坚称自由意志要求“每一个自由决策都是一个小奇迹”，否则就是廉价的替代品。但 Dennett 表示，自己完全支持这些所谓的“廉价替代品”，因为科学足以解释为什么我们拥有“值得企求的自由意志”，这种自由意志足以确保我们过上有意义、有担当的负责任生活，而无须仰赖一种将宇宙生生劈成两半的神秘主义奇迹。 ### [55:22](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=3322s) 惩罚的辩护：反对报应主义与消除主义 当观众提问关于功利主义与道德推理的看法时，Dennett 说明自己不是普通的功利主义者，而是一名结果主义者（consequentialist），注重行为的后果，坚决不走纯粹的道义论（deontological）路线。他以此延伸到自己正在进行的关于自由意志与惩罚的研究。他曾在文章中表示“不想生活在一个没有惩罚的世界”，导致许多人误骂他是“肮脏的报应主义者”（retributivist）。 Dennett 明确澄清自己绝非报应主义者。他举出康德作为典型的报应主义者代表：康德甚至断言，哪怕明天就是世界末日，人类最高的职责之一仍是走进监狱将死囚处决，因为在康德看来惩罚本身就是善，未受惩罚的罪行是对自然或上帝的亵渎。Dennett 坚决反对惩罚是内在之善，但他认为惩罚是基础性的善，理由在于：一个由信任构成的社会离不开“承诺”，而如果不遵守承诺没有任何不良后果，承诺便不复存在；而这种后果不能仅仅是康复改造（rehabilitation）或思想重塑，必须是具有威慑力的“惩罚”。 Dennett 强调，现行的刑事惩罚制度极其残酷丑陋、亟待彻底改革，但这绝不意味着应当像当前一些认知科学家与神经科学家主张的那样彻底废除惩罚。他警告说，这种主张在历史上曾带来惨痛教训——斯大林时代的苏联正是宣称不对异见者进行“惩罚”，而是进行“治疗”（treated）。在界定公民与法律制度的关系时，Daniel Dennett 掷地有声地指出：“I'd rather be a free citizen in a free society. I accept that if I do something really wrong I am eligible for punishment and that's OK.”如果存在明确的规则和惩罚，人们在服刑期满后便能重新作为清白平等的公民回到社会，而不是被当成病人无休止地“被治疗”。 ### [58:38](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=3518s) 软件专利与“实用信息”的哲学界定 最后一位观众抛出了当时美国联邦巡回区上诉法院审理的一起关于软件是否具备可专利性的争议案件：一派观点认为软件代码类似自然法则，只是人人皆可利用的基础工具，其特定组合不应授予专利；另一派则认为计算机载入特定软件后就变成了一台全新的“设备”，自然属于可专利范畴。 Dennett 回应称，自己长期思考专利、版权（著作权）与商业秘密之间的深层边界，并认为解决这一难题需要跳出纯粹的物理信息论。他主张建立一种有别于“香农-韦弗信息”（Shannon-Weaver information）的信息概念：它固然以香农信息为载体，但其核心是语义信息与认知信息，即“有用信息”（useful information）。Dennett 将这种信息定义为：能够为了某种效用或目的而传递适应性（transmitting adaptations）的结构。 从效用角度审视，界定软件领域的“盗窃”行为必须看其核心效用是否被侵占；若是核心效用被盗，法律就必须设立惩罚予以遏制。虽然这一逻辑与现行版权法存在交织重叠，但 Dennett 强调，正如他在书中所讲的各类作为“思维工具”的 App 一样，软件本质上是人类精心设计出来的机制（mechanisms that have been designed）。基于软件具备明确机械设计与效用这一前提，Dennett 倾向于支持软件具备可专利性。演讲在全场热烈的掌声中圆满结束。 ## Notable quotes > [00:04](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=4s) Life on Earth has been generated over billions of years by an algorithmic process. — Daniel Dennett > [03:40](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=220s) You can't do much carpentry with your bare hands, and you can't do much thinking with your bare brain. — Bo Dahlbom > [06:55](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=415s) Philosophy is what you're doing when you're not quite sure what the right questions to ask are. — Daniel Dennett > [19:41](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=1181s) 一旦你能做出一个能够运作的计算机模拟，你就能在道德层面上百分之百确定：这其中没有任何魔法，没有形态共振，没有超心理学能量，也没有灵体。 — Daniel Dennett > [33:54](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2034s) 成为一个负责任的行为主体，拥有真正重要的那种自由意志，与底层物理学究竟是决定论还是非决定论毫无干系。 — Daniel Dennett > [55:22](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=3322s) I'd rather be a free citizen in a free society. I accept that if I do something really wrong I am eligible for punishment and that's OK. — Daniel Dennett > [44:51](https://www.youtube.com/watch?v=4Q_mY54hjM0&t=2691s) For instance, I don't think your competence to sign a contract would be disparaged in the slightest if determinism were true. — Daniel Dennett ## Mentioned Daniel Dennett, Peter Norvig, Rob Pike, Bo Dahlbom, Jim Flynn, John Searle, Douglas Hofstadter, John Stuart Mill, Paul MacCready, Hao Wang, Matt Ridley, Antoni Gaudí, Alan Turing, Leonard Eisenberg, Damiaan Denys, Publishers Clearing House, David Wiggins, Immanuel Kant, Lee Siegel, Google, Intuition Pumps and Other Tools for Thinking, Darwin's Dangerous Idea, Consciousness Explained, Breaking the Spell, Gossamer Albatross, Sagrada Família, Net of Magic ## Video description Professor Dennett comes to Google to talk about his new book, Intuition Pumps and Other Tools for Thinking. Dennett deploys his thinking tools to gain traction on these thorny issues while offering readers insight into how and why each tool was built. Alongside well-known favorites like Occam's Razor and reductio ad absurdum lie thrilling descriptions of Dennett's own creations: Trapped in the Robot Control Room, Beware of the Prime Mammal, and The Wandering Two-Bitser. Ranging across disciplines as diverse as psychology, biology, computer science, and physics, Dennett's tools embrace in equal measure light-heartedness and accessibility as they welcome uninitiated and seasoned readers alike. As always, his goal remains to teach you how to "think reliably and even gracefully about really hard questions." About the Author: Daniel C. Dennett is the Austin B. Fletcher Professor of Philosophy at Tufts University and the author of numerous books including Breaking the Spell, Darwin's Dangerous Idea, and Consciousness Explained. [Source](https://www.youtube.com/watch?v=4Q_mY54hjM0)

[0:04] Daniel Dennett: Life on Earth has been generated over billions of years by an algorithmic process. [3:40] Bo Dahlbom: You can't do much carpentry with your bare hands, and you can't do much thinking with your bare brain. [6:55] Daniel Dennett: Philosophy is what you're doing when you're not quite sure what the right questions to ask are. [19:41] Daniel Dennett: 一旦你能做出一个能够运作的计算机模拟，你就能在道德层面上百分之百确定：这其中没有任何魔法，没有形态共振，没有超心理学能量，也没有灵体。 [33:54] Daniel Dennett: 成为一个负责任的行为主体，拥有真正重要的那种自由意志，与底层物理学究竟是决定论还是非决定论毫无干系。 [55:22] Daniel Dennett: I'd rather be a free citizen in a free society. I accept that if I do something really wrong I am eligible for punishment and that's OK. [44:51] Daniel Dennett: For instance, I don't think your competence to sign a contract would be disparaged in the slightest if determinism were true.

[34:16] Daniel Dennett: Being a responsible agent, having free will in the sense that matters, has zip to do with whether physics is deterministic or indeterministic. [3:40] Bo Dahlbom: You can't do much carpentry with your bare hands, and you can't do much thinking with your bare brain. [55:22] Daniel Dennett: I don't think we want to be treated. I think we want to be punished. Because then there's rules. And you can serve your time and get out and be a citizen again. [50:26] Lee Siegel: The kind of magic you can actually do is not real magic. Real magic is a kind that doesn't exist. [55:22] Daniel Dennett: I want to live in a world in which there are promises. And you can't really have promises unless there's consequences for not keeping your promise. [20:36] Daniel Dennett: You know to a moral certainty, there's no magic. There's no morphic resonance. There's no psionic forces or ectoplasm. It's good old-fashioned, every day, hardened, honest causation. [0:04] Daniel Dennett: Life on Earth has been generated over billions of years by an algorithmic process.

[0:04] PETER NORVIG: Welcome everybody. [0:07] It's our pleasure to have with us at Google [0:09] today, Daniel Dennett. [0:11] He has a long and distinguished career. [0:12] I won't try to summarize all of it. [0:15] You can Google it. [0:16] [LAUGHTER] [0:17] PETER NORVIG: Instead, I'll just try to introduce very [0:20] briefly the book that he's written most recently and is [0:24] talking about today, which is a collection or a [0:27] tool kit for thinking. [0:29] And I think the reason I think it's such a great book is [0:33] because a lot of it is our tool kit. [0:35] It's these computational tools for thinking that we take for [0:41] granted, but Dan is trying to introduce to the rest of the [0:45] world who sometimes aren't aware of how [0:47] to think like that. [0:50] And so what do I mean by that? [0:51] Well, in his 1995 book on evolution, he wrote that, [0:58] "Life on Earth has been generated over billions of [1:00] years by an algorithmic process." And that's kind of a [1:04] natural way for us to think about it, but it's hard way [1:06] for other people to think about. [1:08] And some people don't quite understand that by taking lots [1:12] and lots of small steps that you could arrange and come up [1:16] with this result. [1:18] And other philosophers, who we won't name, have taken the [1:23] approach, well look at these incredible results. [1:25] And if I simulate the steps for three or four steps I [1:29] don't get there. [1:29] So it must be magic. [1:32] Rather than saying, well, maybe a billion or 100 billion [1:36] is actually qualitatively different than a couple steps. [1:39] So Dan reminds us how to think that way. [1:41] He also reminds us that there can be emergent properties [1:46] that aren't implicit in any one piece of an ensemble and [1:51] yet they come out in the end. [1:53] We're all familiar with that. [1:55] Our programs have emergent properties [1:57] that we didn't expect. [1:59] And we bang our heads trying to fix them. [2:02] So we're aware of it. [2:03] But the rest of the world isn't. [2:05] So that's why they need a book like this. [2:08] I just want to talk about one more of the tools. [2:13] And this is the using lay audience as a decoy. [2:17] So this is a trick if you have a philosophy professor who's [2:22] not explaining all the details, rather than ask them [2:25] to explain a few you say, oh, well why did you give a [2:27] lecture to this introductory class. [2:30] And then it will become more clear because the professor [2:33] will have to work harder to get it to that level rather [2:35] than trying to keep it at the professor to professor level. [2:40] One of our Google employees Rob Pike wrote a book where he [2:44] talked about a similar approach. [2:46] Instead of calling it the lay audience as decoy approach, he [2:49] called it the teddy bear approach. [2:51] Which is, there was a introductory class in computer [2:56] programming for which the TAs were overstocked. [2:59] And there was always a long waiting line of students [3:01] waiting to get in to talk with the TAs about why the program [3:03] didn't work. [3:04] So what they instituted was a chair that held a teddy bear [3:09] outside their office. [3:11] Before you were allowed to talk to the TAs you had to [3:13] talk to the teddy bear. [3:15] And so the students would come up and say, oh teddy bear, my [3:18] program doesn't work. [3:19] And it must work because this loop, see, this could never be [3:22] more than one. [3:23] Oh, except, oh, never mind. [3:26] And it turns out the teddy bear successfully solved about [3:30] half the problems. [3:31] And people didn't have to go on to the TA. [3:33] So I think that's sort of in the same toolbox. [3:36] And Dan will tell you about the rest of the tools. [3:40] DANIEL DENNETT: Thanks, Peter. [3:40] That's a great introduction. [3:41] I'm going to add the teddy bear. [3:44] It occurred to me as you were telling us about that, that [3:46] that probably explains the success rate of psychoanalysts [3:49] right there. [3:51] You can replace them with a teddy bear and [3:53] get the same results. [3:54] AUDIENCE: [INAUDIBLE]. [3:56] DANIEL DENNETT: Yeah, Teddy bears are a lot cheaper. [4:01] So I want to talk about better tools for thinking. [4:06] One of my best students, a Swede name Bo Dahlbom once [4:09] said, "You can't do much carpentry with your bare [4:12] hands, and you can't do much thinking with your bare [4:15] brain." He hits the nail on the head right there. [4:22] It's obvious, in a way, when you think of it, even [4:24] something as simple as a word is a tool. [4:27] It's a thinking tool. [4:29] And these thinking tools have been around for a long time. [4:31] And recently we've learned that maybe these thinking [4:35] tools are having a remarkable effect. [4:38] How many of you know what the Flynn effect is? [4:40] Oh, a few of you do. [4:41] And the rest of you, all right, now let me tell you [4:43] about the Flynn effect. [4:44] Jim Flynn, it's named after him, but he's not really the [4:47] one that did the first research. [4:48] He's the one who drew people's attention to it. [4:51] It turns out that IQ worldwide is up. [4:59] And it's up a significant amount. [5:02] If you use today's test, if you gave that test to a person [5:08] who got 100 back in the '30s they get probably an 80. [5:13] And we'd be viewed as almost retarded. [5:16] I mean, there's a big jump. [5:18] It's like 20 or 30 IQ points. [5:20] Of course, IQ tests are always scaled with 100 as average. [5:24] But if you use the same test items, at the very least we've [5:29] gotten a lot better at taking IQ tests. [5:33] And it has nothing directly to do with how much tutoring or [5:36] schooling or your genes. [5:39] It's a very large effect. [5:42] And it can't be due to improvements in our genes as [5:46] we're talking about such a short time periods. [5:49] Jim Flynn thinks-- and I hope he's right. [5:52] That doesn't mean he's right, but I think there's a lot to [5:57] be said for it-- that what's happened is that the thinking [6:01] tools that are first developed and tested and debugged by [6:04] scientists and engineers and other professional thinkers [6:10] have been filtering down into every day people's thinking. [6:17] Even as simple an idea as percentage. [6:21] Which kids, you just pick it up very fast. [6:25] And 50 years ago, 100 years ago, that was a pretty [6:28] complicated concept for a lot of people. [6:31] They just weren't ready for that. [6:32] And so probably it's all these thinking tools that we install [6:39] without even realizing we're doing it. [6:42] We install in on our own brains and in other brains. [6:45] Probably that is a main source of our improved thinking. [6:50] And certainly that is the thesis that I want to explore, [6:54] in any case. [6:55] Because what I'm doing in this book, as a philosopher, I [7:00] discovered that a lot of my colleagues are remarkably [7:04] unselfconscious about the tools of the trade. [7:07] Weird because philosophers are famously self conscious and [7:10] reflective and navel-gazing. [7:12] But a lot of them don't ever ask wait a [7:15] minute, how do I do that. [7:16] And why do I think that's a good way of proceeding? [7:18] What tools am I suing? [7:20] And I wanted them to become more self conscious. [7:22] I want everybody to become more self conscious about the [7:26] tools that they use. [7:30] So what kind of thinking tools are there? [7:32] There's words, as I say. [7:33] And numbers, of course. [7:34] And diagrams and maps and methods. [7:36] And intuition pumps. [7:39] Which is a term I coined more than 30 [7:42] years ago to describe-- [7:45] well in the first instance it was John Searle's notorious [7:48] Chinese room. [7:50] And since I was beating up on that intuition pump, a lot of [7:54] people thought I thought that intuition pumps were bad. [7:57] No, I think they're wonderful. [7:58] It's just got to have good ones. [8:00] There's defective pumps. [8:02] And there's good ones. [8:03] And the way you tell the difference is, as you know, by [8:06] reverse engineering. [8:08] Doug Hofstadter came up with the lovely phrase, you want to [8:12] turn all the knobs on an intuition pump to see what [8:15] makes it do its work. [8:16] You want to try out the variation. [8:18] Nothing mysterious about why that works. [8:20] Is just Mill's method of differences. [8:22] You're just empirically testing to see what variations [8:25] make what differences. [8:26] See how robust the intuition pump is under deformation. [8:31] And sometimes they're remarkably sensitive to tiny [8:35] variations in the way you tell a story. [8:37] Then you know it's probably a defective pump. [8:39] Not doing what you think it's doing. [8:42] So what intuition pumps are are persuaders. [8:45] And it's important that they're persuaders. [8:48] So much good thinking is accomplished in a friendly but [8:55] competitive environment where I am trying to persuade an [9:00] interlocutor, a target, that P and the target is trying to [9:05] persuade me that not P. And this opponent process is [9:10] actually a good way, very often, of [9:11] getting at the truth. [9:14] So they're also discoverers. [9:16] They're attention-holders. [9:18] And they're anchors for a fixed points. [9:20] In my work in philosophy there are no fixed points. [9:27] That's the point a philosophy. [9:28] Philosophy is what you're doing when you're not quite [9:31] sure what the right questions to ask are. [9:33] And everything is up for reconsideration always. [9:37] But how do you proceed then? [9:41] You say, well, maybe this as a fixed point for this [9:44] discussion. [9:44] And then you sort of put a pin down here. [9:47] And here's one here. [9:48] And here's one here. [9:49] And then you say, well, suppose we [9:50] take those on board. [9:52] What can we do with this idea with a few of these candidate [9:55] fixed points? [9:57] One thing you may do with them is throw all three of them out [9:59] or throw out one against the other two. [10:01] It's in this intellectually informal. [10:08] There's no algorithms for doing this, certainly not yet. [10:11] This is an informal exploratory process for which [10:16] there are really no rules. [10:19] But intuition pumps are persuaders to use in those [10:23] harrowing circumstances. [10:28] Here's one that isn't in the book, but I've talked about it [10:30] a lot recently and I think it's important to understand [10:33] the context. [10:34] The MacCready explosion. [10:35] This is Paul MacCready. [10:37] How many of you know Paul MacCready? [10:40] Oh well good. [10:41] I can educate you about that. [10:43] You will know him, not by name, as the inventor of the [10:46] "Gossamer Albatrosss." A visionary engineer. [10:51] The greenest of green engineers. [10:53] Wonderful guy. [10:54] Died a few years ago. [10:56] And in a Ted talk a few years back he came up with the [11:01] following startling observation. [11:05] Let's go back 10,000 years to the dawn of agriculture. [11:10] At that point, human beings all around the world we're [11:13] beginning to domesticate animals and plants, of course. [11:17] And at that point, our species, Homo sapiens, and [11:21] these are, of course, entirely modern Homo sapiens. [11:25] At that point 10,000 years ago, if you put them on one [11:29] side of the balance scale together with their pets and [11:32] livestock, all the domesticated vertebrates on [11:37] one side, including us. [11:39] And you put all the rest of the terrestrial vertebrates on [11:45] the other side. [11:45] Animals, not the insects, not the worms, not the fish. [11:50] He calculated that at that time, 10,000 years ago, we and [11:57] our livestock accounted for a fraction of 1% by weight of [12:03] the terrestrial vertebrate biomass. [12:07] What do you think the percentage is today? [12:11] Any guesses? [12:12] AUDIENCE: 10%. [12:13] DANIEL DENNETT: 10%. [12:15] AUDIENCE: Over 50%. [12:16] DANIEL DENNETT: Over 50%. [12:17] AUDIENCE: 99%. [12:19] DANIEL DENNETT: 99%, are you kidding? [12:21] No it's 98%. [12:24] It is 98%. [12:29] We human beings have engulfed the planet. [12:34] This is one of the most astonishing biological [12:38] phenomena that has ever happened. [12:41] And here's what Paul said about this. [12:44] "Over billions of years on a unique sphere, chance has [12:49] painted a thin covering of life-- [12:52] complex, improbable, wonderful, and fragile. [12:58] Suddenly we humans have grown in population, technology, and [13:03] intelligence to a position of terrible power: we now wield [13:07] the paintbrush." [13:10] I've highlighted the technology and intelligence [13:12] because those are the thinking tools. [13:14] The only difference between us and our ancestors of 100,000 [13:18] years ago is the thinking tools. [13:20] We have the same genes. [13:22] We have pretty much the same justice systems. [13:25] The same muscles. [13:27] The same brains. [13:29] The thinking tools have changed us and made all of [13:33] this power possible. [13:34] It is the source of our power. [13:37] Which raises a sort of chicken egg problem. [13:41] Did evolved tools make us smarter? [13:44] Or did we evolve to become smart enough to make tools? [13:50] And like all good chicken and egg questions, [13:53] the answer is yes. [13:57] It's a coevolutionary, bootstrapping process where [14:01] you have a little evolution makes us smarter. [14:05] And then we can be smarter about making tools. [14:08] We became can become self conscious about [14:10] the tools we make. [14:11] Our ancestors weren't self conscious. [14:13] They didn't invent words. [14:14] They didn't coin words. [14:16] They just found themselves using words. [14:18] And it made them smarter. [14:20] Now we are very deliberate. [14:23] We have become intelligent designers of tools. [14:26] But initially we weren't intelligent [14:28] designers of tools. [14:29] We were simply the beneficiaries of tools. [14:32] Well how did they get there then? [14:34] Not by the genes. [14:35] They evolved, culturally. [14:37] My next book is going to be on cultural [14:39] evolution and that story. [14:41] But I'm not going to say more about it now because I'm going [14:44] to be looking today at recent tools intelligently designed, [14:49] supposedly intelligent designed thinking tools. [14:57] Everybody knows reductio ad absurdum where this is the [15:00] sort of crowbar of logical argument. [15:04] It's what you use. [15:05] You take your opponent, your target's premise. [15:09] You take your target's thesis on board, [15:13] for the sake of argument. [15:15] And then you use the crowbar. [15:17] You try to deduce a contradiction from that. [15:21] And if you succeed you win. [15:24] And we use it all the time very unselfconsciously. [15:26] Very few people, probably you've used to sort of [15:30] disguise reductio three or four times in the last day. [15:36] You've said something like well if that's a bear then [15:42] bears have antlers. [15:44] That's a reductio. [15:48] You'll find yourself making them. [15:49] Now why do I bother mentioning that since it's a very [15:52] familiar tool? [15:53] Because I want to point out that rhetorical questions, the [15:59] questions you're not supposed to answer. [16:01] Many of them, not all of them, but very many of them are, in [16:04] fact, disguised reductio ad absurdum arguments. [16:08] They imply a reductio ad absurdum argument that the [16:13] writer doesn't bother presenting. [16:16] It's too obvious. [16:16] It would be embarrassing. [16:18] So you just get the rhetorical question and nudge, [16:21] nudge you go on. [16:23] So what I want to do is to create in you a little alarm. [16:29] Sort of like a Google alert. [16:32] So that whenever you see a rhetorical question in an [16:36] argument, ding. [16:38] You think, ah rhetorical question. [16:42] You look to see, maybe I can just answer it. [16:46] Maybe there is a perfectly unembarrassing answer. [16:50] I remember years ago seeing a "Peanuts" cartoon and then the [16:54] next to last frame Charlie Brown is saying, "Who's to say [16:59] what the difference is between right and wrong." Lucy in the [17:02] last panel says, "I will." and think about answering a [17:09] rhetorical question but just answering it. [17:11] See if you can do that. [17:15] A similar thing is what I call the surely alarm. [17:17] Whenever you see the word surely a bell [17:21] should ring, ding. [17:25] Let me just give you one. [17:27] Oh, yes, you could make a computer they could sort fruit [17:31] by its color, but surely no computer, ding. [17:36] Surely no computer could ever have a favorite color or [17:40] appreciate color [17:42] When you hear that surely you are being-- [17:46] now maybe you think, I go along with that one. [17:49] That's got to be right. [17:50] Actually in this audience I bet a lot of people say, I [17:52] don't think that's so obvious. [17:54] [LAUGHTER] [17:59] DANIEL DENNETT: Why surely? [18:01] Well the author wants to get this proposition, the [18:06] proposition that follows it into the discussion. [18:09] Thinks it's true. [18:11] Thinks it's obvious. [18:12] But it's not quite obvious enough so that it [18:14] goes without saying. [18:17] If it did you wouldn't have to say it. [18:19] It would go without saying. [18:21] It goes with saying, but he doesn't bother arguing for it. [18:26] So he gives you just a little nudge and says surely. [18:30] So be on the lookout. [18:31] I tested my hunch about this by going [18:33] through, thanks to Google. [18:36] Thanks to string search. [18:37] Going through some dozens of philosophy of mind papers that [18:41] were online. [18:42] Counting up the surelys. [18:43] Most papers don't have the word surely in them. [18:46] But I found several dozen instances of surely. [18:50] And about 1/3 of them were I thought clearly cases of this [18:55] was the weakest place on the whole argument. [18:57] This is the place you want to look first. [19:00] So now if I've alerted you on rhetorical [19:04] questions and on surely. [19:07] Another one is rathering, rather. [19:10] As in, it's not the case of blah blah blah. [19:14] Rather it's ta da da da. [19:17] Rathering signals a dichotomy. [19:21] And very often it singles a false dichotomy and what you [19:24] should do is, again, have a little [19:25] alarm ring for rathering. [19:27] When you see the word rather take a good hard look and see [19:29] if you can't say, well maybe both blah blah [19:32] blah and da da da. [19:34] That hasn't been asserted. [19:36] It's just been presupposed and glossed over [19:39] with the word rather. [19:41] So if I get you to adopt these habits, look what I've done. [19:46] I've installed an app on your necktop. [19:49] [LAUGHTER] [19:55] DANIEL DENNETT: And I mean that literally. [20:00] The thing between your ears is a kind of computer. [20:04] And I don't have to tell you folks, but the power of a [20:09] computer is the apps and the levels. [20:12] And you can build and build and build and build and build [20:14] apps on top of apps. [20:15] That's what the thinking tools do. [20:18] And that's what makes us smarter. [20:22] A CPU without an operating system, you can't do [20:24] very much with it. [20:26] You start building up all the software and that's where the [20:30] power comes from. [20:36] I don't have to tell you that computers are thinking tools [20:38] par excellence. [20:39] And in the book I put an interlude on [20:41] computers in the book. [20:43] And then Peter, I'm happy to say, he mentioned that I want [20:50] everybody who reads the book to come away with a little bit [20:55] more basic understanding of just where [20:56] computers get their power. [20:59] One of the things I do, and this is so eccentric. [21:02] I think that's probably the only trade book in history to [21:04] have this feature. [21:05] In this chapter called, "The Seven Secrets of Computer [21:08] Power revealed," I teach everybody how to program a [21:15] register machine. [21:17] A Hao Wang simple register machine that can just do [21:20] increment and decrement and branch. [21:22] And they teach it how to do arithmetic. [21:25] And then they teach it how to do some other things. [21:27] And there's problem sets. [21:30] I've actually got some simple problem sets in this book. [21:34] And I say, well you don't have to do them, of course, but if [21:35] you really want to understand. [21:37] If you want to be a computer for a while and understand [21:40] what it is. [21:41] And, of course, they learn about loops. [21:43] They learn about counters. [21:44] They learn about all kinds of things. [21:46] That they hear these words in other contexts and they think [21:49] they know what they mean. [21:50] Now they're going to know what they mean. [21:52] The reason I think that's so important, and again, I know, [21:55] I'm preaching to the choir here, is one of the most [22:00] important epistemological facts about computers is that [22:03] if you can make a computer simulation of [22:05] something that works. [22:06] You know to a moral certainty, there's no magic. [22:09] There's no morphic resonance. [22:10] There's no psionic forces or ectoplasm. [22:15] It's good old-fashioned, every day, [22:18] hardened, honest causation. [22:20] And you just got a proof of it because of [22:22] the computer itself. [22:24] That's what it is. [22:26] There's no place for any mystery forces to play a role. [22:33] But people shouldn't just take that on faith. [22:36] They should know in their bones why that's true [22:39] and how it's true. [22:40] And that's what that section is supposed to show them. [22:45] Some intuition pumps don't even have words, [22:48] they're just images. [22:49] I love this one. [22:50] This is from my friend Matt Ridley, wonderful evolutionary [22:54] biologist and journalist. [22:56] And it just startles for a moment. [23:00] But what points can you make with it? [23:01] I'll just make one then move on. [23:04] On the left you see an Acheulean hand-axe. [23:07] On the right, I don't have to tell you what that is. [23:10] The Acheulean hand-axe was made by our ancestors without [23:14] any discernible stylistic change. [23:17] Without any discernible design improvement for [23:20] over a million years. [23:23] Absolute stasis. [23:25] Amazing. [23:28] Some people have even thought maybe it wasn't to be used axe [23:31] at all, maybe it was like the peacock's tail. [23:34] Maybe it was, hey, you want to see my hand-axe collection? [23:41] I am so good that I can make hand-axes all day long and [23:47] still get enough to eat, take care of myself. [23:51] Bragging. [23:53] Sort of like having a Lamborghini or [23:55] something like that. [23:57] But notice at the one on the right, familiar as is it is to [24:01] us, it probably has a rather short lifetime ahead of it. [24:07] How many bet that there will still be mice [24:10] in a hundred years? [24:13] Right, point made. [24:17] Now here's a similar comparison. [24:22] On the left we have a termite castle. [24:27] Australian termite castle. [24:30] And on the right we have Sagrada Familia, the great [24:33] Gaudi church in Barcelona. [24:36] As I think you can see, they're stunningly similar [24:39] structures. [24:40] And even internally, if you look at the cutaways, there's [24:44] a lot of design features that they have in common. [24:47] These are two remarkable artifacts made by animals. [24:53] And yet, for all their similarities, you can see at a [24:56] glance that they are the products of profoundly [25:00] different histories of design and construction. [25:05] The one on the left is, this is bottom up Darwinian design [25:09] and construction. [25:10] The termite, the individual termites, could [25:12] hardly be more clueless. [25:14] They don't know what they're doing. [25:16] There's no boss termite. [25:17] There's no architect. [25:19] There's no blueprint. [25:21] There's no leadership roles. [25:25] There's no hierarchy. [25:26] This is bottom up activity. [25:29] And it produces that. [25:30] Whereas Gaudi, it's almost comical. [25:33] He was like a cartoon caricature of [25:36] the inspired genius. [25:38] Autocratic, dictatorial, laying down the law. [25:42] He had blueprints and manifestos. [25:44] He had accounts of everything. [25:46] And he bossed people around. [25:47] And he told the bosses how to boss the people. [25:51] Classic case of hierarchical design and construction. [25:55] Here is one way of looking at the problem, I think, that [25:59] evolution and cultural evolution has to explain. [26:03] How did our planet get from the left to the right? [26:06] How did we get from termite type R&D to Gaudi-type R&D? [26:15] I think the answer has to be, well, Gaudi did not have a [26:19] bare brain. [26:20] He had all those thinking tools. [26:22] And those thinking tools are the key to opening up top [26:26] down, as it were, intelligent design type design. [26:31] What Darwin showed us is we don't need the intelligent [26:34] designer that lots of bottom up, mindless grubbing around, [26:38] trial and error, will eventually produce all the [26:40] wonders of the biosphere. [26:42] But then the process keeps right on going [26:44] with cultural evolution. [26:45] And pretty soon it starts to reverse itself. [26:48] And we get more and more intelligent design. [26:51] To where we now have people like touring, better example [26:55] than Gaudi, because he really had the goods. [26:57] Who notice, when he built his computer, he already had proof [27:02] of concept. [27:03] He had the blueprint. [27:04] He had the design. [27:05] He proved it all. [27:05] Nobody would have ever given him the money to make that [27:07] computer if he hadn't proved it in advance. [27:09] This is where the idea really is leading the way. [27:13] That's intelligent design. [27:15] And we've got to explain how it has happened to be one of [27:18] the fruits of evolution. [27:21] Because it wasn't the cause. [27:25] The Bible says in the beginning was the Word. [27:27] No, no, the word, very recent development. [27:33] The word is the beginning of intelligent design. [27:37] The Bible is right about that. [27:43] Another thinking tool which is just a picture, but what a [27:45] lovely picture. [27:47] And, by the way, this is Leonard Eisenberg's version of [27:49] the Tree of Life. [27:50] And you can get it on t-shirts, big shiny posters [27:53] that you put up in your office. [27:55] There's all sorts of ways you can get it if [27:56] you go to his website. [27:58] And what I love about it, there's many different ways of [28:01] representing the tree of life. [28:02] This has a few nice features. [28:03] I just want to sort of point them out to you. [28:06] I ran out of battery at some point. [28:08] OK. [28:09] There's no other pointer up here, is there? [28:12] I'll just do it like this. [28:16] Here's a pointer. [28:18] A thinking tool! [28:19] Right here on the end of my finger. [28:21] How wonderful. [28:24] So time goes out radially. [28:28] Here's the birth of the earth. [28:29] Here's the birth of life, the origin of life. [28:31] And as you can see, the bacteria and the archea are [28:35] the first single-celled organisms. [28:38] And then we have a period of single-celled eukaryotes. [28:40] This is a great, great, great moment in the history of life. [28:44] It's the eukaryotic revolution. [28:46] When the endosymbiotic event put two of these types [28:51] together they joined forces and became a [28:53] more powerful thing. [28:55] This is technology transfer. [28:57] This is what happens when you have two independent lineages. [29:00] They've been evolving independently [29:02] for a billion years. [29:04] And now they join forces. [29:05] They don't all have to reinvent the wheel. [29:07] They get the benefit of each current, and [29:10] those eukaryotic cells. [29:13] All the rest of life is eukaryotes. [29:18] You've heard of the Cambrion explosion. [29:20] There it is. [29:21] In a very short period of time tremendous flourishing of [29:25] different kinds of life. [29:27] A period of fantastic fecund R&D. [29:34] How was it explained? [29:36] That's a story I'd love to tell, but for another day. [29:38] You can see mass extinctions. [29:41] And then way over here, this last little fork here, that's [29:47] about 7 million years on that scale. [29:49] Which is about the length of time since we shared an [29:52] ancestor with our nearest relative, the chimpanzee. [29:56] And, of course, of that seven million years, only a million [30:01] years at most has any language. [30:06] Some people will push it back a little further. [30:09] The Acheulean hand-axe comes before language. [30:15] Some people think that's a controversial issue. [30:17] But so now notice, Cambrion explosion. [30:21] Fabulous. [30:23] The MacCready explosion happens there. [30:30] And transforms the planet. [30:32] All thanks to thinking tools. [30:40] So now, a few recent intuition pumps that are in the book. [30:47] Because I haven't really talked about them. [30:49] These are these little stories. [30:50] So here's an intuition pump I recently devised to discomfort [30:54] neuroscientists and cognitive scientists who have been going [30:57] around saying I think ill-considered things about [31:00] what neuroscience teaches us about free will. [31:04] So first a little science fact. [31:07] Damiaan Denys, a very brilliant researcher in [31:10] Amsterdam, has made a little microchip which controls [31:13] obsessive-compulsive disorder. [31:15] You implant it in a person's brain. [31:16] Really does a very good job. [31:20] Science fiction. [31:22] So one day this guy with OCD goes to see his neurosurgeon. [31:25] And she implants one of in these Denys [31:28] microchips in his brain. [31:30] Sews him up. [31:32] The anaesthesia wears off. [31:34] And as he's about to leave she debriefs him a little bit. [31:37] She says, oh and by the way, my staff will be monitoring [31:40] you 24 hours a day, seven days a week from our [31:44] control room here. [31:45] And you're going to think you have free will. [31:46] You're going to think you're making your own decisions. [31:47] But, in fact, we're going to be making all the [31:49] decisions for you. [31:51] Free will is just an illusion for you. [31:53] Have a nice life, goodbye. [31:57] And he believes her. [31:59] Well, shiny lab, lab coat, all the rest. [32:03] And as you might imagine he becomes a little negligent in [32:06] his habits. [32:07] And a little bit aggressive. [32:09] And indulges in some of his worst behaviors. [32:14] And soon enough he gets in trouble with the law. [32:17] Gets hauled into court, charged with a [32:20] fairly serious crime. [32:21] He says, but Your Honor. [32:23] I don't have free will. [32:25] How do know that? [32:26] Well my neurosurgeon told me, free will is just an illusion. [32:30] It feels like I'm making these decisions, but I'm not really. [32:33] So they call the neurosurgeon to the stand [32:36] and ask her to testify. [32:37] And under oath she says, Yeah, that's what I told him. [32:40] But it was just a joke. [32:42] I was just messing with his head. [32:46] Now how many of you agree that she did, not just a mean [32:50] thing, a socially irresponsible thing. [32:52] But she did very serious harm to this guy. [32:57] With her language, she did nonsurgically she claimed to [33:02] do surgically. [33:03] She disabled him as a moral agent. [33:07] By just convincing him that he didn't have free will. [33:11] Now I ask my neuroscientist friends, aren't you just doing [33:17] the same thing wholesale? [33:19] When you write these articles saying, we have learned in [33:21] neuroscience that we don't have free will. [33:24] The reason they say that turns out to be very bad reasons. [33:29] Having to do with, well everything in your brain is [33:32] determined by what happens in your brain [33:34] a little bit before. [33:34] Yeah, of course. [33:35] So what? [33:36] That doesn't show you don't have free will. [33:38] They have a very simplistic idea of what free will is. [33:41] But it's the idea that would actually ground moral [33:45] responsibility. [33:47] If you didn't have it you wouldn't have moral [33:49] responsibility. [33:50] But they're not showing that at all. [33:52] And I think that's irresponsible. [33:54] Now you might say, well, if your brain is determined, and [34:00] then you don't have free will. [34:02] How many of you are inclined to think, if every event in [34:05] your brain is determined then you don't have free will? [34:09] The general run of people think that's sort of, even [34:12] obviously true. [34:13] So here's a little intuition pump to [34:15] dissuade you from that. [34:16] Two lotteries. [34:18] In lottery A the winning ticket is chosen after the [34:23] tickets are sold. [34:24] Most lotteries are like that. [34:27] In Lottery B the winning ticket is chosen before the [34:30] tickets are sold. [34:31] And the stub is secretly put in a safe for safekeeping [34:35] until the tickets are sold. [34:37] Now how many of you think that you're a fool to buy a ticket [34:42] in Lottery B? [34:43] That only lottery A gives you any chance of winning? [34:48] How many think that they're even? [34:53] Yeah. [34:54] And so does most of the rest of the world. [34:57] When you put in these terms. [34:59] They recognize that, no, you've got a [35:00] chance either way. [35:02] And Publishers Clearing House actually says, you may [35:05] already have won. [35:09] People see that they still have a chance if the winning [35:14] tickets are chosen in advance. [35:17] But now think, if determinism is true then all your lottery [35:20] tickets were chosen before you were born. [35:24] At the Big Bang, if you like. [35:26] And sort of put in on a little secret safe in your body. [35:30] And as life goes on, whenever you need a coin flip, whenever [35:35] you need a little chaos, you just take the next one out of [35:38] the envelope and you win some you lose some. [35:42] That's just as free as if there's a quantum miracle that [35:45] happens right then and there. [35:48] So what if they're chosen first. [35:52] Well somebody might say, well you see, if all your lottery [35:56] tickets are chosen before you're born, you're determined [35:59] to have good luck on some occasions [36:01] and bad luck on others. [36:02] You're determined. [36:03] Well that's true even if indeterminism is true. [36:09] Even if indeterminism is true, you're determined on some [36:12] occasions to win and some occasions to lose, you just [36:14] don't know which. [36:17] Fairness isn't always winning. [36:20] You've got to accept the fact that there's luck. [36:23] And that luck is consistent with free will. [36:26] But a lot of philosophers haven't seen it. [36:28] A very good one, David Wiggins, once wrote some years [36:30] ago about what he called the cosmic unfairness of [36:35] determinism. [36:36] And you can see, or many people think, yeah right, it's [36:40] a bummer if determinism is true. [36:45] But I want to ask what about the cosmic unfairness of [36:48] indeterminism? [36:51] Just as bad. [36:53] You know. [36:56] Life isn't always a picnic. [36:58] That doesn't mean that we can't be responsible agents. [37:02] Being a responsible agent, having free will in the sense [37:04] that matters, has zip to do with whether physics is [37:09] deterministic or indeterministic. [37:11] Or whether your brain is deterministic or [37:13] indeterministic. [37:14] Has to do with the software you've got in your brain. [37:19] It has to do with the competence that you have had [37:21] installed in your brain since you were a little kid. [37:25] If you were trained with a good, moral upbringing then [37:28] you make the right decisions at the right [37:29] time most the time. [37:31] And you are, in general, competent enough so when you [37:34] screw up, you are held responsible. [37:37] Why? [37:38] Because you're good enough to be responsible. [37:42] That's a long story. [37:42] I'm not going to say more about that. [37:45] Time to quit this talk. [37:47] So the book is sort of a tapas book. [37:50] There are 77 chapters. [37:53] And some of them are just a page. [37:55] Some of them there are maybe as much as 10 pages long. [37:58] And they each introduce a tool making it autonomous. [38:02] I tried to make them all portable and accessible. [38:05] So my recommendation is to try them out. [38:08] Take them for a spin. [38:10] And above all, turn all the knobs to see how they work. [38:14] Thanks for coming. [38:15] [APPLAUSE] [38:20] DANIEL DENNETT: Questions? [38:23] AUDIENCE: Hello. [38:26] So intelligent design is better in some [38:29] domains than others. [38:31] We can design a very nice mobile phone or computer. [38:36] When it comes to an economy, well, some may argue, but so [38:41] far we've yet to show we can [38:43] intelligently design an economy. [38:46] Do you have, well two questions. [38:49] First of all, what you think the profit is of things which [38:53] are not very tractable to intelligent design? [38:58] And do you think eventually all problems we'll have the [39:02] right tools in order to design everything? [39:07] DANIEL DENNETT: No, I think there's a trade-off which [39:10] other people have described pretty well. [39:14] Natural selection is a great engine. [39:19] And it feeds on chaos and so-called randomness. [39:25] But it works, remember, by trial and error. [39:26] And there's an awful lot of errors. [39:28] And it's a very wasteful product. [39:30] Very wasteful way of doing business. [39:33] You've got to have a lot of failures. [39:38] And so that's what happens, let's say, with an economy. [39:43] And to that extent, that grubby old social Darwinist [39:48] line about how the survival of the fittest in the economy has [39:53] some truth to it. [39:55] But that doesn't mean that we should simply honor that and [39:59] make that a moral principle. [40:02] Because I think we have very good reasons to think that an [40:05] unfettered, completely chaotic, laissez-faire, [40:09] dog-eat-dog economy is not a good way to go. [40:14] Too much harm can accrue. [40:17] You can get these winner take all situations [40:20] which can be very bad. [40:22] You know about winner take all algorithms. [40:24] You know about opponent processes in the brain. [40:28] In the work that I've been doing on how brains are [40:31] organized, I think, first of all, they are computers. [40:37] But they're very, very different from laptops. [40:41] Precisely because you've got 100 million [40:47] or 200 million neurons. [40:50] Each one is like a little termite. [40:51] They actually have their own agendas. [40:53] There's no two alike. [40:56] And the way software works in the brain is by creating [41:00] coalitions and getting them to work together. [41:03] And unity is an achievement, it's possible. [41:07] But it isn't built in from the top down. [41:12] And once we think about the brain being organized that [41:14] way, we should think that's probably economies are best [41:17] built in that way too. [41:18] Without too much top down central control, but with some [41:26] policing to protect us from the excesses that otherwise [41:30] would accrue. [41:32] AUDIENCE: Could you give your opinion on Newcomb's paradox? [41:36] DANIEL DENNETT: Yeah. [41:37] Do you know what Newcomb's paradox is? [41:39] This is the one boxer or two boxer. [41:42] OK, so you've got. [41:45] First I'm going to have to explain it. [41:47] See if I can get it right. [41:48] So you're on a sort of quiz show and you're given the [41:51] following weird choice. [41:53] You could be a one boxer, or you can be a two boxer. [41:57] In one box there's a $1,000. [41:58] In the other there's $1 million versus $0. [42:04] It would be hard enough if you had to choose either box A or [42:06] box B. Would you take a chance on $1 million, or would you go [42:12] for the sure $1,000? [42:16] But that's not your choice. [42:18] Your choice is either to go for a box of B which has [42:20] either got the $1 million or nothing. [42:23] Or to go from boxes. [42:25] Got it? [42:27] Well one line of research, one line of thinking, says well, [42:31] dummy, of course you go for both boxes. [42:34] That way at least you get $1,000 and maybe get a $1 [42:38] million and $1,000. [42:39] That dominates decision tree, draw it out. [42:42] That's the one you should do. [42:45] And so it would be, except for one thing. [42:49] The MC of the program tells you, oh, by the way, you're [42:54] the 5,000th contestant on the show. [42:59] And we're very good at predicting whether you're [43:03] going to be a one boxer or a two boxer. [43:06] And if we predict that you're going to make the two boxer [43:10] choice, then there's only the $1,000 in box A. We never put [43:17] the $1 million. [43:18] We only put the $1 million in when we predict that you're [43:21] going to go for one box. [43:24] And do you want to know our track record? [43:26] It's almost perfect. [43:29] They might add, well you know, one poor devil went for box A [43:35] and it was empty, or box B and it was empty. [43:39] And one lucky devil out of 500 or whatever, went for both and [43:43] got both rewards. [43:45] But in general it's a very strong statistical [43:49] correlation. [43:50] So now I ask you. [43:53] How many of you go for both boxes? [43:58] And how many of you go for just the one box? [44:02] What's nice about Newcomb's paradox is that it ain't [44:05] obvious at all. [44:07] And an awful lot of ink has been spilled trying to get [44:10] clear about this. [44:12] And I have, myself, not spilled any ink because I [44:15] still don't have a firm view on it. [44:23] I tend to think there must be something a little bit screwy [44:28] about the setting up of this thought experiment. [44:32] But it's very hard to show that. [44:37] And I highly recommend it to you. [44:40] There's a lot of good things have been written about it. [44:44] So I'm sorry I can't answer the question. [44:47] I go back and forth. [44:48] Some days I'm a one boxer, some days I'm a two boxer. [44:51] AUDIENCE: I wanted to ask you about [44:53] determinism versus free will. [44:56] Since you were talking about earlier. [44:57] And, in particular, I think Hofstadter would talk about a [45:00] lover's disparity there. [45:02] That determinism is kind of a totally different question [45:05] then free will. [45:06] I wondered if you could talk about that. [45:08] DANIEL DENNETT: Yeah Doug and I have talked about [45:09] this over the years. [45:10] And Doug disagrees with me. [45:12] It's one of the few points where he and I have a serious [45:15] disagreement. [45:15] And he's actually quite pessimistic about this and [45:18] thinks that the sense of free will that really matters is [45:22] one that we can't have because of determinism. [45:25] And I've done my level best in several books and many hours [45:29] to convince him otherwise without success yet. [45:35] Unlike the fellow in my tale, it doesn't lead him to behave [45:41] in a irresponsible way. [45:45] So there's at least to could be said for it. [45:52] I think that Doug-- [45:54] here's what I think the problem is. [45:57] That the term free will. [45:58] Where does it come from? [46:00] It comes from every day. [46:02] It comes from the law and from tradition. [46:04] It's not a scientific term. [46:06] It's a term you get from what's called the manifest [46:09] image, in philosophy. [46:11] The world of colors and dollars and [46:14] baseball and so forth. [46:16] Not the world of neurons and proteins and atoms, or for [46:23] that matter, scientific laws. [46:26] When you look at it's home territory, free will, you can [46:33] see it does two jobs. [46:36] And they both seem to be definitive or constitutive. [46:39] On the one hand, it seems definitive that well free will [46:43] means uncaused, it means it's free. [46:45] There's no sufficient causal antecedent. [46:50] For many people I think that's just obvious that's what free [46:53] will means. [46:55] But the other thing it means is it's the standard condition [47:00] for being morally competent. [47:03] For instance, how many of you believe yourself qualified, [47:08] eligible, to buy a car to take out a loan? [47:13] Yeah. [47:14] And you sign the paperwork of your own free will. [47:17] Now there are people who don't have that competence. [47:20] They're senile or they're too young. [47:23] There's something you've got that they don't have. [47:25] And it is your very condition of your living a free life as [47:29] a citizen and conducting business, making promises that [47:32] count and so forth. [47:34] I think that's the best notion of free will. [47:36] That's the one that matters. [47:37] And I think that notion has nothing to do with [47:40] determinism. [47:44] For instance, I don't think your competence to sign a [47:48] contract would be disparaged in the slightest if [47:54] determinism were true. [47:57] So I have argued for years there's just no reason to [48:01] prefer that your brain be deterministic over [48:05] indeterministic. [48:06] No reason at all. [48:07] No good reason. [48:08] I've recently had to back off that a little bit by a very [48:14] persistent lover of indeterminism and free will, [48:19] who finally showed me that there was, by my own lights, a [48:22] condition where I would agree with them. [48:24] I would wish for indeterminism in that case. [48:27] And I'll tell you what it is. [48:29] You all know the game rock-paper-scissors? [48:32] Well the best way to play that game, of [48:33] course, is to play random. [48:35] It's very hard to fake randomness or to put together [48:39] a random series. [48:40] If you're random you give you give your opponent no pattern [48:44] to track at all. [48:45] So they can't do any better than play random against you. [48:48] You'll break even. [48:55] So since that's the case, you might just as well memorize a [49:00] few hundred entries in a table of random numbers and use that [49:07] to guide your choices. [49:10] You'll do fine as long as you don't let the other [49:11] guy see your list. [49:13] Which would be terrible. [49:15] So now I can tell you, here's the condition under which I [49:18] would hope that indeterminism is true. [49:20] If I were playing rock-paper-scissors for high [49:23] stakes with God. [49:28] I don't really anticipate that happening in the future, so I [49:33] am quite content to live out my life in a deterministic [49:36] world and I'll take responsibility for my acts, [49:39] even though they are, for all I know or care, determined by [49:43] the kazillion things that happened both before [49:47] and after my birth. [49:50] Yes. [49:51] AUDIENCE: Is consciousness an epiphenomenon of advanced [49:55] biological systems and tends to disappear as you look back [49:58] across the Tree of Life? [49:59] Or is it a fundamental dimension to nature and [50:01] reality like the dimensions of space and time? [50:05] DANIEL DENNETT: OK you said or. [50:07] This is sort of a rathering. [50:10] You didn't say rather, but there was a sort [50:12] of rathering there. [50:15] But I'll take it on because I do not think that it is a [50:21] fundamental category that splits the universe in two. [50:26] I think that the idea that it is deeply rooted in culture. [50:32] And I suppose it might be right. [50:33] But I think that consciousness is like reproduction, [50:38] metabolism, and self repair, healing, blood clotting. [50:42] It's a very interestingly complex and deeply recursive [50:48] biological phenomenon. [50:50] It is not in the interesting sense, and there's many senses [50:54] of consciousness. [50:56] In the sense that we in this room are conscious, which is [50:59] the anchoring that's home base from our own experience and [51:06] our reading of literature. [51:07] We read novels and things like that. [51:09] We have the sense of what a conscious life is like. [51:13] I think we're the only creatures that have that kind [51:16] of conscious life. [51:16] I don't think dogs have it. [51:17] I don't think dolphins have it. [51:19] They're sentient. [51:20] They're awake. [51:21] They feel pain. [51:22] But they don't have most of what makes human consciousness [51:27] so important. [51:30] And I think that phenomena is, in the end, software. [51:40] And I think it has to be installed. [51:42] So it is an implication, in my view, and [51:46] don't try this at home. [51:47] Don't test this. [51:49] That if a human baby could be raised entirely without any [51:53] interaction with other human beings and [51:55] cut off from language. [51:56] If it could grow to adulthood without benefit of all the [52:02] thinking tools that would normally be installed in the [52:06] course of just growing up in a social community. [52:09] I think that human being would have a mind so profoundly [52:15] different from ours that we would all pretty well agree he [52:19] or she wasn't really conscious. [52:22] Now that may be a very eccentric notion of [52:26] consciousness. [52:26] But if you're wanting to know, do I think that, if you look [52:31] for a better notion of consciousness or one that [52:33] isn't so eccentric. [52:35] And you start thinking about what it is. [52:37] The capacity to experience color and flavor and to have [52:43] qualia, as philosophers say. [52:45] I think that's all pretty badly confused. [52:47] And I've tried to say how and why over the years. [52:52] But part of what makes it confusing is it people want [52:57] consciousness to be, to use a technical term, [53:02] supercalifragili sticexpialidocious. [53:06] And if you're not talking about something that is [53:08] supercalifragili sticexpialidocious then you're [53:11] not talking about consciousness. [53:14] And there's a quote my friend Lee Siegel, he's a philsopher [53:21] and a magician. [53:22] A very fine stage magician. [53:24] Wrote a wonderful book on Indian street magic called, [53:27] "Net of Magic." [53:29] And there's a passage in there I love to quote. [53:30] He says, I tell my friends I'm writing a book on magic and [53:34] they ask, do you mean real magic? [53:38] And by real magic they mean magical. [53:46] They mean thaumaturgical acts and so forth. [53:52] And he says, no. [53:53] Not real magic. [53:54] Sleight of hand. [53:56] Stage magic. [53:57] And then he goes on, he says, in other words, the kind of [54:02] magic you can actually do is not real magic. [54:10] Real magic is a kind that doesn't exist. [54:13] Well what I've come to realize, my whole career, I've [54:18] been confronted with people. [54:19] Because what I want to deal with are topics that a lot of [54:21] people think of as real magic. [54:23] Free will and consciousness in particular. [54:26] And if you don't mean something that as one [54:32] proponent of it says, every free [54:36] decision is a little miracle. [54:39] If you don't think that that's what free will is then you're [54:41] not talking about free will. [54:43] Not real free will. [54:45] You're talking about some cheesy substitute. [54:48] And I'm all for this cheesy substitutes, I'm afraid. [54:51] I think they're quite good enough. [54:53] And science can explain why we have varieties of free will [54:56] that are worth wanting, even if they aren't real free will. [55:02] They'll secure our lives as meaningful and [55:07] responsible and so forth. [55:09] And we can have consciousness even if it isn't something [55:13] that sunders the universe in twain in this [55:17] very special way. [55:19] Sorry it's such a long answer, but you got me going. [55:22] AUDIENCE: What do you think about [55:23] utilitarianism and moral reasoning? [55:26] DANIEL DENNETT: I'm not a utilitarian, or at least not [55:29] an ordinary utilitarian. [55:31] I don't that any. [55:32] I'm a consequentialist. [55:35] That is to say, I think the consequences matter. [55:38] I'm not one who takes a strictly [55:44] deontological view of ethics. [55:48] Using some technical terms. [55:49] I want to try to get away from them. [55:52] This is an issue that's very close to the service in work [55:55] that I'm doing right now on free will and punishment. [56:00] Because I say, no, we don't. [56:02] I notoriously said in my last book on free will, I don't [56:06] want to live, no in an article on free will. [56:08] I don't want to live in a world without punishment. [56:12] And these people jumped all over me and said oh, you dirty [56:15] retributivist, you. [56:18] I said, no, I'm not a retributivist. [56:22] Kant was the great retributivist. [56:25] And if you want to know what bullet he bit, he said, if we [56:31] know the world was going to end tomorrow, one of our [56:36] highest duties would be to go into the prisons and execute [56:39] the people on death row. [56:46] By that he wanted he wanted us to think that punishment was [56:49] in itself good. [56:50] That a crime unpunished is a an offense to [56:55] nature or God or something. [56:57] And that's what a retributivist is one who [57:00] thinks that punishment is a good in itself. [57:04] I don't. [57:06] But I think it's a very foundational good for a very [57:09] simple reason. [57:12] I want to live in a world in which there are promises. [57:17] And you can't really have promises unless there's [57:19] consequences for not keeping your promise. [57:22] And those consequences are not rehabilitation. [57:26] They're not reform. [57:27] They're punishment. [57:30] And it's OK. [57:31] And we should all be willing to say, no, I'd rather take my [57:35] punishment. [57:36] I'd rather be a free citizen in a free society. [57:39] I accept that if I do something really wrong I am [57:42] eligible for punishment and that's OK. [57:45] I think the world we live in, when we have a system of law [57:49] and order with punishment, I think our current system of [57:54] punishment is obscene and should be completely reformed. [57:57] But I don't think we should eliminate punishment as some [58:00] of the current crop of cognitive scientists and [58:04] neuroscientists are saying. [58:07] I don't think they've look closely at the last time this [58:10] was considered a really wise move. [58:12] And that was back in the Stalinist era when people [58:18] weren't punished, they were treated. [58:22] I don't think we want to be treated. [58:24] I think we want to be punished. [58:28] Because then there's rules. [58:30] And you can serve your time and get out and [58:33] be a citizen again. [58:35] An that dire note, I'll stop. [58:38] AUDIENCE: There was a decision this week about patents that [58:41] touched on some deeply philosophical issues. [58:44] The question was whether software should be patentable. [58:47] And one of the opinions, which unfortunately, didn't gather [58:50] majority support was that basically things you do in [58:55] software are like natural laws. [58:57] They're tools that everybody can use. [58:59] And some particular arrangements with these tools [59:01] should not be patentable. [59:03] Other judges felt that if you have a computer and it's [59:05] programmed with software, that's a new device. [59:08] And devices are clearly patentable. [59:11] Do you have some wisdom to offer us on that? [59:14] DANIEL DENNETT: Well I thought about it in the past. [59:18] And I've thought about the difference between patent and [59:20] copyright and trade secret. [59:22] And actually I'm trying to think about this in quite [59:29] fundamental terms. [59:30] Because I think that what we need to develop is the sense [59:35] of information, which is distinct from Shannon-Weaver [59:38] information. [59:39] Which depends on Shannon-Weaver information, [59:41] but is semantic information. [59:43] It's cognitive information. [59:45] And, to me, the best way of thinking about this is that [59:50] it's useful information. [59:54] It is information defined as that which permits the [60:00] transmission of adaptations for one purpose, one utility [60:04] or another. [60:05] And then under that view then, of course, software, I think [60:10] intuitively when we're measuring whether software [60:14] what counts as theft, we have to look at the utility. [60:22] And if it's the utility that's been stolen, then we should [60:29] have penalties for that. [60:32] That should be discouraged. [60:35] The trouble is that copyright law also involves this idea. [60:45] It's a good question. [60:46] And I'm not surprised that the Justice-- [60:49] Is this at the Supreme Court? [60:52] AUDIENCE: It's the Court of Appeals for the Federal [60:54] Circuit that handles patent cases. [60:56] DANIEL DENNETT: I want to know more about it. [60:57] I'm not surprised that they are divided. [60:59] I think it's a tough issue. [61:01] But obviously when I talk about apps, as tools for [61:06] thinking, I am thinking of these as mechanisms that have [61:10] been designed. [61:10] And so on that ground I would tend to come down on the side [61:15] that you can patent it. [61:17] Thank you. [61:17] [APPLAUSE] [61:20] [MUSIC PLAYING]