Mazzucato Urges Nations to Rebuild State Capacity, Cut Consultants

马祖卡托呼吁各国重建国家能力，削减外部咨询 将核心政策决策外包给私营咨询公司会削弱国家能力，并阻碍政府引导技术创新服务于公众需求。 在本期内容中，Mariana Mazzucato认为，政府必须构建内部动态能力，并采用“使命型经济”方法来应对重大全球挑战，而不是将核心公共职能外包给私营咨询公司。她阐释了受NASA阿波罗计划启发的成果导向型采购如何推动跨部门创新，同时对人工智能领域的市场榨取以及公共科研人才流向大型科技公司的现象发出警告。此外，Mazzucato主张在公私伙伴关系中纳入预分配条件，重塑公共政策的尊严，并借鉴使命驱动型治理的国际模式。 西班牙官员基于Mariana Mazzucato的研究编写了一份题为《西班牙创业型国家》（El Estado Emprendedor en España）的报告，指出风险投资家将失败视作常态，而公务员因单次失误就会面临严重的政治后果，两者形成鲜明对比。 (5:00) 为防止产业政策沦为企业补贴，政府应放弃挑选特定行业或技术，转而通过与致力于宏伟社会目标的私营部门合作来“挑选有意愿者”。 (8:12) 政府效能可划分为基础能力（预算和人员编制）、行政常规（稳定的学习环境）以及动态能力（塑造市场而不仅是修补市场失灵的敏捷性）。 (8:12) 过度依赖私营管理咨询公司会侵蚀国家内部能力，并造成结构性利益冲突，因为咨询公司会从政府对其外部建议的持续依赖中获利。 (15:00) 借鉴NASA的阿波罗计划，Mazzucato主张采用成果导向型采购取代传统的成本加成合同，以在保持公共导向的同时推动跨部门的私营创新。 (20:00)

Economist Mariana Mazzucato argues that governments must rebuild their internal capabilities to steer industrial strategy toward ambitious, problem-oriented 'moonshots' instead of outsourcing core public functions to consultancies and subsidizing private corporations without public-interest conditions. Shifting Industrial Strategy Toward Problem-Oriented Moonshots (8:12) Massive public spending packages like the US Inflation Reduction Act and the European Union's €2 trillion NextGenerationEU fund risk socializing investment risks while privatizing profits if they simply dole out subsidies. Mazzucato argues that states should avoid trying to 'pick winners' or favor specific business sectors; instead, policy should 'pick the willing' by directing contracts toward private partners committed to solving explicit societal challenges. This mission-oriented model draws precedent from NASA's Apollo program, where procurement chief Ernest Brackett replaced basic cost-plus contracts with outcomes-oriented procurement, steering roughly 400,000 private workers toward explicit goals and yielding technological spillovers like camera phones and home insulation. Contemporary applications include Germany's state bank KfW conditioning steel industry loans on material decarbonization, and Sweden's innovation agency Vinnova reshaping municipal supply chains around healthy, sustainable school lunches. Dismantling the Public Sector Consulting Trap (15:00) Decades of ideological downsizing since the 1980s have eroded internal state capacity, creating an over-reliance on external management consultancies. Mazzucato contends this reliance produces an institutional conflict of interest: commercial advisory firms have no economic incentive to build self-sufficient public institutions because doing so eliminates their billable revenue. Furthermore, consultancies frequently represent conflicting interests, such as advising regulatory agencies while simultaneously consulting for the private entities under their oversight. Outsourcing core administrative functions also serves as a political mechanism to diffuse blame, leaving civil services susceptible to what NASA procurement officials historically called 'brochuremanship.' To counteract this drain, governments require dedicated internal innovation hubs and public capabilities. The creation of the UK's Government Digital Service—which brought digital design in-house to build Gov.uk after private contractor Serco failed—demonstrates how restoring internal technical expertise can improve public services and attract top engineering talent back to the state. Recapturing Public Value and Governing AI (30:00) Foundational digital breakthroughs, including early AI algorithms, GPS, and the internet, were underwritten by public agencies such as DARPA and the National Science Foundation. However, private technology monopolies now capture disproportionate economic rents from these public foundations, precipitating a talent drain where well-funded tech firms hire researchers directly out of universities and state laboratories, undermining the public sector's technical literacy. To correct this imbalance, Mazzucato advocates for 'pre-distributive' conditions on public funding, citing Oxford University requiring AstraZeneca to cap vaccine prices and enter a patent pool during COVID-19. In artificial intelligence, she warns that absent public direction and disclosure standards, the technology is largely being deployed to generate friction within broken frameworks—such as healthcare insurance algorithms disputing claims with hospital billing bots—rather than confronting critical global challenges like ecological collapse or water-cycle disruptions. Insulating State Ambition and Restoring Dignity To shield long-term public missions from short-term electoral cycles, governments can emulate mission-driven agencies like the US agency DARPA, Finland's Sitra, and Sweden's Vinnova. These institutions deploy fixed-term secondments where civil servants are evaluated on systemic impact and calculated risk-taking rather than political survival. Mazzucato also emphasizes that economic policy must actively confer dignity to effectively counter populist political backlashes. For example, in the London borough of Camden, restructuring adult social care procurement by bringing both caregivers and recipients directly into the policy design process restored civic agency far more effectively than top-down, passive administrative programs. While mission-driven public institutions have successfully directed technical revolutions in the past, an unresolved challenge remains whether governments possess the political leverage and technical capacity to govern artificial intelligence—a general-purpose technology whose capital concentration, computational demands, and systemic risks are currently dominated by private firms.

经济学家 Mariana Mazzucato 指出，政府不能仅充当被动的“市场修复者”或将战略责任外包给咨询公司，而应重塑内部动态能力，通过设定明确公共目标的“登月式”产业政策引导私营部门创新，并在合作中设置严格的前期约束条件以捍卫公共利益。 重塑国家能力与使命导向型产业政策 (8:12) Mazzucato 认为，单纯扩大财政支出（如美国的《通胀削减法案》或欧盟的下一代基金）并不等同于战略治理。政府如果仅提供无条件的税收激励或补贴特定行业，往往会导致“风险社会化、收益私有化”。她主张政府应停止在微观层面“挑选赢家”，转而“挑选有意愿者”（picking the willing），即与愿意共同解决重大社会挑战的企业结成伙伴关系。在政策评估上，静态的净现值（NPV）或成本效益分析会扼杀长期雄心；如果 NASA 在 1960 年代采用常规商业收益测算，人类绝不可能实现登月。 在阿波罗计划中，NASA 采购主管 Ernest Brackett 摒弃了仅关注压减短期支出的“成本加成”采购合同，转而推行面向明确技术挑战的“以结果为导向”采购，不仅有效协调了约 40 万名私营部门雇员，还催生出相机传感器、高级软件和隔热材料等大量民用技术外溢。Mazzucato 指出，真正的国家能力不仅包括预算规模和编制等基础资源，更依赖于稳定的行政流程，以及敢于承担试验风险、主动塑造市场而非事后“打补丁”的内部动态能力。 咨询依赖与公共部门能力空心化 (15:00) Mazzucato 在其著作《The Big Con》中指出，自 20 世纪 80 年代里根和撒切尔推行意识形态化的政府精简以来，核心公共职能逐渐被外包给私营管理咨询公司。这种咨询模式存在根本性的利益冲突：咨询公司若帮助政府建立自给自足的解决能力，就会终结自身的业务来源。与此同时，咨询公司普遍存在跨利益主体的双重业务，例如在南非同时为国家电力公司 ESKOM 及其监管方财政部提供建议，或在澳大利亚普华永道同时服务医疗器械企业与监管机构。 官员过度聘用麦肯锡等机构往往是为了转嫁政治责任，在项目失败时将其作为规避公众问责的挡箭牌，导致政府陷入 1960 年代 NASA 官员所警惕的“宣传册统治”（brochuremanship）与幻灯片管理陷阱。为扭转这种能力流失，政府应在法律上禁止双重业务，在采购合同中强制推行知识转移与内部培训，设立诸如智利“政府实验室”（Laboratorio de Gobierno）的内部创新机构，并借鉴英国政府数字服务局（GDS）重构 Gov.uk 的经验，建立以公民权利与使用尊严为核心的公共部门技术团队。 大科技巨头的寻租垄断与 AI 治理困境 (30:00) 尽管现代人工智能、互联网和 GPS 的底层研发最初都依赖 DARPA、NSF 等公共资金资助，但当前技术收益与顶尖人才却高度集中在少数私营巨头手中。随着科技公司以数倍甚至百倍于学界的薪资挖角科研团队（如 Uber 曾直接整体聘走卡内基梅隆大学机器人团队），公共和学术机构面临严重的人才流失，导致监管机构在技术理解上面临巨大落差。Mazzucato 强调，监管者无需读懂每一行底层代码，但必须建立类似于环境披露的算法透明度标准，遏制技术垄断带来的算法寻租与市场操纵。 针对当前的技术飞跃，政府必须在研发资助初期确立“预先分配”机制。以新冠疫情期间牛津大学与阿斯利康的合作为例，校方要求企业加入专利池、保持低价并向全球开放技术分享，为公私合作树立了范式。反之，若任由以股东价值最大化为导向的企业主导，技术将被大量浪费在低效环节——例如美国医疗体系中保险公司算法与医院账单算法互相争斗的“机器人互掐”内耗，而真正关乎人类生存的水资源危机、粮食安全和气候破坏等重大挑战却被严重忽视。 跨越政治周期的长期使命与政策尊严 针对选举更迭导致政策反复翻转的挑战，Mazzucato 指出各国需要设立具有制度缓冲机制的使命型组织。诸如美国的 DARPA、智利的 Corfo、芬兰的 Sitra 以及瑞典创新局（Vinnova），均通过聘任固定年限的专业专家团队，以承担创新风险和系统性成果作为考核标准，从而避免被四年的短视选举政治所绑架。在实践工具上，德国复兴信贷银行（KfW）通过附加具体减排条件的贷款促成了钢铁产业向绿色钢铁转型，巴西国家开发银行（BNDES）将生态转型确立为核心放贷门槛，瑞典则将联合国可持续发展目标落实为全国性的健康校园午餐采购计划。 Mazzucato 进一步强调，宏观经济政策若要抵御极端政治民粹主义的冲击，必须在微观执行中修复民众的尊严与自我认同。仅靠自上而下的冷冰冰救济往往让底层感到被蔑视，例如在伦敦卡姆登区的成人照护改革中，政府让看护者与被照顾群体共同参与采购方案设计；在应对贫困时，将传统慈善型食物银行改组为受助者拥有投票决策权的绿色食物合作社，让公民切实获得参与政策制定的权利与主体性。 播客主持人在总结中指出，实现跨时代的突破性目标必须借助私营部门的研发制造能力，但前提是公共机构必须保有足够的内部专业判断与方向把控力。目前人工智能不仅面临被过度炒作且难以匹配底层制度改革的困境，其自身的技术特性也带来了加剧不平等和安全风险的双重挑战；若不对医疗、基建等失灵的社会系统进行深层治理，仅在表面叠加自动化算法，技术进步最终只会沦为增加系统复杂性的无效内耗。

## Background 本期节目由彭博（Bloomberg）旗下的知名播客《Odd Lots》于 2026 年 5 月 8 日在 RSS 及各大网络平台发布。节目由 Joe Weisenthal 与 Tracy Allaway 主持，在西班牙马德里举办的彭博 CityLab 峰会现场录制，特邀伦敦大学学院（UCL）创新与公共目的研究所（IIPP）创始所长玛丽安娜·马祖卡托（Mariana Mazzucato）教授，深入探讨“任务导向经济”、重塑国家能力、防范咨询业导致的公共职能空心化，以及从阿波罗登月到现代人工智能治理等核心议题。 ## TL;DR 马祖卡托在对话中犀利批判了将政府角色矮化为单纯“修补市场失灵”的传统经济学教条，主张政府应借鉴阿波罗登月工程，转向由宏大社会目标驱动的“任务导向型”产业政策，从垂直“挑选赢家”转变为“挑选意愿者”。她系统拆解了“国家能力”的三层架构，痛斥公共部门过度依赖商业咨询公司导致的核心职能空心化与利益冲突，呼吁重塑政府内部的动态创新能力。针对当下席卷全球的人工智能浪潮，她警告顶尖科研人才向极少数科技巨头的单向失血正破坏民主监管能力，提出必须推行“事前分配”与透明的“算法地租”披露框架，避免 AI 退化为缺乏系统改革的算法内卷工具。 ## Atmosphere & style 对谈在马德里峰会现场的轻松调侃中开启，随后迅速切入极具学术穿透力与批判锋芒的思想交锋。马祖卡托语速飞快、直言不讳，以极具感染力的激情打破官僚术语，既有对“番茄酱算蔬菜”或“PPT 治国”等政治闹剧的辛辣讽刺，又有对底层公民参与治理与重获尊严的深切人文关怀。两位主持人以生活化的纽约市公立学校午餐、华尔街算法监管等现实案例精准抛接，使整场对话在宏大政治经济学理论与接地气的微观治理现实之间自如穿梭，充满智力激荡与振聋发聩的现实批判感。 ## Bulletin - `05:00` 玛丽安娜·马祖卡托联合彭博推出“公共部门能力指数”，呼吁政府重塑“动态能力”并打破对通用咨询巨头的重度依赖，防止核心治理职能空心化。 - `08:00` 真正的产业战略应从垂直“挑选赢家”转向以重大社会挑战为核心的“任务导向”（Mission-Oriented），并将传统的成本加成采购重塑为以结果为导向的创新杠杆。 - `30:00` 当前 AI 发展正重演“索洛悖论”，顶尖科研人才加速由公共机构流向少数私营科技巨头，若缺乏“事前分配”和公共顶层设计，技术将滑向算法地租与内部消耗。 - `44:30` 巴塞罗那吸纳白帽黑客治理城市数据、英国设立自主数字服务（GDS）等实践表明，体制内部建立敢于试错的创新文化完全能够跑赢纯外部外包。 ## Key takeaways - `00:00` 地方市长治理往往比国家政治更务实，直面住房、贫富差距与技术落地等具体挑战，更易跨越意识形态壁垒展开政策试错与经验交流。 - `06:30` 公共部门决策长期受困于静态的成本效益分析与净现值（NPV）测算，忽视了像阿波罗登月工程那样创造新市场和跨领域技术溢出的动态价值。 - `08:00` 天量补贴与税收减免不等于真正的产业政策，若缺乏明确的约束条件与战略方向，极易重蹈“风险社会化、利润私有化”的覆辙。 - `10:00` “任务导向型”政策的核心是从“挑选赢家”转变为“挑选意愿者”，学校优质午餐等接地气的民生目标同样可以被构建为需跨部门协同的登月任务。 - `11:42` 国家能力包含三个层次：资源规模、行政惯例与动态能力，其中最稀缺的是跨部门敏捷协同与主动塑造市场的动态能力。 - `16:08` 商业咨询公司的盈利模式基于客户的长期依赖，将核心公共职能外包不仅削弱政府自身学习能力，还带来了严重的双向利益冲突。 - `20:00` 公私合作并非“全盘国有化”与“甩手外包”的二元对立，政府必须保有足够的内部专业认知，以防陷入依赖宣传材料和 PPT 治理的困境。 - `28:00` 现代智能手机及 AI 的底层核心技术大多源自数十年公共研发投入，但当下的科技红利与知识资产正被极少数科技巨头私有化并形成垄断。 - `32:15` 公共部门亟需推行“事前分配”（pre-distribution）理念，在研发合同签署之初便锁定专利共享、合理定价与公共利益条款（如牛津与阿斯利康的疫苗合作）。 - `34:30` 针对 AI 黑箱与算法寻租，监管者应推动类似 ESG 或巴塞尔协议的信息披露框架，并将算力能耗与水资源消耗等生态指标纳入国家治理仪表盘。 - `40:00` 单点数字技术无法替代基础设施建设，缺乏公共制度支撑的 AI 应用最终只会陷入如医疗账单“机器人互搏”般的无意义内卷与寻租。 - `44:30` 巴塞罗那市政厅招募白帽黑客利用市民流动数据治理交通与住房，提供了将关键技术攻坚能力收回体制内（insourcing）的成功范例。 - `46:00` DARPA 等机构通过错开选举周期的任期设计、鼓励冒险的考核指标及借调制度，证明了公共部门完全有能力建立跨越政治周期的创新韧性。 - `50:00` 抵御民粹主义不能仅依靠自上而下的经济施舍，必须让边缘群体直接参与政策制定（如伦敦卡姆登区的参与式护理采购），重塑公民尊严。 - `51:49` 巴西国家开发银行与德国复兴信贷银行通过将信贷与绿色脱碳指标深度挂钩，成功推动了重工及钢铁等传统行业的低碳技术转型。 - `54:30` 英国政府数字服务（GDS）在经历外包失败后组建内部顶尖团队，将纳税人重新定义为“享有权利的用户”，创造了兼具人性化与高效率的公共数字底座。 ## Detailed walkthrough ### `00:00` 马德里现场：市长政治、地方治理与全球前沿挑战 在西班牙马德里举行的彭博 CityLab（城市实验室）大会现场，Odd Lots 主播 Joe Weisenthal 与 Tracy Allaway 开启了本期对谈。二人不仅调侃在西班牙狂吃火腿与奶酪、笑称自己快变成行走的“伊比利亚火腿”（jamón），更将视线投向了本次大会所聚焦的城市政治议题。Joe 指出，虽然节目平时很少专门讨论市长层级的政治，但身处峰会现场，能够极为清晰地感受到市政议题与他们长期探讨的创新、科技、政策执行以及公私部门交界点等主题天然契合。 Tracy 进一步补充，无论是人工智能（AI）、住房负担能力还是贫富差距，城市都处于直面这些全球性挑战的最前线，地方政策的实际执行成效不仅直观可见，而且极易被公众检验。Joe 与 Tracy 都注意到一个有趣的政治反差：不同于国家级政治往往充斥着严重的意识形态壁垒，市长层级的治理往往显得更为务实且非意识形态化。以巴尔的摩市长为例，市长之间的交流核心永远是极其具体的“经验互换”——哪项政策在你的城市奏效了、哪项彻底失败了。相比之下，根本无法想象像特朗普与习近平这样的国家级领导人会坐在一起，轻松交流各自搞了哪些很酷的国家交通项目，并问对方“你要不要也试一试”。 ### `05:00` 玛丽安娜·马祖卡托登场：从“共同利益经济”到告别咨询依赖 随后，本期嘉宾玛丽安娜·马祖卡托（Mariana Mazzucato）正式受邀加入对话。作为伦敦大学学院（UCL）创新与公共目的研究所（IIPP）的创始所长及著名经济学家，她是 Odd Lots 开播以来听众呼声最高、等待最久的重磅嘉宾之一。谈及此次马德里之行，马祖卡托透露了她与西班牙政府的深度合作：西班牙政府在以新经济思维重塑政府职能方面走在前列，她此行专程会晤了西班牙首相以及经济部长卡洛斯·奎尔波（Carlos Cuerpo），并共同成立了一个名为“全球共同利益经济委员会”（Global Council for a Common Good Economy）的全新机构。 除了国家级战略层面的沟通，马祖卡托参加彭博 CityLab 峰会的一个核心任务，是发布并介绍 IIPP 与彭博合作开发的“公共部门能力指数”（Public Sector Capability Index）。该指数旨在推动对公务员系统内部能力的再投资，重建政府应对极端棘手复杂问题（wicked problems）的核心能力。马祖卡托言辞犀利地痛陈了政府内部能力流失的恶果，直斥在新冠疫情期间，英国政府竟以“每天 100 万英镑”的天价聘请德勤（Deloitte）等咨询巨头来搭建检测与追踪（Test and Trace）系统，最终却落得一地鸡毛、惨遭失败的荒唐现状。 ### `06:30` 重构经济学模型：为什么登月工程不能按成本效益分析？ 当被问及像西班牙首相这样的国家领导人究竟想从她那里获得何种指导时，马祖卡托坦言自己非常幸运，其学术著作（尤其是 2013 年出版的《创新型国家》/ *The Entrepreneurial State*）被很多领导人反复研读；西班牙政府甚至曾基于其理论撰写了题为《西班牙的创新型国家》（*El Estado Emprendedor en España*）的官方报告。领导人最常向她请教的核心问题是：“创新型国家”在实操中究竟意味着什么？特别是在“容忍失败”的机制上，政府该如何破局。 马祖卡托指出，资本市场存在一种极不平等的文化双重标准：风险投资人（VC）总是肆无忌惮地到处吹嘘自己经历了多少次惨痛失败才换来一次成功，而公务员、政府部长或国家首相只要遭遇一次项目挫折，就会立刻登上各大报纸的头条并遭到严厉抨击。因此，政府亟需改变舆论叙事、推动组织文化重塑，但最根本的任务，是推翻财政部或经济部内部根深蒂固的陈旧经济学底层逻辑： - **陈旧静态指标的局限**：如果政府在决策时永远只盲目依赖成本效益分析（cost-benefit analysis）和净现值（NPV）等静态测算工具，那么在 20 世纪 60 年代，人类根本连“登月”的念头都不会动。 - **动态公共目的导向**：评估重大战略不能仅看眼前的财政收支比，而必须建立起一套衡量创造全新市场、催生跨领域溢出效应的新框架。 ### `08:00` 产业政策的真伪：“任务导向”颠覆部门补贴 随着 2020 年疫情暴发，全球宏观经济思维发生了巨变，产业政策（Industrial Policy）强势重回主流视野。面对 Tracy 关于 2020 年后的政策转向是否证明了其前瞻性的提问，马祖卡托却给出了清醒且严谨的辨析。她强调，不能把“政府开始砸大钱”简单等同于“真正的产业战略回归”：无论是美国的《通胀削减法案》（IRA），还是欧盟总额高达 2 万亿欧元的“下一代欧盟”（Next Generation EU）复苏计划，如果缺乏战略性构建，这些天量资金很容易退化成新一轮盲目的税收激励与针对特定行业的补贴施舍。这种模式的终局，只是再次陷入“风险社会化、利润私有化”（socializing risks and privatizing rewards）的旧泥潭。 马祖卡托打趣道，美国政府从来都拥有产业战略，只是长期以来“嘴上说着杰斐逊式（小政府自由放任）的辞令，背地里走的却是汉密尔顿式（主动作为的强国家战略）路线”。她开玩笑说，直到百老汇音乐剧《汉密尔顿》大火之后，公众才终于听懂了她常年谈论的宏观历史脉络。 为了摆脱传统补贴模式的陷阱，马祖卡托在其著作《登月任务经济学》（*Mission Economy*）及领导的 IIPP 中极力倡导转向**“任务导向”（Mission-Oriented）**框架： ```mermaid flowchart TD accTitle: Traditional Industrial Policy vs Mission-Oriented Policy accDescr: Comparison showing the flaw of sectoral picking versus the multi-sector coordination driven by missions. subgraph OLD["传统产业政策（垂直挑选）"] direction TB O1["选择特定部门 / 技术 / 中小企业"] --> O2["直接发放补贴与税收激励"] O2 --> O3["风险社会化 / 利润私有化"] end subgraph MISSION["任务导向型策略（以问题为中心）"] direction TB M1["确立宏大社会目标（如登月工程）"] --> M2["设定清晰时限与具体指标"] M2 --> M3["跨行业跨部门协同攻关"] M3 --> M4["基于创新意愿与条件筛选支持对象"] end ``` 马祖卡托明确主张，政府必须彻底停止把资源单纯聚焦于“特定部门、特定技术或特定企业类型”（例如对中小企业的过度执迷），转而将目光锁定在如阿波罗登月工程般宏大、清晰且有时限的重大社会挑战上。在“任务导向”框架下，政府支持特定行业或企业，绝非因为它们挂着某种“行业标签”，而是因为它们真正具备攻克这一宏大目标、推动创新破局的意愿与实际行动力。 ### `10:00` 从“挑选赢家”到“挑选意愿者”：学校午餐也能成为登月任务 玛丽安娜·马祖卡托（Mariana Mazzucato）指出，产业政策的核心逻辑必须从传统的“挑选赢家”（picking winners）转变为“挑选意愿者”（picking the willing）——即挑选那些愿意与政府携手攻克极其棘手挑战的企业与机构。这类挑战不一定非得是火箭升空，它完全可以具体到“确保全国每一个孩子都能吃上一顿健康、美味且可持续的学校午餐”。她犀利地调侃道，这绝非当年罗纳德·里根（Ronald Reagan）推行削减成本政策时的做法——当时里根政府甚至荒唐地宣称“番茄酱是一种蔬菜”（ketchup is a vegetable），以至于大家都穿着印有里根头像和“番茄酱就是蔬菜”标语的文化衫抗议。 将学校食堂变成一项“登月任务”（moonshot），需要跨越多个维度的创新与协作。马祖卡托强调，这绝不可能单靠一个教育部独自搞定，而是需要教育部、农业部、卫生部与财政部坐到同一张桌子前协同发力。这种跨部委的协同机制并非空想，在新冠疫情（COVID-19）危机期间，各国曾设立“作战室”（war room）、情势分析室，将军方、教育、卫生等部门紧密聚集在一起攻坚；然而一旦危机告一段落，官僚系统立刻又迅速退缩回各自为政的“筒仓化”（siloed）孤岛之中。 主持人对此产生了极大的生活共鸣。主持人提到自己家有两个孩子在纽约市公立学校就读，孩子们对纽约市长的认知几乎完全取决于学校午餐政策：“他们甚至会煞有介事地讨论，‘以前我们还能吃到非常好吃的华夫饼，结果埃里克·亚当斯（Eric Adams）上市长之后就给取消了’。”主持人戏称，纽约的小学生们简直成了“只看有没有华夫饼投票的单一议题选民”（single-issue voters on waffles），这也恰恰证明了公共政策哪怕细微到一顿餐食，对公众生活的影响都极其切近。 - 里根政府时期的预算削减曾引发将番茄酱列为蔬菜的荒诞丑闻，成为反对公共服务紧缩的标志性文化符号。 - 登月型午餐计划需要农业（供应链与可持续种植）、卫生（营养标准）、教育（校园执行）与财政（采购预算）的深度跨部委协同。 - 危机时刻形成的战时跨部门协作（如疫情作战室），往往在平时难以制度化保留，暴露了政府治理模式的惯性倒退。 --- ### `11:42` 拆解“国家能力”：资源规模、行政惯例与动态能力 面对主持人关于“国家能力”（state capacity）这一流行词确切定义的追问——例如在疫情初期搭建检测基础设施的迟缓常被归咎于国家能力不足——马祖卡托给出了层次清晰的理论剖析。她将通常被混为一谈的概念拆解为三个明确维度：资源能力（Capacity）、行政惯例（Administrative Routines）与动态能力（Dynamic Capabilities）。 ```mermaid flowchart TD accTitle: 三层国家能力模型框架 accDescr: 展示资源能力、行政惯例与动态能力三者构成的国家能力体系及导向结果 subgraph Layers ["国家能力三层架构"] C["资源能力 (Capacity)<br/>财政空间、编制预算、公务员基础培训"] R["行政惯例 (Routines)<br/>稳定的运作环境、干中学 (Learning-by-doing)"] D["动态能力 (Dynamic Capabilities)<br/>敏捷度、跨部门协同、生态学习与公私伙伴关系"] C --> R --> D end subgraph Outcome ["治理模式输出"] Fix["传统被动修补<br/>市场失灵理论 / 贴创可贴"] Shape["主动塑造市场<br/>登月式协同 / 服务人与星球"] end D -.缺乏.-> Fix D ==> Shape ``` 马祖卡托解释道，“资源能力”（Capacity）主要关乎基本盘：你手头是否有财政空间？是否有明确的资金预算？政府团队编制了多少人，公务员是否接受过合规培训？但光有资源是死的，必须依赖“行政惯例”（Administrative Routines）提供一个稳定的组织运作环境，唯有如此才能形成“在实践中学习”（learning by doing）的经验沉淀；若机构人员流动频繁、组织架构朝令夕改，最基础的运作惯例都会瓦解，正如她在高校管理中亲眼目睹的混乱。而最关键也最匮乏的，是“动态能力”（Dynamic Capabilities）——这决定了政府拿到资源后究竟能办成什么事：政府是否足够敏捷、灵活？能否在疫情等突发危机中迅速打破壁垒展开跨部委协同？能否与外部构建健康的伙伴关系？市长之间是否具备相互学习和适应的能力？ 马祖卡托指出，投资动态能力的先决条件是彻底抛弃传统经济学的教条。传统主流经济学对政府角色的设定，最差不过是“赶紧让开别插手”，最好也只是被动地“修补市场失灵”（fix the market failure）。如果政府被预设为只配当“缝缝补补的修理工”，那系统自然只会准备一大堆应急创可贴，政策便陷入在极化之间反复摇摆的被动填坑状态；唯有承认政府的核心使命是“共同塑造与共创一个造福人类与地球的新型经济体系”，政府才会真正有动力去自建强大的内生动态能力。 - **资源能力（Capacity）**：侧重硬件与投入要素，包括财政空间、预算规模以及公务人员的编制与基础通识培训。 - **行政惯例（Administrative Routines）**：提供持续稳定的组织记忆与制度环境，是公务系统积累“干中学”经验的关键蓄水池。 - **动态能力（Dynamic Capabilities）**：体现为危机应对中的战略敏捷性、协同创新力、生态伙伴构建力与组织迭代学习能力。 --- ### `16:08` 咨询业的泛滥与“大骗局”：政府能力的空心化与利益冲突 主持人提出尖锐疑问：外部咨询机构（consultants）是否已经彻底沦为国家能力、行政惯例与管理技能的“替代品”？在某种程度上，每次政府高薪聘请外部顾问，难道不都相当于公开承认自身能力的溃败吗？为什么这种现象反而成为全球公共部门心照不宣的标准惯例？ 马祖卡托提到了她合著的著作《大骗局》（*The Big Con*），该书副标题针砭时弊地揭示了咨询业如何“削弱企业、弱智化政府并扭曲我们的经济”（How the Consulting Industry Weakens Our Businesses, Infantilizes Our Governments, and Warps Our Economies）。然而马祖卡托出人意料地表示，这绝不能单纯怪罪咨询公司——问题根源在政府自身，是政府主动把大门敞得太开。政府借助专家顾问建言献策本无可厚非，但致命之处在于将**核心公共职能**拱手外包。例如在英国疫情期间，极其关键的“检测与追踪”（Test and Trace）系统竟被外包给外部咨询公司主导。这种现象可追溯至20世纪80年代里根和撒切尔推行的新自由主义“精简政府”运动，随后甚至在英国工党执政期间进一步变本加厉；出于意识形态而非战略需要的一刀切瘦身，非但没有消除赘肉，反而让政府在面对洪灾、英国脱欧或全球疫情时，瞬间丧失了应对基本危机的核心内生能力，最终反噬并耗费了更多纳税人资金。 不仅如此，马祖卡托揭露了咨询外包机制背后可怕的结构性利益冲突。首先，商业咨询公司从根本上缺乏“让政府变得自立自强”的激励——如果客户自己学会了解决问题，后续价值千万的续约合同该去哪里找？主持人形象地比喻道：“这就像看了一辈子心理医生；如果你的心理医生看了你一辈子都没让你好转，那他大概算不上一个好医生。”咨询模式的盈利逻辑正是建立在客户的持续依赖上。其次是触目惊心的“两头通吃”（double-dealing）： - **南非 Eskom 案**：咨询巨头一边为深陷困境的南非国家电力公司（Eskom）提供咨询，另一边却同时服务理应严格监管 Eskom 的南非国家财政部（National Treasury）。 - **澳大利亚普华永道（PwC）丑闻**：咨询机构一边为医疗器械公司提供商业策略，另一边又作为顾问直接参与甚至协助制定医疗器械行业的监管政策。 马祖卡托疾呼，这种两头下注的行为在法律上必须彻底予以禁止。解决这一顽疾的首要途径，是重新对政府内部能力进行战略性投资，大幅降低对通用咨询公司的过度依赖；而在必须引入专业咨询服务时，合同条款中必须强制嵌入“知识转移与组织学习机制”，确保顾问带来的经验沉淀在公共系统内部，并确保引入的是具备扎实专业背景的真行家（例如制定肿瘤学战略必须依托真正的医学专家），而非仅凭模板套话揽财的通用咨询团队。 ### `20:00` 咨询机构的局限与纽约垃圾清理的“外包迷思” 玛丽安娜·马祖卡托（Mariana Mazzucato）深入剖析了当今政府过度依赖咨询公司所带来的结构性弊病。许多顶尖战略咨询顾问进驻公共部门后，往往缺乏真正的垂直领域专业知识，甚至频繁反向骚扰那些本就负荷过重的公务员，索要内部报告和已有规划来充当自己的研究材料。马祖卡托直言，全球学术界有大批耗费数十年深耕特定课题的研究中心却被束之高阁；以澳大利亚为例，政府当年并未动用自身深耕气候科学的专家，反而豪掷巨资聘请麦肯锡（McKinsey）来制定国家气候战略，最终交付的方案堪称灾难。政府不仅未能持续投资内部治理能力，还往往为了买一份“安全感”而找咨询公司盖章背书。 针对这一观点，播客主持人提出了不同视角的现实反例：纽约市近期因斥巨资聘请外部顾问重新设计全市垃圾清理系统（trash collection system）而引发热议。支持外包的逻辑在于： - **一次性设计 vs. 长期日常运营**：维护、推行和日常服务属于政府固有的常态职能，但全局重新规划是一次性任务，政府没必要在内部永久保留一套昂贵且低频使用的“专项设计肌肉”。 - **第三方外部视角的独立价值**：引入外部机构能跳出既有体制局限，协助理清架构。 马祖卡托对此明确回应，这恰恰暴露了关于公私合作的一大误区——非黑即白地认为“要么政府把一切都大包大揽（乃至全面国有化），要么政府彻底甩手私有化、全权委托咨询公司”。现实永远处于中间状态。政府固然不需要掌握全部具体技能，但必须具备足够的内部专业认知，去理解外部生态体系、精准识别合格的合作方，并有能力拟定真正具有战略前瞻性的重塑方案。 --- ### `22:15` “阿波罗登月”与以结果为导向的公共采购 为了阐明何为正确的“任务导向”（Missions），马祖卡托重温了 NASA 当年登月计划的组织模式。NASA 制定登月目标时，既没有试图闭门造车，也没有单纯依靠航天防务部门，而是动员了约 40 万名私营部门从业人员。NASA 的核心做法是设定极其清晰的攻坚方向，以问题为导向与私营部门协同破局： ```mermaid flowchart TD accTitle: 从成本加成到以结果为导向的采购转型 accDescr: 对比传统采购模式与NASA任务导向采购模式的机制与创新溢出路径 subgraph OLD["传统采购模式"] A1["成本加成采购 (Cost-Plus)"] --> A2["单纯追求成本最小化"] A2 --> A3["过程死板 / 抑制外溢创新"] end subgraph NEW["NASA 任务导向模式"] B1["以结果为导向的采购 (Outcomes-Oriented)"] --> B2["明确核心痛点：饮食、排泄、通讯、宇航服"] B2 --> B3["公共部门引导构想 + 私营机构研发解法"] B3 --> B4["广泛跨界技术溢出：相机手机、保温箔毯、新型绝缘材料、基础软件"] end ``` 公共采购（Procurement）通常占到政府总预算的 30% 左右，无论是巴巴多斯这样的小岛屿国家还是美国这样的超级大国，采购都是最关键的财政杠杆。NASA 彻底抛弃了单纯压低成本的“成本加成”（Cost-Plus）采购，转向“以结果为导向”的采购模式。他们围绕一系列极端具体的现实问题立项——宇航员在太空中如何上厕所（哪怕在当下的阿尔忒弥斯登月计划中，太空马桶依然是技术难题）、宇航员吃什么、穿什么、地月之间如何通信。在 NASA 的战略指引和先导投资下，私营部门在攻克这些具体问题的过程中催生了海量民用外溢技术，涵盖手机摄像头、航天保温箔毯、家用建筑绝缘材料、尖端材料科学以及现代软件系统。 马祖卡托强调，无论面对的是登月、城市垃圾治理、学校营养午餐，还是下一场不可避免的大流行病预防，政府都必须建立起这种“问题导向、方案导向”的敏捷思维。这正是彭博城市实验室（CityLab）以及公共部门能力指数（Public Sector Capability Index）所倡导的——总结试错经验、突破行政瓶颈，并重拾约翰·肯尼迪所倡导的雄心：“我们选择做这些事，不是因为它们容易，而是因为它们艰难”。当代政策文件充斥着“促进（Facilitate）”、“去风险（De-risk）”或“使能（Enable）”等消极词汇，但如果政府仅仅将自己定义为外包中介或修修补补的兜底者，就永远无法与私营部门建立平等且富有雄心的合作关系。 --- ### `25:00` 责任稀释、PPT治国与 NASA 的警示 针对“咨询崇拜”的根源，主持人一针见血地指出其背后的避责动机：正如商界老话所言，“没有人会因为买了 IBM 的产品而被解雇（You'll never get fired for buying IBM）”；官员只要聘请了麦肯锡，万一项目砸锅便可以推卸责任：“我已经尽力了，如果失败了那是麦肯锡的错”。 马祖卡托赞同这一剖析，并指出了深层的文化病灶： - **责任分散与怯懦文化**：政府内部对失败采取零容忍态度，公务员缺乏风险投资和创业生态中对试错的包容，于是通过引入第三方咨询公司来分担政治责任。这种“管理咨询化（Consultification of management）”在企业界同样泛滥——管理层在推行裁员、并购或大规模股票回购时，只要搬出“这是麦肯锡建议的”，就能免受问责，本质上是一种怯懦的推诿。 - **打造政府实验室**：智利的政府实验室（Laboratorio de Gobierno）提供了一种全新样板，即在公共部门内部建立起敢于实验的创新空间。马祖卡托讽刺道，在意大利语或西班牙语里，“公共行政改革”听上去十分高大上，但往往“拆穿了其实就是缩减开支（literally just cuts）”。 马祖卡托重温了 20 世纪 60 年代阿波罗计划时期 NASA 采购主管欧内斯特·布拉克特（Ernest Brackett）的名言。布拉克特不仅重构了采购体系，还极具前瞻性地警告走廊里游荡着太多外部顾问。他尖锐地指出，如果政府放任自身能力退化，终将被“宣传画册术（Brochuremanship）”所俘获。在没有 PowerPoint 的年代，顾问依靠光鲜的印刷画册夸夸其谈；在今天，则演变为了“凭 PPT 统治（Ruling by PowerPoint）”。布拉克特当年的警世格言至今振聋发聩：公共部门若自己丧失了智力内核，不仅写不出精准的采购规格与招标文件（Terms of reference），更将彻底丧失与私营部门聪明、平等合作的能力。 --- ### `28:00` 生成式人工智能的转折点：公共投入与私人垄断的危机 对话顺势转向当下的核心焦点——人工智能（AI）。主持人提出：在学校午餐或太空探索中，政府可以率先定义使命再调用私营力量；那么面对耗资巨大的 AI 浪潮，政府是否也应率先确立明确的国家使命？ 马祖卡托回顾了技术的演进史，强调当今包括大语言模型（LLM）和语音识别在内的 AI 底层技术，追根溯源都来自数十年来由公共部门主导的先导投资： - 从早期的 DARPA（美国国防高级研究计划局）为了解决军事国防需求而研发出阿帕网（互联网雏形），到奠定当今 AI 技术根基的早期算法； - 正如她在《创业型国家》（The Entrepreneurial State）中所论证的，现代智能手机中所有使其“智能”的底层核心技术——互联网、GPS、触控屏、Siri 语音助手等——最初全部由公共资金孕育而成。 但马祖卡托发出严正警告：**今天的 AI 格局正在发生令人惊恐的异变**。以往“公共投资开拓前沿、社会共享发展红利”的良性机制濒临瓦解。极少数科技巨头将全社会集体创造的巨大科技价值彻底私有化，攫取了天文数字级的超额“经济租金（Economic rents）”而非正常市场利润。这些企业坐拥动辄数万亿美元的资金体量，开出公立机构望尘莫及的天价薪酬争抢稀缺人才。这种前所未有的资源垄断，正在彻底动摇公共部门在下一代通用技术前沿领域的主导权与话语权。 ### `30:00` 人才失血：顶尖科研力量向少数 AI 巨头的倾斜 玛丽安娜·马祖卡托（Mariana Mazzucato）发出警告：当下公立及私立大学的顶尖研究人员，以及曾经在 NASA、DARPA 甚至智利的 CORFO（生产促进局）等公共研发机构效力的专家，正在以极快速度流向大型科技公司。她认为这场科研人才的严重流失（hemorrhaging of talent）是当代技术变革中最深远、但讨论却最不充分的转变。这与以往通用技术（如电气化）的崛起截然不同——过去虽然也重塑了生产和分销，但知识从未像今天这样高度向少数私营巨头聚拢。乔·维森塔尔（Joe Weisenthal）补充道，优步（Uber）当年在研发自动驾驶汽车时，就曾直接开出一张支票，把卡内基梅隆大学（CMU）整个国家机器人工程中心的核心教研团队整体挖走。 马祖卡托强调，如果最关键的知识只集中在少数私营 AI 企业内部，而公共部门既缺乏人才也无法真正理解其底层技术运作，那么民主选举的政府就根本无法对其进行符合伦理、有利于人类和地球的治理。虽然像 NSF（美国国家科学基金会）等公共资金资助了早期的大量基础研究，但一旦人才彻底外流，公共部门的监管能力就会被直接掏空。民主政府有义务确保重大技术红利造福社会整体，但这不仅需要重建国家能力，更需要一种新型监管：深入理解算法是如何被设计出来的。 ```mermaid flowchart TD accTitle: Public Research to Private Monopoly Transition accDescr: Diagram illustrating how public funding generates base talent, which then drains into AI mega-corporations. subgraph PublicSector["公共部门与学术界"] A["公共研发资助<br>(NSF / DARPA / NASA)"] --> B["顶尖学者与大学教研团队<br>(如 CMU 机器人团队)"] end subgraph BrainDrain["人才严重失血 (Brain Drain)"] B -- "巨额薪酬挖角" --> C["少数私营 AI 巨头"] end subgraph GovernanceCrisis["治理困境与失衡"] C --> D["知识与算法黑箱高度垄断"] D --> E["公共部门失去技术理解与治理能力"] C --> F["巨额超额利润<br>(避税 / 压榨劳工)"] end ``` - **CMU 案例**：Uber 在开发自动驾驶汽车期间，几乎将卡内基梅隆大学机器人团队的科研骨干全盘收编，成为科技巨头凭资本收割学术界研发能力的典型注脚。 - **监视资本主义**：马祖卡托引用肖莎娜·祖博夫（Shoshana Zuboff）在《监视资本主义时代》（*The Age of Surveillance Capitalism*）中的名言——“你以为你在免费搜索 Google，其实是 Google 在免费搜索你。”她指出，如果最初在算法架构设计中就植入伦理规范，这种掠夺性商业模式本可以避免。 --- ### `32:15` “事前分配”理念与阿斯利康新冠疫苗的合作典范 讨论进一步转向财富和技术的分配机制。马祖卡托提出了“事前分配”（pre-distribution）的概念，即在技术合作最初阶段就锁定伦理、定价与公共利益条款，而不是等到市场失灵、产生超额利润后再试图通过事后征税、反垄断或市场修补来解决。她列举了 COVID-19 疫苗研发的例子：当时全球疫苗研发的使命不仅是制造出疫苗，而是要让全球人口都能公平获取。 然而，在所有主要新冠疫苗中，唯有牛津大学与阿斯利康（AstraZeneca）的合作践行了这种机制。牛津大学作为一所公立研究型大学，在最初与阿斯利康达成合作协议时，便设立了明确的附加条件（conditionality）：阿斯利康必须加入专利池、共享知识、保持透明，并在疫情期间以成本价平价供应疫苗。这正是马祖卡托所倡导的“好合作”的具象体现，也是未来公私合作该遵循的范式。 针对主持人特蕾西·阿洛韦（Tracy Alloway）提到的现象——大学教授留在学术界意味着主动“放弃巨额金钱”（leave money on the table），去私营部门能赚几十甚至上百倍——马祖卡托直接质疑了科技巨头高薪资本的合法来源。她指出，巨头的原始知识积累本质上来自公共资助，但在获得垄断地位后，它们通过跨国避税在许多国家几乎不缴税，并在物流等环节极度压榨基层员工（例如疫情期间有调查曝光亚马逊甚至不在高温和劳累的仓储区外配备救护车）。这些凭借公共支持与避税掠夺而来的“超额利润”（excess returns），如今却成了它们反噬公共科研、收割顶尖人才的筹码。 - **预先分配 vs 事后修补**：传统经济学侧重于事后通过税收或福利政策进行再分配（redistribution），而马祖卡托主张在公私合作起草合同时就规定好专利共享、定价策略与利润分成（pre-distribution）。 - **牛津/阿斯利康范例**：作为接受公共资助的科研主体，牛津大学在合作源头确立了非营利、成本定价及专利共享约束，打破了医药巨头利用专利牟取暴利的惯性。 --- ### `34:30` 算法黑箱、信息披露与“算法地租” 维森塔尔抛出监管的现实痛点：早前他们采访高盛（Goldman Sachs）的首席技术官（CTO）时，询问那些审查高盛高度复杂 AI 系统与算法的银行监管员是否真能看懂底层代码。CTO 的坦白回答是：监管人员根本看不懂底层代码，他们能做的只是建立外围控制机制，防止算法“失控暴走”（running amok）。那么在 AI 时代，面对学术与公共部门技术官僚的急剧萎缩与私营部门的技术垄断，公众究竟能指望监管官员理解到何种程度？ 马祖卡托对此阐述，监管者确实不需要通读每一行算法代码，但必须像建立气候披露制度（climate disclosures）一样，建立针对 AI 的透明披露框架。她坚决反对所谓的“丰裕理论”（abundance theorem，该理论认为外部充满无限机会，是繁琐的政府规划、监管和附加条件扼杀了创新）；相反，她指出恰恰是追求股东价值最大化的现代企业治理模式抑制了真正的技术潜能。 为推进这一主张，马祖卡托透露她正与奥莱利媒体创始人蒂姆·奥莱利（Tim O'Reilly）合作，在奥米迪亚基金会（Omidyar Network）的资助下展开一项名为“算法地租”（Algorithmic Rents）的研究项目。该项目旨在探索如何设计出一套类似于金融领域的 ESG 指标或银行业巴塞尔协议式的 AI 披露标准，通过迫使科技巨头披露关键指标，减少算法对公共注意力和经济的掠夺性抽取，并尝试联合一批有社会责任意愿的 AI 企业共同构建披露联盟。 ```mermaid flowchart LR accTitle: Algorithmic Rent Disclosure Framework accDescr: Flowchart showing the shift from unmonitored algorithmic rents to systemic disclosure and dashboard governance. subgraph CurrentState["当前监管困境"] A["底层代码黑箱<br>(监管者无法看懂)"] --> B["算法地租滋生<br>(注意力和经济剥削)"] end subgraph Solution["系统性治理路径"] B --> C["奥莱利与马祖卡托项目<br>算法披露框架"] C --> D["环境/资源能耗关联<br>(水资源与数据中心算力)"] D --> E["政府综合仪表盘<br>(超越单一 GDP 考量)"] end ``` - **AI 与水资源/能耗消耗**：马祖卡托指出，AI 与气候问题密不可分。目前数据中心极为耗能耗水，例如每一次 ChatGPT 对话交互，其背后服务器冷却系统所消耗的水量相当于消耗了一整瓶小型塑料瓶装水，然而现行政策完全缺乏将 AI 治理与水资源、能源问题贯通的系统视角。 - **治理仪表盘（Dashboard）隐喻**：如果驾驶汽车时仪表盘上只有一个数字（要么只有剩余油量，要么只有行驶车速），车子一定会撞毁。政府治理现代技术同样如此——不能仅仅盯着 GDP 这一个指标，必须配备涵盖算力消耗、生态影响、收入公平与劳工权益的多维度“国家能力仪表盘”。 ### `40:00` 通用技术幻觉：AI 的索洛悖论与公共系统缺失 关于人工智能（AI）的技术狂热与资本喧嚣，让人联想起历史上各类通用技术（General Purpose Technology）初期的泡沫与困惑。AI 无疑具备改变一切的技术底色，但回顾技术史便会发现，电力在问世后经历了长达近三十年的漫长周期才真正改造了政府与产业的运作模式；互联网同样走过了漫长阵痛，正如经济学家罗伯特·索洛（Robert Solow）曾尖锐指出的那样：“计算机无处不在，唯独没有出现在生产力统计数据中。”如今的 AI 正处于极其早期的渗透阶段，根本问题在于：它到底在系统性地帮人类解决哪些能够规模化推广并促进跨政府学习的核心难题？倘若技术仅仅是在绕开公共监管以攫取天价租金，而未能 ethical（伦理化）且深度协同公共部门去解决气候危机与公共卫生等根本挑战，那么社会将面临巨大的扭曲。 针对政府行动迟缓的疑问——即便 OpenAI 发布了详尽的产业政策蓝图，山姆·奥特曼（Sam Altman）也公开坦陈 AI 会产生严重的负外部性并需要政府介入规制，各国政府却依旧显得跟不上节奏——这背后的核心根源是：任何单点技术凭自身都不可能凭空解决结构性问题。在洛克菲勒基金会位于科莫湖畔的贝拉焦（Bellagio）别墅举办的一场高端对话中，曾上演过一场意味深长的思想交锋：前谷歌首席执行官埃里克·施密特（Eric Schmidt）断言未来人们只需要一个健康 App 就能解决一切医疗问题；而印度 IT 巨头印孚瑟斯（Infosys）联合创始人南丹·尼莱卡尼（Nandan Nilekani）当场反驳道，在印度，如果没有健全的基础医疗体系与公共卫生网络，手机里装再多的 App，底层的苦难也丝毫不会减少。 当公共部门陷入财政紧缩并主动贬低自身职能、导致公共医疗与基础设施根本无法适配现实需求时，指望单凭 AI 拯救医疗无异于痴人说梦。这种资源错配在生态危机中尤为刺眼：由于对全球水文循环的破坏，目前全球 55% 的粮食系统正面临严重风险，亚马逊热带雨林的生物多样性丧失正在世界其他地区诱发严重的干旱与洪涝。然而，几乎没有人真正将强大的算力与 AI 算法投入到挽救水资源危机的登月任务中。这正是为何社会必须由政府主导、携手私营部门确立清晰的“登月目标”，将受到审慎规制的 AI 力量引向最具体的紧迫问题。 - 索洛悖论再现：当前围绕 AI 的金融市场泡沫与生产力实际落地的断层，重演了电力花费约 30 年才重塑政府运作的历史周期。 - 粮食与水系统警报：全球 55% 的粮食体系正因人类对水文循环的破坏而处于危险边缘，但极少有前沿 AI 资本聚焦于此。 - 贝拉焦辩论焦点：埃里克·施密特主张的“一个 App 解决医疗”遭到南丹·尼莱卡尼的坚决否定，凸显“数字技术唯有嵌在完备的公共体系中才能产生真实效用”的常识。 --- ### `44:30` 算法内卷与巴塞罗那的“黑客治城”实验 在没有公共顶层设计介入的情况下，市场驱动的 AI 往往会滑向荒诞的内耗。医疗领域就是一个典型缩影：大量初创公司开发 AI 帮助医院简化并最大化账单报销，而保险公司随即部署更强大的对抗性 AI 算法来审核、压低和拒付账单。这种“我的机器人跟你的机器人互相掐架”的军备竞赛没有让任何患者受益，也没有改善医疗系统的效率，唯一的赢家只有兜售对抗软件的“机器人制造商人”（bot makers）。 ```mermaid flowchart TD accTitle: AI 市场内卷与巴塞罗那公共数据治理对比 accDescr: 对比缺乏系统导向的算法内耗与巴塞罗那利用公共数据服务社会挑战的闭环设计 subgraph 内耗陷阱["商业内耗：缺乏公共目标的算法对碰"] A["医院收费端 AI<br/>（优化账单以索赔）"] <-->|"算法博弈军备竞赛"| B["保险审核端 AI<br/>（自动识别以拒赔）"] A & B --> C["结果：患者医疗无改善<br/>仅算法软件商获利"] end subgraph 公共导向["公共创新：巴塞罗那黑客进市政厅"] D["市民出行与生活数据<br/>(Uber / Citymapper)"] --> E["市政厅招募白帽黑客<br/>建立敏捷实验机制"] E --> F["精准公共政策制定<br/>(破解住房危机与公交网络优化)"] end ``` 针对这种技术扭曲，巴塞罗那前市长阿达·科劳（Ada Colau）提供了一种极具启发性的应对示范。作为基层住房维权运动出身的市长，科劳上任后敏锐地意识到：当市民每次在手机上点击 Uber 或 Citymapper 时，海量的城市流动数据即被沉淀。她提出质问：“为什么城市管理者不能利用这些市民日常创造的数据，来改善公共决策、深入洞察公共交通难题与保障性住房挑战？” 为此，巴塞罗那市政府并未将数据分析层层外包给昂贵的咨询巨头，而是反向操作，将一流的计算机黑客直接招入市政编制。市政当局为这群技术人才营造了一个敢于试错、打破官僚层级的工作环境，明确要求他们：“不要害怕失败，放手承担风险，帮助市政府直击住房不公、交通拥堵与资源准入不平等的深层顽疾。”这种“将核心创新能力收回体制内”（insourcing）的举措，生动展示了公共部门如何主动驯服技术工具为公共利益服务。 --- ### `46:00` 跨越选举周期的机构韧性：从 DARPA 到公共价值重塑 公共部门践行重大使命时面临的一个普遍现实是选举周期的制约：政客的首要本能往往是赢得连任，一旦选民换届，新政府很容易将前任的政策推倒重来，造成 90 度甚至 180 度的大转弯。然而，研究全球顶尖的任务导向型组织（如美国的 DARPA、智利的 CORFO、芬兰的 SITRA、丹麦的 MindLab、瑞典的 Vinnova）后可以发现，成功的创新机构在制度设计上早已对政治周期筑起了防火墙。 以 DARPA 为例，项目主管的聘期通常为 5 年，有意错开了 4 年一届的美国总统选举周期。其内部绩效考核标准不是“四平八稳、从不出错”，而是“你是否承担了足够大的风险”以及“成功项目到底撬动了多大的战略影响力”。此外，机构广泛推行借调制度（secondment），吸引顶尖学者与技术专家以公职人员身份服务数年，直接在体制内部为公共目标效力，而非以外部咨询顾问的身份抽空公共机构的学习能力。 长期以来，主流经济学叙事存在严重偏差：哈佛商学院积累了海量的私营企业商业案例库，但几乎从未为公共实体系统性建立案例研究，因为政府长期被贬低为单纯的“再分配者”、“市场修补匠”和私营部门的“辅助协调者”，人们理所当然地认为只有私营部门才能创造价值和发挥创意。为了打破这种成见，伦敦大学学院创新与公共目的研究所（IIPP）在其公共管理硕士（MPA）及实务培训课程中，专门为全球公务员开发针对公共机构的全新案例库： - 剖析英国广播公司（BBC）等公共广播机构如何定义并量化“公共价值”（Public Value）； - 探寻公共开发银行与数字化部门如何主动“塑造市场”而非仅仅“修复市场”； - 提炼公共部门建立实验与冒险文化、抵御因过度规避风险而沦为“咨询化”（consultification）俘虏的制度架构。 体制内部的不稳定并非公立部门所独有，私营部门同样充斥着动荡与短视；将官僚机构改造成具备韧性、富有创造力且横向联通的组织完全可行。然而，这也引出了当下全球民主制度中一个深刻的现实悖论：为何像拜登政府这样其经济与产业政策在红州（共和党选区）落地了大量先进制造业与基建投资的执政团队，在现实政治与选举选票的转化上却依旧步履维艰？这项全球性的治理难题，正向公共政策学者提出了更深层的挑战。 ### `50:00` 找回尊严与参与感：以人为本的“使命导向”政策 马祖卡托探讨了西方许多发达国家（如美国、英国、意大利）所面临的深层政治与社会脱节：即便某些经济政策取得了成效，为何曾经被边缘化（left behind）的群体依然在选举中投向民粹主义？她指出，仅仅提供经济利益是远远不够的。如果人们感受不到自己的价值，没有重获尊严，且持续感受到体制自上而下的傲慢与居高临下，不满情绪就不会消退。政策设计的关键不仅在于“自上而下给予”，而在于让受影响的公民坐到桌前共同设计。 她以自己在伦敦居住的卡姆登区（Camden，约 25 万人口）为例，介绍了如何将使命导向采购应用于成人社会护理。他们在 10 个公营住宅区（housing estates，相当于美国的 projects）中开展试点，直接邀请护理人员与受护理者坐到同一张谈判桌前共同制定政策，充分尊重他们的亲身生活经验（lived experience）。这种设计不仅让政策切中要害，更赋予了边缘群体自豪感与尊严。同样，她指出英国在 21 世纪依然广泛依赖慈善性质的“食物银行”（food banks）是一种倒退且野蛮的现象；更好的路径是将其改造成“绿色食物合作社”（green food cooperatives），让原本排队领救济粮的人成为治理者、参与审议与共同决策。她观察到，当人们不再是弯腰接受施舍，而是挺直腰板参与集体管理时，整个人的面部表情与体态尊严完全不同。 - **参与式社会护理采购**：卡姆登区跨 10 个公房社区，让护理员与被护理人直接参与政策设计，改善服务品质。 - **重构食物救济体系**：从被动排队领取临期食品的食物银行，转向赋予受助者共同治理权的绿色食物合作社，以此重塑尊严并抵御民粹主义情绪。 --- ### `51:49` 全球工业政策实战：巴西、德国与瑞典的使命创新 当被主持人问及全球有哪些国家或城市在“以战略使命为核心”的现代工业政策上表现出色时，马祖卡托强调，不应盲目去给某个整体政府贴上“完美”的标签，而应剖析其具体机制与工具的创新。在巴西，政府将生态转型置于核心，这倒逼财政部以及全球最大的公共开发银行之一巴西国家开发银行（BNDES）重新构想融资工具：公共信贷不再只是简单给农业综合企业发补贴或兜底濒临倒闭的钢铁厂，而是设立明确的转型条件，推动这些行业主动进行技术升级。 这一模式与德国通过复兴信贷银行（KfW）推进能源转型（Energiewende）的举措如出一辙。KfW 向面临生存压力的德国钢铁业提供有条件贷款，要求行业降低生产材料消耗，但不干涉具体技术路线，从而在保护创新的同时诞生了全球最绿色的钢铁产业。此外，瑞典创新署（Vinnova）将抽象的气候与可持续发展挑战聚焦为具体使命——“无化石燃料的福利国家”，并细化为“健康、美味、可持续的学校供餐政策”，成功撬动跨部门协同与地方供应链革新。 ```mermaid flowchart TD accTitle: 战略使命驱动公共工具转型框架 accDescr: 展示从宏观挑战到具体使命，再通过公共金融与采购等工具落地为产业创新的全流程 subgraph Challenge["顶层战略挑战"] A["宏观挑战：生态转型 / 无化石福利国家"] end subgraph Mission["明确的使命制定"] B["具体使命：绿色钢铁 / 可持续学校供餐"] end subgraph Tools["公共工具重塑"] C["有条件贷款：BNDES / KfW 设定环保指标"] D["公共采购：政府牵引本地绿色制造与供应"] end subgraph Outcome["产业创新成果"] E["企业自主技术突破：再利用、再循环的绿色生产"] end A --> B B --> C B --> D C --> E D --> E ``` - **巴西 BNDES**：以生态转型为核心，改变传统对农企及重工的盲目兜底，将贷款与行业脱碳深度挂钩。 - **德国 KfW**：在能源转型中为钢铁业提供约束性贷款，仅规定“降低生产材料消耗”的方向而不包办细节，催生了高复用、高回收的绿色钢铁工艺。 - **瑞典 Vinnova**：将宏观减碳转化为“健康可持续学校餐”具体抓手，打破部委壁垒，联动地方农业与食品制造。 --- ### `54:30` 重新定义公民：英国政府数字服务（GDS）与 Gov.uk 案例 马祖卡托指出，即便英国在过去十四年中饱受财政紧缩和政局不稳的困扰，其公共组织内部依然诞生过卓越的实践案例——英国政府数字服务（Government Digital Services, GDS）。在 2000 年代初期，英国政府为了避免让公众依靠 Google 去检索下载政策白皮书，决定打造自主数字平台。起初政府走上了传统的委外采购老路，将其外包给大型外包商 Serco，结果遭遇惨败。随后，来自 BBC iPlayer 团队的数字化人才入驻内阁办公室（Cabinet Office），创立了 GDS，打造出荣获国际设计大奖的集成平台 Gov.uk。 GDS 成功的关键在于组织文化的根本重塑：开发团队在窗前贴出箭头指向窗外，郑重申明“外面的人不是客户或买家，而是享有基本人权的用户”。在传统的福利申领体系中，繁复冰冷的行政壁垒常常让人痛苦不堪（正如肯·洛奇 Ken Loach 电影中所揭露的那样）；而 GDS 致力于通过人性化的技术流程，让公民在办理驾照、护照或选民登记时感受到顺畅与尊严。这一愿景使 GDS 成为极具吸引力的顶级工作场所，一度吸纳了最顶尖的软件工程师，以至于私营科技巨头都为其人才虹吸效应感到头疼。 - **外包失败与自主重构**：外包给缺乏创新能力的 Serco 失败后，政府引入 BBC iPlayer 核心人才组建内部攻坚团队。 - **核心理念跃迁**：拒绝将纳税人视为商业化的“客户”，确立“享有权利的数字用户”理念，消除类似肯·洛奇电影中描绘的官僚折磨感。 - **顶尖人才虹吸效应**：使命感驱动的 Gov.uk 项目迅速成为最酷的雇主，在招聘高端软件工程师和技术黑客上反超私营企业。 --- ### `57:00` 主持人复盘：国家核心能力、使命导向与 AI 双刃剑 在马祖卡托教授离席后，主持人 Tracy Alloway 与 Joe Weisenthal 在马德里现场展开复盘讨论。他们一致赞同马祖卡托对“外包”的核心洞察：外包本身并非原罪，宏大项目必然需要私营部门的大规模参与，但其前提是公共部门必须具备强大的“内部肌肉”（internal muscle）——懂得该在何时与谁沟通、如何评估项目成效。如果一味依赖外部咨询公司（如麦肯锡）去定义问题与解决路径，公共机构便会彻底丧失核心能力与协同活力。 随后，主持人将话题延伸至当前最受瞩目的人工智能（AI）领域。AI 作为一项通用技术，表面上能够嵌入到医疗改善、新药研发、效率提升等公共使命之中，但它与过往技术的不同之处在于其自身巨大的风险与负外部性。科技界本身正陷入对 AI 潜在反噬的极端焦虑中——不仅包括急剧扩大的社会不平等，还有模型能力提升后可能引发的对齐失效与失控风险。因此，AI 很难仅仅被当作服务于其他使命的被动工具，围绕“AI 的发展应走向何方”与“如何建立有效规制”，本身就需要确立独立而严谨的技术使命与治理框架。 - **外包的底线在于内生能力**：政府不能当“甩手掌柜”，必须建立评估绩效和识别合作伙伴的内部专业度，而非事事外包给麦肯锡。 - **AI 技术的特殊悖论**：既能作为辅助实现健康、环保等使命的高效工具，又因自身可能加剧不平等、产生对齐风险而必须成为被专门治理和引导的独立使命目标。 ### `01:00:00` AI 监管诉求的真伪与系统性隐忧 在节目的收尾讨论中，主持人探讨了科技巨头高管呼吁全社会共同商讨人工智能应对策略的真实动机。有人质疑这只是一种营销手段——通过宣称 AI 的潜在市场规模（TAM）涵盖全人类劳动乃至人类生活本身，以此推高企业估值。但主持人认为，这种做法不仅引发了极大的公众反弹，让许多人深感不安，而且从效果上看并不利于公关。此外，许多前沿人工智能实验室从成立第一天起，就建立在“AI 不是普通技术”的认知之上，甚至特意采用了非营利组织等特殊治理架构。因此，科技巨头 CEO 们对技术失控的担忧在很大程度上是真诚的，他们自身确实感到焦虑。 更深层的担忧在于，如果没有政府的战略干预或顶层设计，单纯部署 AI 根本无法解决现有的结构性问题，反而会变成“在边缘修修补补”并导致局面恶化。最具代表性的场景便是美国医疗计费系统：未来可能演变成保险公司的 AI 机器人与医院的 AI 机器人互相对轰。双方都宣称自己在简化医疗计费流程，但由于没有人去真正改革医疗账单的运作逻辑以及付费机制，最终只会陷入“机器人互搏”（bots fighting bots）的怪圈，不仅没有人类能从中受益，反而使系统愈发繁琐复杂，彻底沦为“无止境的机器人套娃”（bots all the way down）。 ```mermaid flowchart TD accTitle: AI修补式应用与根本性改革的对比 accDescr: 对比缺乏政府战略干预的AI自动化与根本性系统改革的不同结果。 subgraph EdgeFix["缺乏战略干预：边缘修修补补"] A["底层结构性顽疾（如复杂医疗账单）"] --> B["医院部署 AI Bot 索赔"] A --> C["保险公司部署 AI Bot 拒赔/审核"] B <-->|"算法互博与恶性消耗"| C B & C --> D["系统复杂度激增，无人类受益"] end subgraph SystemReform["顶层设计与战略改革"] E["政府制定使命战略与规则重构"] --> F["厘清付费模式与根本利益"] F --> G["AI 用于实质性提效与公共价值"] end ``` - **科技高管的真正焦虑**：AI 实验室早期的非营利架构及其面临的社会抵制表明，科技领袖呼吁监管并非单纯的商业估值炒作。 - **系统性风险的反思**：将 AI 叠加在扭曲失效的旧系统上，只会制造更多技术寻租与对抗，而非真正的生产力跃升。 ### `01:01:21` 节目结语与鸣谢 讨论结束后，主持人 Tracy Alloway 和 Joe Weisenthal 进行了节目的例行致谢与收尾。他们公布了各自主播及嘉宾 Mariana Mazzucato 的社交平台账号，并向幕后制作团队成员表达了谢意。同时，节目呼吁听众加入 Discord 社区参与日常讨论、订阅每日简报，并在播客平台留下好评，尤其是“如果听众喜欢节目对‘国家能力’（state capacity）展开的深入探讨”。最后，节目在关于 Bloomberg 无广告收听指南以及赞助商信息（包括企业定制平台 4imprint、商业保险机构 The Hartford 以及壳牌加油奖励计划 Shell Fuel Rewards）中正式落幕。 - **社交媒体与制作团队**： - 主持人：Tracy Alloway（@TracyAlloway）、Joe Weisenthal（@TheStalwart） - 嘉宾：Mariana Mazzucato（@MazzucatoM） - 制作团队：Carmen Rodriguez（@CarmenArmand）、Dashiell Bennett（@Dashpot）、Kale Brooks（@KaleBrooks）、Kevin Lozano（@KevinLloydLozano） - **官方资源与互动**：彭博《Odd Lots》每日简报（bloomberg.com/odds）、Discord 讨论社区（discord.gg/oddlots），以及 Apple Podcasts 彭博无广告订阅频道。 ## Notable quotes > `06:30` 如果政府在决策时永远只盲目依赖成本效益分析和净现值等静态测算工具，那么在 20 世纪 60 年代，人类根本连“登月”的念头都不会动。 — Mariana Mazzucato > `08:00` 嘴上说着杰斐逊式的辞令，背地里走的却是汉密尔顿式路线。 — Mariana Mazzucato > `10:00` 番茄酱是一种蔬菜 — Mariana Mazzucato > `16:08` 这就像看了一辈子心理医生；如果你的心理医生看了你一辈子都没让你好转，那他大概算不上一个好医生。 — Joe Weisenthal > `22:15` 我们选择做这些事，不是因为它们容易，而是因为它们艰难 — Mariana Mazzucato > `25:00` 没有人会因为买了 IBM 的产品而被解雇 — Joe Weisenthal > `25:00` 凭 PPT 统治 — Mariana Mazzucato > `30:00` 你以为你在免费搜索 Google，其实是 Google 在免费搜索你。 — Mariana Mazzucato > `40:00` 计算机无处不在，唯独没有出现在生产力统计数据中。 — Mariana Mazzucato > `54:30` 外面的人不是客户或买家，而是享有基本人权的用户 — Mariana Mazzucato > `01:00:00` 机器人互搏 — Joe Weisenthal ## Mentioned Mariana Mazzucato, Joe Weisenthal, Tracy Alloway, Carlos Cuerpo, Ronald Reagan, Eric Adams, Ernest Brackett, Shoshana Zuboff, Tim O'Reilly, Robert Solow, Eric Schmidt, Nandan Nilekani, Ada Colau, Sam Altman, Ken Loach, UCL Institute for Innovation and Public Purpose (IIPP), Bloomberg CityLab, Global Council for a Common Good Economy, Deloitte, McKinsey, PwC, Eskom, NASA, DARPA, NSF, CORFO, Uber, Carnegie Mellon University (CMU), Google, Oxford University, AstraZeneca, Goldman Sachs, Omidyar Network, OpenAI, Infosys, SITRA, MindLab, Vinnova, BBC, BNDES, KfW, Government Digital Service (GDS), Serco, 《创新型国家》 (The Entrepreneurial State), 《登月任务经济学》 (Mission Economy), 《大骗局》 (The Big Con), 《监视资本主义时代》 (The Age of Surveillance Capitalism), Gov.uk, ChatGPT ## Video description <p>Today's guest Mariana Mazzucato is one of our most requested. Mazzucato, a professor of economics at University College London and the founding director of its Institute for Innovation and Public Purpose, specializes in the political economy of technological development and public sector investment. In our conversation, recorded in Madrid while at the Bloomberg CityLab conference, she explains her concept of the "mission economy," her definition of state capacity, how to prevent top talent from fleeing to the private sector, and whether consultants or governments should be blamed for inefficiencies and civic failures. It's a wide-ranging interview, one that covers everything from the initial public financing of Silicon Valley algorithms to the history of moonshots.<br><br><a href="http://bloomberg.com/subscriptions/oddlots" data-link-cookie="1" data-destination="http://bloomberg.com/subscriptions/oddlots">Subscribe to the Odd Lots Newsletter</a><br>Join the conversation: <a href="https://discord.gg/oddlots" data-destination="https://discord.gg/oddlots">discord.gg/oddlots</a></p><p>See <a href="https://omnystudio.com/listener">omnystudio.com/listener</a> for privacy information.</p> [Source](https://podtrac.com/pts/redirect.mp3/tracking.swap.fm/track/UVBrz8bN8aM2Xe47PEPu/traffic.omny.fm/d/clips/e73c998e-6e60-432f-8610-ae210140c5b1/8a94442e-5a74-4fa2-8b8d-ae27003a8d6b/258befee-0a77-412f-bc07-b44300b368b2/audio.mp3?in_playlist=982f5071-765c-403d-969d-ae27003a8d83)

[6:30] Mariana Mazzucato: 如果政府在决策时永远只盲目依赖成本效益分析和净现值等静态测算工具，那么在 20 世纪 60 年代，人类根本连“登月”的念头都不会动。 [8:00] Mariana Mazzucato: 嘴上说着杰斐逊式的辞令，背地里走的却是汉密尔顿式路线。 [10:00] Mariana Mazzucato: 番茄酱是一种蔬菜 [16:08] Joe Weisenthal: 这就像看了一辈子心理医生；如果你的心理医生看了你一辈子都没让你好转，那他大概算不上一个好医生。 [22:15] Mariana Mazzucato: 我们选择做这些事，不是因为它们容易，而是因为它们艰难 [25:00] Joe Weisenthal: 没有人会因为买了 IBM 的产品而被解雇 [25:00] Mariana Mazzucato: 凭 PPT 统治 [30:00] Mariana Mazzucato: 你以为你在免费搜索 Google，其实是 Google 在免费搜索你。 [40:00] Mariana Mazzucato: 计算机无处不在，唯独没有出现在生产力统计数据中。 [54:30] Mariana Mazzucato: 外面的人不是客户或买家，而是享有基本人权的用户 [60:00] Joe Weisenthal: 机器人互搏

[8:12] Mariana Mazzucato: Stop focusing on sectors, technologies, and firms, focus on big problems... as bold as going to the moon and back in a short amount of time. [20:00] JFK: because it is hard, not because it is easy [25:00] Ernest Brackett: captured by brochuremanship [30:00] Shoshana Zuboff: You think you're searching Google for free, they're searching you for free. [40:00] Robert Solow: you can see the computer age everywhere but in the productivity statistics [40:00] Nandan Nilekani: no matter how many apps you have, we will continue to have misery. [54:39] Mariana Mazzucato: those are not clients and customers. Let's stop talking about people as clients and customers. They're users with human rights and how they will access their driver's license, their passport, their voter registration has to enhance their souls and not make them want to die.

[0:00] Whatever your goal, trade show giveaways, client gifts, or team gear, 4imprint has the promo products to match. With thousands of options, from apparel and drinkware to tech and totes, it's easy to find the right fit for your brand and budget, with standout choices at every price point. And with their 360-degree guarantee, you can be 4imprint certain your order will show up just right, right on time. Explore more at 4imprint.com. 4imprint. 4certain. The thing about AI for business? It may not automatically fit the way your business works. At IBM, we've seen this firsthand. But by embedding AI across HR, IT, and procurement processes, we've reduced costs by millions, slash repetitive tasks, and freed thousands of hours for strategic work. Now we're helping companies get smarter by putting AI where it actually pays off, deep in the work that moves the business. Let's create smarter business, IBM. When you're running a business, The best days are the ones where priorities stay on track. For midsize and large companies, that isn't always easy. Risk can touch multiple parts of an organization at the same time often in ways that aren immediately obvious It might involve property liability or cyber It could stem from regulatory requirements or challenges tied to a specific industry or the scale of an operation At that level managing risk becomes an ongoing discipline not a one decision At The Hartford, the focus is on helping businesses manage risk before it turns into something more disruptive. That means working with companies to identify where they're exposed, decide what matters most, and put practical standards in place so risk is managed as part of day-to-day operations. And when losses do happen, The Hartford can pair that risk control work with insurance coverage grounded in underwriting, risk engineering, and claims experience developed over time. Learn more at the Hartford dot com slash risk mitigation. Bloomberg Audio Studios, podcasts, radio news. Hello and welcome to another episode of the Odd Lots podcast. I'm Jill Weisenthal. And I'm Tracy Allaway. We're still here in Madrid, Tracy. How are you? You have a good time? I am. I've eaten a lot of ham and cheese That pretty much all I can say Yeah I going to turn into a jam by the time I leave I certain of that So we are at the Bloomberg City Lab conference You know it funny, like the mayoral level of politics, not something we spend a ton of time typically on. But I would say like it definitely, you know, when I think about it, when I'm like here and like listening to a lot of the conversations, it's just so obviously like connected to a lot of the themes we talk about, because so much of our discussions have to do with something related to, you know, innovation or technology or implementation of policy and how it spans both the public and private sectors. Yeah. I mean, also, when I think about a lot of AdLot's topics like AI or housing affordability or inequality, like cities, I think CityLab actually used this phrase at one point, but cities are really at the front line of all of those challenges, right? And trying to implement policy in that local level in a way that's like, very easy to see and observe and also to judge. Totally. Yeah, no, it's exactly right. And I think maybe there's something like, you know, I don't, it feels like to some extent, the mayoral level of any governance is sort of, maybe the least ideological and the most like you know we talked to the Baltimore mayor for example and so much of his theme was just like talking to other mayors like what working in your town what working in your city and what not working and so forth They all have such similar challenges So many things have been like repeated you know one time after another that like they all can sort of speak the same language and all have the same issues. Yeah, it's funny. You kind of get that local idea swapping at a mayoral level that I cannot imagine necessarily happening in national politics. Like, can you imagine Trump and Xi Jinping getting together and be like, oh, we implemented this really cool national transportation program. Have you tried it? Yeah. You know, you hear these stories, right, of a governor sometimes or a mayor will go to another country. It's like, oh, we can learn from what the city did on transit or whatever. But no, you don't really hear that the same way at the real national level, which is where a lot of our discussions tend to sit. Absolutely. Anyway, I'm really excited about today's episode. We're going to be speaking with a guest who I would say, since we've been doing OddLots, is actually one of the more frequently requested long requested and so it's sort of a failure on our part that like we just like never made it happen before but yeah someone who like really like whose whole career is like dedicated to a lot of odd lotsy things well we waited until we could do it at madrid just like do it at any other random venue yes it was all very strategic [5:00] designed. So we really do have the perfect guest, someone that a lot of guests have wanted to listen to, hear from for a long time. We're going to be speaking with Mariana Mazzucato, professor at University College London, founder of the Institute for Public Purpose, author of several books sort of touching on these themes of technology, public sector, private sector, the roots of innovation, how these things actually get deployed. So Professor Mazzucato, thank you so much for coming on Adla. It's thrilled to finally have you on. Finally. Finally. Completely our fault for not having made it happen sooner. But what are you doing here? What brings you to this particular conference? What attracts you here? Well, first of all, I'm here also because the government of Spain is really kind of leading the way in many areas that I'm very interested in, especially the kind of new economic thinking that needs to underpin how we rethink government. And so I was meeting with both the Prime Minister and the Minister for the Economy, Carlos Cuerpo, who I'm on a panel today with, about a council that we've just set up. It's called the Global Council for a Common Good Economy. Anyway and besides that I also here to speak at the Bloomberg City Labs event about this new public sector capability index that we been developing with them as a partner which is really about reinvesting inside the civil service so that they can really tackle those wicked complex problems instead of you know hiring Deloitte during COVID and giving them 1 million a day to do test and trace which they failed at in the UK I definitely want to get you to go off on consultants. But before we do, like when you meet with someone like the prime minister of Spain, what is it that they want to know from you? Like what information or expertise are they seeking? Right. Well, I'm actually quite lucky that my books actually get read by the prime minister. So usually what happens is that they've, especially the entrepreneurial state, which I wrote in 2013, that had quite a bit of an effect here. They even wrote a report called El Estado Emprendedor en España. And they really want to ask me, you know, what does that mean? What does it mean for actually even being able to fail, for example, right? So venture capitalists are always bragging about all the failures that they had in order to get a success, you know, whereas as soon as a civil servant or a minister fails or a prime minister fails, fails, front page of the papers. So they're very interested in the kind of narrative change, but also the cultural change, but especially the theoretical underpinning within, say, a finance ministry or an economics ministry that needs to then accompany rethinking government Because otherwise if you continue to have the old economic models where we judge things by cost benefit analysis net present value all these really static metrics we would have never even bothered going to the moon Okay first of all in the 1960s had we thought of it as a cost benefit calculation. And so what should replace that? Right. So you mentioned the entrepreneurial state being published in 2013. And I feel like I need to emphasize this because like, there has been this mind shift since the 2020 pandemic on industrial policy. And we kind of take it almost as a given, especially on this show, that there is a role for governments to play when it comes to innovation and entrepreneurialism. But you were there very early. Did you feel vindicated by the 2020 shift? Well, I think there's a bit of a misunderstanding that just because government spending a lot of money, investing a lot of money, whether it's the IRA or the next gen EU, two trillion recovery plan in Europe, that that means that industrial strategy is back or that, you know, government is being really strategic. It depends. And I mean, that's the whole point, right? You know, how are we actually framing it? Is it just a lot of tax incentives? Are we focused too much on sectors? To what degree is industrial strategy actually helping to create a more inclusive sustainable innovation driven you know economy Or is it just another wave of handouts and subsidies to particular sectors And then we end up socializing risks and privatizing rewards And also I mean I used to joke that the U government has always had an industrial strategy but you know pretended not to And they've always talked the Jefferson talk, but acted the Hamilton talk or the Hamiltonian kind of more, you know, proactive government strategy. But by pretending that wasn't there, they kept a lot of things under the radar. and the joke bit is just that finally people understood what the hell it was talking about because of the musical before the musical came out anyway so i think that's really important to recognize that industrial strategy has always been there yeah except that in certain phases including now in some countries it's not really strategy in terms of being driven by public purpose which as you said is the title in this institute for innovation and public purpose i direct, it's been just kind of vertical strategies, focus on sectors, technologies, types of firms, right? The whole focus on small, medium enterprises. So what I've been trying to do is to say, stop focusing on sectors, technologies, and firms, focus on big problems, right? As bold as going to the moon and back at a short amount of time. This is the mission idea. The missions that then require sectoral support, but you're not getting support because you are a particular sector. You're getting support if you're willing. So moving forward, [10:00] away from picking winners to picking the willing, if you're willing to work with government around these very difficult challenges, which could be as, I don't want to say simple, but as concrete as making sure that every child in a country has healthy, tasty, sustainable school lunch, right? Not just school lunch. That was the Reagan thing where he said ketchup is a vegetable so we can reduce the cost. We all had t-shirts of Reagan and ketchup, which was the vegetable is what the t-shirt said. So having moonshots, even on something as simple as the cafeteria, that then requires innovation across many different areas. It requires government to wake up and not just have that be done, for example, from the Department of Education, Department of Agriculture, Health, Education, Finance would work together on a moonshot of healthy, tasty, sustainable school lunch. So that inter-ministerial coordination, which we saw during COVID, right? So we had the war room, the situation room, the military, education, health, and so on together because we had these really difficult challenges. But as soon as COVID stops, we go back to very siloed ways of thinking. As someone with two kids in the New York City public schools, I can absolutely confirm that the degree to which they associate the mayor with school lunch policy is extremely real And they like talk about like oh we used to have these like really nice waffles or something And then Eric Adams got rid of them Or so like, they're really like keyed in. Single issue voters on waffles. The real like single issue voters. They're like really pay attention to the degree to which school lunch policy shifts with administration. Like this is like, they're really, they really think, what is state capacity? We've been using this term for years on the podcast. And we were like, during COVID, We're like, you know, it took a took, you know, a long time to stand up testing facilities for sense on people's lack state capacity. But what is it? What's the definition? So we actually distinguish between the word capacity. OK. Routines, kind of administrative routines and then capabilities. OK. So what states are often what governments are often lacking are those kind of dynamic capabilities. So capacity partly is like literally do you even have fiscal space? Do you have a budget? But also how are you thinking about that budget in terms of do you put it this way Capacity is you know number of people working in your administration the budget that been allocated perhaps also the training you know that the civil service actually has But capabilities are what you actually then do with it Are you agile Are you flexible Are you able to pivot during COVID and actually start working in this more again interministerial way Do you know how to work with others Right You know Do you set up good partnerships or are they problematic partnerships And also you were talking before about what you learning here which is fantastic about how mayors can learn from each other, right? Have you invested in your ability to learn, to adapt? And so I think that ladder, this concept of dynamic capabilities is a much more complex area to be investing in. And the only reason you would do it is if you actually have a theory about government that is more than what traditional economists think about government, which is at best, well, at worst, get out of the way, at best, fix the market failure. So as soon as you say, actually, it's about shaping and co-creating a different type of economy and society that works for people and planet, then the question is, what does that mean for the capabilities that you need? If you're just fixing, then you just need a lot of bandages. And that's, in fact, what we get. We get very reactive, kind of filling the gap kind of policies. Yeah, the pendulum constantly swinging back and forth. Exactly But I mean capacity of course is essential without you know a budget and fiscal space You can do nothing That third category that I mentioned quickly administrative routines That you know are you also you know do you have a stable environment where you can learn by doing Because if you constantly changing what you doing it going to be hard to have a learning by doing dynamics So those kind of administrative routines I even see this in my university where as soon as you get a lot of turnover even those kind of basic routines aren there But capability So these three areas capacity routines and capabilities are equally important but the capabilities are really what I find are lacking. And it goes back, as I was mentioning before, to the underlying economic dogma that has underpinned the way that we think about policy, government at different levels, that by design, not by coincidence, is reactive. Thank you. for any audience, purpose, or budget. You can customize it all, your logo, your message, your look, and many items come with no setup charge to help you save. And if you're really watching the bottom line, you'll find standout choices at every price point so you can make a real impact while staying on budget. Plus, you'll get expert help, fast turnaround time, [15:00] and their 360-degree guarantee. So you can be 4imprint certain your order will arrive on time and look exactly right. Whatever your goal, 4imprint makes it easy to find your perfect promo match. Explore the possibilities today at 4imprint.com. 4imprint. 4certain. Support for this show comes from public.com. If you're actively involved in your portfolio, you probably catch yourself repeating the same actions. Buying the dip, manually sweeping idle cash, putting on a hedge. On public, you can now create AI agents that handle all these tasks on your behalf. Just describe what you want to do in plain English. Like, if the VIX hits 25, buy a put option on the S&P 500. Or, if my cash balance goes above $20,000, move the excess into my direct index. You approve of the workflow and your agent handles the rest. Monitoring the market, watching for your conditions, and executing your strategies exactly as defined. an investing platform driven by your intent not just your clicks you can also get full read and write access to your account via the public api go to public.com slash market and fund your account in five minutes or less that public slash market paid for by public investing brokered services by Open to the Public Investing Inc member FINRA and SIPC Advisory services by Public Advisors LLC SEC Registered Advisor Complete disclosures available at public.com slash disclosures. The thing about AI for business, it may not automatically fit the way your business works. At IBM, we've seen this firsthand. But by embedding AI across HR, IT, and procurement processes, we've reduced costs by millions, slash repetitive tasks, and freed thousands of hours for strategic work. Now we're helping companies get smarter by putting AI where it actually pays off, deep in the work that moves the business. Let's create smarter business, IBM. Have consultants become a substitute for state capacity, slash routine, slash administrative ability? Right. I always wonder how we got to the point where consultants are so big anyway, because it feels like every time you hire a consultant, it's almost an admission of failure on your part to be able to do something right. And nevertheless, it's a widely accepted practice across governments. So my last book actually was called The Big Con. So Con for Consulting. And the subtitle was how we weakened businesses infantilized our governments and warped our economies basically due to this consultification I wouldn blame the consultants I mean I actually blame governments right? Like, why are you opening the door so wide to consultants? And it's fine to have advisors and some consultants. The problem is when they are actually doing the core tasks that government should do. Again, test and trace during COVID was a core task. And so I think it really stems from, I'd say the 80s when we you know the kind of Reagan Thatcher years if you want. It always goes back to Reagan. Yeah but actually in the UK for example it even increased more during the labor government so it's not it's not you know one party but it did begin I'd say in the 80s with this kind of downsizing of governments which then ironically cost them more because as soon as you start downsizing without really strategically thinking what you need and what you don't need of course you should trim the fat. There's no reason to have a bloated government structure. But when it's done for ideological reasons and not strategic reasons, then, ironically, then you end up not having those capabilities that you need as soon as you have a flood or Brexit or COVID. And so I think then what happens with the consultants is it not their fault that they invited in I do think it very problematic what they end up doing once they in So there huge conflicts of interest Well the biggest conflict of interest being that they have no incentive really to make government better later because they wouldn have no contracts. It'd be like having a therapist your whole life. That therapist is probably not very good. Which is how the therapy model works. Well, do you remember this came up in the episode we did about construction in New York and this idea that like one of the reasons it's so expensive and take so long for like public funded projects is because consultants have no incentive to actually get the project done. Yes. But also they're often working on both sides of the street, right? So there'll be, for example, consulting for, I don't know, a state-owned enterprise like ESCOM in South Africa, as well as the Treasury, which should be regulating ESCOM. Or in Australia, there was a famous case with PwC where they were consulting for a medical device company, as well as the regulators of the medical device companies. Like, come on. So that should just be illegal, right? And again, you know, getting the right kind of regulation that makes sure that we don't have these kind of scams. So what we also argued was, you know, the first thing is start investing back inside government. So you don't need so much consulting, but also when you do bring in the consultants, make sure the contracts actually embed learning within them and that you are also bringing in the right people. You know, if you have an oncology strategy, of course, you [20:00] get the top doctors and consultants and cancer to advise. So the other huge problem is that these consultants, when they're coming in, they often actually don't know very much. And they end up really bothering the poor public servants that they end up emailing. Oh, would you mind telling me what you think about or sending me your plan that we can study? It's like, why are you even working with government if you don't have within the consulting companies that deep expertise, which, you know, I'm not saying this just because I'm an academic, but I don't think academics are used enough. You know, if you have a research center that's been thinking about climate change for the last 40 years, use them. Don't ask McKinsey, as Australia did, to design your climate strategy, which ended up, by the way, being terrible. So there is a bit of why is it that governments, A, don't invest in their own capacity and capabilities, and B, when they do go out there and look for the advisors and consultants, are kind of getting the ones that simply kind of provide a rubber stamp, right, that makes them feel more secure. They haven't even done the homework to make sure they are getting the top people in the world to help advise them on doing difficult tasks. Just to take the other side of the coin. So recently in New York City, it was, I don't know I think scandal is too strong a word but there were a lot of headlines about how much the city had paid consultants for working on redesigning the trash collection system in the city And on the one hand it a trash collection core city function Why do you have to bring in a consultant? On the other hand, reimagining the trash collection system is like that's a one time thing. So maintaining it, implementing it, servicing it. OK, that's like a permanent government function. But the actual like, OK, we have to do a redesign. Yeah. That is that necessarily something that like we need to have in house in government, the capacity, because do you want to permanently have that muscle? Because that's a one shot thing. Doesn't that like to me, that makes sense as a time to bring in a consult. People also argue for something like that. There's a value to having an external viewpoint, right? That makes sense. Like the redesign, that's a one time job. Like to me, I was like, all right, that doesn't seem crazy to bring in a third party to like help figure out what that plan is. Exactly. So that's the myth, right? That somehow we're talking about either government does everything, right? Or even nationalize everything, or it does nothing and it privatizes and brings in the consultants. So the truth is obviously somewhere in the middle. So of course, you're absolutely right. Government doesn't have to have all those skills. It definitely needs the skills to know who to work with outside of government but it also needs to even understand that kind of outside landscape to even think about what might we need how might we start developing a strategy that reimagines, say, the trash collection process. So that's why I talk about missions. So when, you know, NASA wanted to go to the moon and back in a short amount of time, they didn't say we're going to do it all by ourselves. And they also didn't say we're just going to do it with the aerospace sector. Right. They said we're going to have to work with so many different private sector people. They ended up working with something like 400,000 people in the private sector. They said, we have a lot of problems, but we don't know the solutions. But we're going to set very clearly a direction for working with the private sector in a problem-oriented way. So the first thing they did was change procurement. Procurement, you know, government purchasing is often like 30% of a government's budget. It's a very important part of their budget, whether it's Barbados, a small island state, or the US, a very large government. Procurement is there. How are we using it. So they realized they had the wrong type of procurement. It was just, again, minimizing costs. It was cost plus procurement. They changed it to outcomes oriented procurement. And they started to ask themselves, what are the outcomes that we need? We need to figure out how are the astronauts going to go to the bathroom? Right. Which, by the way, was just a problem again with Artemis The toilet was always a problem up in space What are they going to eat What are they going to wear How will we communicate with them and it was the solutions to those problems that happened within mainly not only private sector institutions with nasa's also kind of leading investment but especially leading kind of thought process of what the problems were okay that ended up getting us camera phones foil blankets home installation software yeah so many different innovations across many different sectors aerospace nutrition materials electronics. That itself is what we're talking about, right? So whether it's going to the moon, whether it's trash collection, whether it's school meals, whether it's getting prepared for the next pandemic, which unfortunately the science tells us will happen, how are we even thinking within government in a problem-oriented way, a solutions-oriented way? And this, by the way, is why the CityLab conference is so wonderful and Bloomberg's government innovation team is so important for so many cities is because they then share their experiences of solving problems. And then they ask, and we're trying to help them do this with this public sector capability index. What did we learn along the way that we were missing? Where were the bottlenecks? What can we do better? But especially in terms of that flexibility, adaptability, willingness to experiment. Right. Remember what Kennedy said, we're doing it because it's hard, not because it's easy. Yet all the words and policy papers are about making things easier, facilitating. I'm Italian. [25:00] Fácila. We're in Spain. Fácila. Right? So if you're facilitating someone, it's not going to be a good contract. If you're de-risking someone, it's not going to be a good contract. If you're simply enabling, facilitating, fixing, it's going to be a very bad public-private relationship. Just going back to consultants for a second, and I think this is actually relevant to the discussion of having cooperation among different parts of the government on big projects, but how much of the consultancy fetish just has to do with diffusing responsibilities? So I always think back to the old saying about the purchasing manager thing. You'll never get fired for buying IBM, right? Like how much of it is just like, well, you know, I did my best. I hired McKinsey. What more can I do if it goes wrong? It's McKinsey's fault. Accountability. Yeah, absolutely. So that's a really important point because, you know, one problem is when government doesn't have those capabilities for the reasons we said before. Another is even when they have it, why are they not using it? Australia, again, is an interesting example because they had really interesting capability within government with their innovation agency, CSIRO. And yet they gave this massive contract to McKinsey to do their climate strategy. And it absolutely about diffusing that responsibility but also because of the culture we have again within government where if they do make mistakes unlike in the entrepreneurial ecosystems and VC and so on we don accept that But it also in the private sector right So we also talk in the book about the consultification of management, basically. And there as well, if you're going to be doing a merger or downsizing or a massive share buyback scheme, isn't it great if you have, you know, McKinsey told us to do it. So also just not taking on, I mean, it's kind of cowardly, Right. Like you're not kind of owning your decisions. I do think it's different in government. I think that, you know, changing that culture, having more gov labs like they have in Chile, Laboratorio de Gobierno, everything sounds better in Spanish. You can say that in Spanish. I always say that in Italian, every time we have a reforma de la publica administración, it's public sector reform, it's just cuts. But it sounds nice. It's literally just cuts. And so that idea that what we need is kind of a laboratory within government, but also between governments, I think is really important. By the way, the head of procurement in NASA, Ernest Brackett, in the 60s with the Apollo program, not only did he help change the procurement policy of NASA, but he also said, we got to watch out. There's too many consultants in these corridors. and his exact quote was if this continues we will get captured by brochuremanship which is kind of endearing because they didn have PowerPoints at the time right So now the idea that you ruling by PowerPoint and that basically all they know how to do At the time it was kind of shiny brochures. But he didn't say we don't want to work with the private sector, right? He said, we won't know how to work with the private sector. We won't know how to write the terms of reference if our own brains are becoming weak. So investing within in order to work also outside with us, others. So it's not working with others. It's working smartly with them. So this gets to the sort of one of the big questions of the day, obviously, and that is AI, right? And so you mentioned, okay, school lunches, you said a mission. It doesn't mean like the private sector isn't going to play a role in providing food or whatever, but first you establish what the mission is. Then you talk about the moon. It doesn't mean that the private sector isn't going to play a role. Private sector played a significant role through various technologies and procurement, but there was an overall mission. Well, how are you thinking about AI? Should governments first decide, like, what is our mission for this? What what missions could it theoretically enable? Like, does this feel different than other endeavors? We obviously know a lot of public sector money is going to wind its way up in AI and already has a lot of it already has But how are you Does this fit Is this different Talk to us about how you think about this particular moment of extreme sort of technological ambition Right I mean there so much to say First, the investments that went into what we call today artificial intelligence, including, you know, LLM models, language models, speech recognition dates, you know, decades and like so many other areas that was led by government. So if you look at even DARPA, which, as you know, was the lead investor in the internet, it came from problems mainly in the military industrial complex that then required ultimately what we're calling artificial intelligence today. But again, in the entrepreneurial state, I talked about how everything in our smartphones that make them smart are not stupid, internet, GPS, touch, Siri, and so on, were government finance. What's very scary today, what makes today different with AI is that that's not really necessarily going to be true much longer. Why? Because these massive economic rents, and I call them rents, not profits, so excess profits and excess of what these companies actually did, because we privatized all the rewards from this massive social and collectively created value in this area. They have so much money, right? Trillions, not billions, trillions. The salaries they [30:00] paying to the top researchers in universities, both public and private universities, and to people who used to work in the NASA's, the DARPA's, the COFCO's in Chile, they are going now to work in these companies. And that hemorrhaging of talent, of top research expertise, I don't think people are talking about this enough. It literally is the biggest change, right? Because otherwise, just, you know, the fact that we've had, you know, big technological changes, general purpose technology is completely affecting you know how we produce how we distribute and you know from the rise of electrification the thing like at one point i mean i don't think it actually panned out that great for them but didn't uber hire the entire carnegie melon robotics team i think they did i think when when they were doing self-driving car technology they just wrote a check for the entire faculty of the cmu yeah robotics team and it was just but not enough people are talking about this. So it's very hard to govern a process for good, ethically and so on, if you don't understand it. So if this talent is in fact leaving these publicly financed institutions, even private universities most of the research as you know has been funded by NSF and so on Once they leave and the knowledge is so concentrated in these few AI companies I don think we thought about that enough But sorry just the other thing just coming back to your question, which I don't think I answered properly. Yes, of course, government should be thinking because of its, at least a democratically elected government, right? This is different for dictatorships, what we would expect from them. But we would expect governments to make sure that any big change, large kind of opportunities around technological change are done in ways that are, again, good for people and planet, right? But that requires not only that capacity that, as I mentioned, is being decimated, but also a certain type of regulation, which is making sure we even understand the process of, for example, how algorithms are currently being designed and whether we have a situation like Shoshana Zuboff in her great book, Surreal and Capitalism, she says, you think you're searching Google for free, they're searching you for free. You know, that could have been avoided, right, through the design of the algorithms. And given that the algorithms initially, at least were publicly financed, there is this issue of how do we, in a pre-distributive way, instead of kind of exposed with redistribution or regulation or market fixing, think about innovation collaborations that have some of these really important ethical concerns thought about upfront And an example of this would be during COVID where you know the vaccines were not the mission The mission was to allow these vaccines to be produced and available globally given it was a global health pandemic. We were all better off if the world was vaccinated. And yet only one of the vaccines, the one that was the collaboration between Oxford University and AstraZeneca, had that kind of conditionality about what good looks like at the start in terms of how they collaborated. So it was the publicly financed researchers, Oxford is a state school, that put that as a condition with their relationship with AstraZeneca, that they would share the knowledge of it, they would join the patent pool, keep costs and prices low. So more of that. Every time I hear, you know, Tracy, like someone talk about studying AI in college and like their professors, I'm like, that's great, but why are those professors not in the private sector making a hundred times? I mean, many of them are, but I'm always like amazed that given how much money People leave money on the table. Yeah, they're like, what are you doing at a university? You're probably making 10 times the money. But where did that money come from? This is the issue, right? So this money that's being used to pay these very large salaries that you're talking about is very important to recognize that that didn't come about because of, you know, early on this amazing kind of entrepreneurship and all the knowledge was in these companies They gathered that knowledge Now they using it Of course they pushing the frontier Of course they doing research But that idea of also making sure that companies are not earning in excess Right So why did the public purse not benefit in those early days Why are they evading so much tax I mean literally in so many countries paying almost no tax you know also labor exploitation in the case of amazon you know you don't hear it from me you hear it from all sorts of different investigations even during covid by the way that they wouldn't even put the ambulances apparently outside the warehouse i don't know i've read that i'm not sure if it's true but anyway the point is that they're you know what does a good company look like and how does that then affect the returns that they're earning and what is a just return versus these excess returns that now are being used to you know higher end you know you mentioned the black box of the algorithms, which I think is a really interesting point. And we did an episode with the CTO of Goldman Sachs recently. And we asked him like, Goldman Sachs is a highly regulated bank. When you have bank supervisors who go in and want to understand your AI system, what are they actually understanding? Do they like understand the underlying code that's driving the algorithm? And the answer was, well, no, not really. You can't expect them to understand all It's more about having the right controls in place that prevent the algorithm. [35:00] from running amok. And so my question is with AI, where we have a lot of experts, highly paid experts and engineers in the private sector developing all this new stuff versus a shrinking body of government officials, like what should our expectation be about how much they understand this technology? Right. Well, it actually comes back also to that previous point. Do you need all that expertise inside? Or also how do you regulate this in such a way that also stimulates more innovation? Because you don't want to put a cap on the innovation just by over regulating it. This is where, by the way, I disagree with the abundance theorem, which makes it sound like, you know, there's all these opportunities out there and it was regulation and too much planning and too many conditions that kind of hurt that, which I think actually corporate governance and shareholder value maximization has really actually stifled the opportunities that we have today. So they don't have to, of course, understand the algorithms, but they do need to do exactly as you said, like we do with climate, right, where there's climate disclosures, we need to know what it is that we want to be disclosed. In fact, we have a project that we just did, that we just finished with, do you know Tim O'Reilly? Yeah. So it was Tim and myself with a grant from the Amityar Foundation a project we called Algorithmic Rents and how to reduce these rents and how we designing algorithms also through disclosures So we thought about what would be the equivalent of AI related disclosures that could do exactly, as you said, be almost the equivalent of what banks now have to do, but also companies have to do around kind of ESG kind of metrics. And so, I mean, the project is still underway. And what we want is actually to find kind of a coalition of AI companies that would be willing to think about this with us. But that assumes, right, that there's also enough of the science as there has been with climate change, which said this is an urgent problem. And unless we fix it now, we're going to reach a tipping point where there's no coming back. That's really when things started to change around the climate disclosures. We haven't, for some reason, reached that yet, even though there's such a link also, by the way, between AI and climate, like in terms of these data centers guzzling, you know, like for every, you know, chat GPT search you do, apparently it's like using a bottle of water, like the small plastic bottles of water, like that's literally how much we're using. And yet, you know, we're not making that kind of systemic understanding of, you know, AI problems and how they are connected to climate and water problems and so on And that requires again government not just having an indicator like GDP to think about but a dashboard right When you your car if you had one number how much gas you have or how fast you going you would crash it So what does a dashboard look like for government that would allow it to make sure it's on track to making sure that also with the evolution of technology, it's thinking about these kind of more systemic features and not just thinking about it as innovation policy. Thank you. . and their 360 guarantee So you can be 4imprint certain your order will arrive on time and look exactly right Whatever your goal 4imprint makes it easy to find your perfect promo match Explore the possibilities today at 4imprint.com. 4imprint. 4certain. Support for this show comes from public.com. If you're actively involved in your portfolio, you probably catch yourself repeating the same actions. Buying the dip, manually sweeping idle cash, putting on a hedge. On public, you can now create AI agents that handle all these tasks on your behalf. Just describe what you want to do in plain English. Like, if the VIX hits 25, buy a put option on the S&P 500. Or, if my cash balance goes above $20,000, move the excess into my direct index. You approve of the workflow, and your agent handles the rest. Monitoring the market, watching for your conditions, and executing your strategies exactly as defined. an investing platform driven by your intent not just your clicks you can also get full read and write access to your account via the public api go to public.com slash market and fund your account in five minutes or less that's public.com slash market paid for by public investing brokered services by open to the public investing inc member finra and sipc advisory services [40:00] This is by Public Advisors, LLC, SEC Registered Advisor. Complete disclosures available at public.com slash disclosures. So there's a lot of noise about AI, but time's too tight for more promises. So let's talk about results. At IBM, we work with our employees to integrate technology right into the systems they need. Now, a global workforce of 300,000 can use AI to fill their HR questions, resolving 94% of common questions. Not noise. Proof of how we can help companies get smarter by putting AI where it actually pays off. Deep in the work that moves the business. Let's create smarter business. IBM. Just from a pure sort of history of technology standpoint, does AI feel different? I think it definitely fits with the kind of characteristics of a general purpose technology in the sense that it really does kind of change everything. However, there's also a lot of kind of myths around that, kind of a bubble in terms of how we're thinking, not only in terms of the financial market bubble around it, but just like with electricity, it took about 30 years for it to actually really affect how governments were operating. Similarly, with the internet, you'll remember when Robert Solo was saying, you know, there's computers everywhere except in the productivity statistics I think we still very early in the phase with AI in terms of it having a really meaningful impact on improving Again coming back to the point about problems right This ultimately should be helping us solve problems. Which problems is AI really helping us to solve in a systemic way that can scale and that there's learning between governments? We're definitely using AI. Is it actually helping us solve some of the biggest problems of our time? Again, health problems, climate problems. Surely there's some of that. But until we manage this process, until we govern it both ethically, but also in a way that is with government instead of kind of sidetracking government in order to extract these mega rents, then we're going to have a huge problem. Why do you think that hasn't happened yet or governments haven't stepped in faster? Because I do think AI is kind of unusual in the sense that, you know, you have people like Sam Altman who will say very publicly in interviews like, yeah, this creates a lot of negative externalities and society is going to have to figure out how to deal with this. and governments are going to have to figure out how to actually best deploy this technology. OpenAI published a big industrial policy document, which, again, I think is kind of unusual to previous technological developments And yet governments seem kind of slow I don want to generalize too much but many governments seem kind of slow to get in and really start shaping what they want to do with AI Yeah, I mean, I think, again, it seems like I'm saying the same thing, but I really it's because I believe in it. This technology on its own will not solve anything. If it's accompanied by a strong, for example, health system that with AI thinking about health problems, then I can see that being absolutely revolutionary. And in fact, I once heard a really interesting discussion between Nandan Nilekani, you know, this incredible entrepreneur of computer systems. He was one of the co-founders of Infosys in India. And Eric Schmidt at one of these dialogues we have in Bellagio, this beautiful villa in Lake Como run by the Rockefeller Foundation. and we had kind of Eric Schmidt on the one hand saying, you know, in the future, all we're going to really need is an app, a health app, and that's going to be fine. And Nandan Nielkani, who really is one of the biggest innovators around computing, was like, what? Like, let me tell you, in India, without a proper health system, no matter how many apps you have, we will continue to have misery. And so it's not one or the other, but what you need is more people like Nandan who think about what is the relationship between you know the power of AI and the structure of a health system And who thinking about that And if we do have at the same time not just austerity but this you know kind of dumbing down of what we think government is for. So even the health systems we have are not kind of, you know, fit for purpose, then I don't see any sort of future of AI kind of helping us with health problems. And by the way, just look at water. I mean, like 55% of the global food system right now is at risk because of how we're treating the global hydrological cycle. You know, biodiversity loss also in the Amazon is affecting droughts and floods in other parts of the world. That's a huge problem. To what degree are we really using AI to fix that problem? Yeah, not much. So it's also about where we're putting kind of the emphasis. And for that, I think we do need these moonshots, government led, working with the private sector. And again, using the power of AI, well regulated to solve very concrete problems. Yeah, Joe, I was thinking about this specifically related to health care recently because I've seen a bunch of startups that are saying they're going to simplify the billing process for hospitals. And then you also know that there are a bunch of insurance companies that are also using AI. And it's like, well, if we're just going to have more computers debating with computers. Yeah, exactly. And like if the system itself doesn't change, nothing's going to improve. Right. No, I've totally thought about the same thing. [45:00] It just feels like we're going to have this arms race where it's like my bot will argue with your bot. And then the only entities that make any money are the bot makers, the bot makers. Do you know that Ada Colau, who was the mayor of Barcelona, another city here in Spain that you know well, she when she was mayor, she came from a housing movement. So she was really, really concerned with housing issues, but also, you know, public transport, public schools and so on. And her thing was, why is it that when the citizens of Barcelona, you know, like click on Uber or CityMapper, this data that's created from that, right? Because every time we click on something, data is created. She said, why aren't we in the city using that data to improve our decisions and understanding of our public transport and public housing challenges? And so she ended up hiring computer hackers into the city government and made it a really cool place to work. And I think that, again, kind of insourcing back in those kind of, you know, cool hackers that currently are working in these companies, but to come and work with the city administration that says, we want you to come in, fail, but help, you know, like, don't worry about failing. Sorry not fail but you know take risks to help us though kind of really target our big challenges around housing transport and so on inequality in terms of access So obviously we talking about the public sector We talking about the government role in facilitating or guiding various technologies. But there's also just like politics, right? Winning elections. And the fact that, you know, as you mentioned, failures become a scandal, and they're on the paper, and maybe like politicians lose their jobs, then someone comes in. When you just in your own personal work, when you think about this stuff, how much do you have to calculate the reality that one big job of politicians is to win reelection and that a loss of election has the potential to just take the government in 90 or 180 degree turn from whatever the previous administration at any level in any country does? And how much do you think about that reality when you're thinking about strengthening government? So the first time I started to work on missions wasn't on mission-oriented policy. It was on mission-oriented organizations, right? To better understand the DARPA kind of organization, right? Corfo in Chile is similar. Citra in Finland. MindLab in Denmark. Vinova in Sweden. These are innovation agencies that are, in fact, making these big bets that are mission But my question was how are they organized Are they also unstable due to the electoral cycle Yeah And they have thought about this I mean it not a coincidence that DARPA for example people come in for five years but it not the four year electoral cycle They actually told, come in and do take risks. That's how you'll be evaluated, you know, not by just if you're succeeding all the time, that means you're not taking those risks, but also the impact that your successes have. And so that kind of cultural shift, but also the fact that people are coming in kind of on secondment you know they're not there to be a civil service civil servant their whole life so we started studying well for some for some areas i think that's fine especially around innovation right you want to coming back to the idea that we want to bring in and crowd in the top talent into government you know having like a five-year period that you're going to help as a civil servant paid by the government not as a consultant you know working with not at the civil service i do think there's lots of kind of room for that so that idea that because we value the private sector, there has been, for example, think of Harvard Business School, where they have this case study methodology of businesses. We've never really done that with government entities, because we don't value basically government as a value creator. It's just seen as a redistributor, a fixer a facilitator and enabler of the private sector And that of course where then we expect creativity and value to be created which is not right And so one of the things that we do in the Institute which is actually a department so we train up civil servants around the world also you know, through our own MPA, Master's in Public Administration, but also through applied learning programs, was to start developing these cases. You know, what do we know about the BBC? You know, how is it different from other public broadcasters? How do they measure what they call public value. What is public value? So even having, you know, comparison learning between say a public bank, the BBC, a government digital agency on what it means to crowd in or crowd out the private sector, what it means to shape markets, not fix them. What does it mean to have a culture of experimentation versus this huge risk averseness that, as we said before, is a cause for the consultification. So I think a lot about that, but it's not, you know, there's so much instability, obviously also in the private sector. So there is a bit of a myth that it's all and stable in the public sector because of electoral turnover. We can't shape these bureaucracies to be creative bureaucracies, resilient bureaucracies. They don't have to be vertical and so inertial. But the other point I think that's sort of stemming, I think, in your question, tell me if this is not related, is literally winning the election. What are we learning globally? You know, why is it that Biden, whose economic policies were actually quite successful in the red states? [50:00] starting to be quite successful. Why in those states did he not win? And I think there's something going on in a lot of countries, definitely also in Italy and the UK, where people who have been, let's just use the concept, left behind on terms of the economic benefits, at least in the past, even when new economic policies work, that's not enough if people don't feel valued, if they don't have their dignity back, if they continue to feel condescended upon. So one of the really cool things I've been working on with city governments, but even councils. So my neighborhood in London is called Camden. It's about 250,000 people. I worked with the council on mission-oriented procurement for adult social care across 10 housing estates, what you call projects in the US. And we brought the carers and the caries to the table to design that policy. So working with people, really valuing their lived experience to help design policies that are meaningful and will improve their lives. I think it's just so important, firstly, to get those policies to be designed right, but also to give people, again, dignity and self-worth. And I've seen it also, you know, because we have so much inequality in the UK Unfortunately we have food banks which is barbaric if you think about it in the 21st century food banks like you know we should not have that People should have food you know on the table enough and healthy food We don have that So transforming food banks into food cooperatives green food cooperatives, where the people benefiting from what was a food bank are now also in the place of governing, of having real deliberation, of thinking together. I can tell you the people I saw working in the food banks who are also receiving the food, the facial expression, the dignity, just even how people are standing is completely different from a system where, you know, you hear someone's expired. Yeah, exactly. From last Thanksgiving. No, but even if it's good food, you know, having again, you know, bringing back dignity and value. It's so important, I think, to fight populism. I'm going to ask what is potentially an unfair and loaded question, but I think it might be quite illustrative of everything that we've been discussing. When you look across the world, are there particular countries or cities that you think are doing industrial policy right in the sense that they're taking maybe a holistic approach with a defined strategy slash mission? So I tend to also look at like very specific things that a government did instead of just saying the whole government's perfect. Right So for example in Brazil something they done that I think has been very positive is that they put what I call missions at the center of government So the ecological transition is at the center And that then required the Department of Finance for example to rethink its own tools For example public bank right So BNDS which is one of the largest public banks in the world it can either just again give money out to say the agrobusiness industry or save the steel industry when it going bust Or because there's an ecological transition, it can think about how these sectors themselves need to change in order also to access the loans that the bank is giving. Germany, by the way, did that when they had the Energiewende policy, their public bank, the KFW, the way they provided support to the steel sector, which in the US and the UK and so many countries, steel is under pressure. the loans to the steel sector were conditional that the sector lower the material content of production, which they did in their own way. Had government told them how to do it, you kill innovation. But strong direction, conditional loans, and they ended up now having the greenest steel in the world. It might not be competitive yet, and that's a scale issue that has also to do with regulation. But repurpose, reuse, recycle technology in steel only happened because the public bank that was mission aligned So I very interested in procurement public loans state enterprises digital public infrastructure examples where they not just things that are there but they used to really transform and help development Sweden is also really interesting because they had a high challenge, right? Missions are somewhere between a challenge, like the space race, and a sector, like aerospace, right? So the moon mission required lots of different sectors, but it was very concrete with the challenge of, say, climate change or all the sustainable development goals, Those are kind of very broad. So transforming them into missions. And their challenge was, I think they said they wanted a fossil free welfare state. And that's why then they said, well, what are the missions? We were working with them on this. So we were kind of stimulating some of this thought through Vinova, their innovation agency. What are the missions that will help us achieve that? And that's where then the healthy, tasty, sustainable school meals policy came from. We worked with them on that also in Brazil. We just actually put out a report about this with the World Food Program. But what was interesting, again, was that then that required government to work in a different way, again, interministerially, catalyzing bottom up experimentation, local manufacturing, but through also the redesign of the tools themselves. The UK, which I don't think is a very good example right now. I mean, there's lots of instability. There's also been 14 years of austerity. Some things that were really interesting. And this is why I look more at the organization. [55:00] kind of examples was government digital services, GDS, which basically began by government back in the early 2000s saying, why does everyone have to go to, say, Google to download a white paper? Why don't we have our own kind of digital platform? They did what most governments do, outsourced it to a company called Circo, which is not a very innovative company at all, gets lots of government contracts. They failed miserably. And then people from the iPlayer team And the BBC said, we'll do it. So they went over to the cabinet office, set up government digital services, came up with this incredible digital platform called Gov.uk, which won an international design award. But what was interesting to me from that example was that the first thing they did was look out the window and said with arrows pointing out the window, those are not clients and customers. Let's stop talking about people as clients and customers. They're users with human rights and how they will access their driver's license, their passport, their voter registration has to enhance their like their souls and not make them want to die. If you see the Ken Loach movies where you know people literally want to die when they accessing their welfare payments because of just the complications around it And so making you know having kind of like a user friendly government digital platform that changes the experience of a citizen with those rights that they have through the technology just requires a very different kind of mind shift And it became the coolest place to work. So if you were a top software engineer or even, you know, whatever, these hackers that Adekolao wanted to hire, that's where they wanted to work to the point that lots of private companies were having a hard time finding the top talent because they all wanted to work in GDS. So again, those are the examples I think we need to look for, not like which is the government that's doing everything perfectly all right professor matsukato thank you so much great to finally catch up with you it took us all like randomly being in the same location in madrid but thank you so much thank you so much for coming on odlos thank you so much for having me. Tracy, I'm glad we finally caught up with Professor Matsukato. But this was the place to do it. This is definitely the place to do it. I think it really did. Her framing makes a lot of sense. This idea that outsourcing is not per se bad, that obviously any major mission is going to have to have significant private sector involvement and innovation even if it somehow publicly led but that there no chance of going anywhere if the public sector doesn have the internal muscle of who to talk to or who to talk to at the right time. Or how to judge performance. How to judge performance seems absolutely a key question. And that kind of gets back to the mission idea, which I do. I really I like the idea of focusing on what you're trying to achieve rather than just how you're actually going about achieving it. Like that makes a lot of sense to me if you're dealing with a vast bureaucracy with lots of different silos. And also, I think like getting back to that expertise point does mean that you do develop that muscle internally within the organization rather than just like, OK, we're going to hire McKinsey to figure out how we're going to do something. Instead, you say, well, we want to do X. Let's let's all get together and figure out how to do it. Yeah. You know, AI seems like a weird thing as a technology because, all right, on the one hand, you could say like, all right, we want to massively improve our healthcare system. We want to massively improve healthcare outcomes. And I think you could like very easily say well AI is going to be a really big part of that right Yeah And maybe make things a lot more efficient maybe give information access to a lot of people identify experts et cetera Develop new medicines Yeah totally It so great. So maybe the mission has something to do with health and AI plays an important component of it. I guess what's strange, though, is that AI itself creates its own potential pitfalls. I mean, the industry, as you mentioned, is obsessed with the pitfalls of its own making, And so there is almost no way that AI can just be a tool in service of some other mission, because almost everyone who knows something about AI sees potential for extreme exacerbation of inequality, potentially AI robots that will be misaligned and want to kill us all when they have sufficient capability. Just little things. Yeah, right. And so like on the one hand, like, yes, as the technology might fit into some of these other big missions. But on the other hand, it sort of feels like it has to there has to be some like AI specific goal of like, where do we want this technology to go? Or how do we how do we curb it or whatever it is that seems very distinct? No, totally. And I do think the unusual part of this moment in time and a lot of people will argue that maybe it's marketing or whatever, but you do see the big tech CEOs like basically going [60:00] and saying like, we as a whole society need to figure out what we want to do here. And I don't think it is marketing or just marketing because, A, there's already this very big tech backlash, right? So like, it's not working. You know, if the idea is, oh, we want to plump our valuations. And so we do that by saying the TAM is all human labor. And human life. Yeah. Well, you're really upsetting a lot of people by saying this. It's not that's not obviously good. And B, you know, as we've written about or talked about, you know, some of these big labs, like they were founded from day one with the premise that this is not normal technology, which is why it's like housed in a nonprofit or something like this. So I tend to think that when the CEOs of these companies talk about this stuff, they kind of mean what they say. They, too, are they, too, are concerned. I mean, I also worry without some sort of government intervention or government strategy here, we are going to get to that situation where we deploy AI. And because we not fixing the underlying system we just sort of nippling at the edges and making it worse per that idea of like OK the insurance bot is going to talk to the hospital bot And both of them are going to say that they streamlining the billing process for medical services But because we're all doing the same thing without actually fixing how medical bills work and who pays for what in the U.S., like it's just going to be bots fighting bots. No one's going to benefit from that. No. And most likely there will be sort of, you know, just ongoing increased complexity. It's bots all the way down. Bots all the way down. All right. Shall we leave it there? Let's leave it there. OK, this has been another episode of the All Thoughts podcast. I'm Tracy Alloway. You can follow me at Tracy Alloway. And I'm Joe Weisenthal. You can follow me at The Stalwart. Follow our guest, Mariana Mazzucato at Mazzucato M. Follow our producers, Carmen Rodriguez at Carmen Armand, Dashiell Bennett at Dashpot, Kale Brooks at Kale Brooks, and Kevin Lozano at Kevin Lloyd Lozano. And for more Odd Lots content, go to Bloomberg.com slash Odd Lots. We have a daily newsletter and all of our episodes. And you can chat about all these topics 24-7 in our Discord, discord.gg. And if you enjoy OddLots if you like it when we define state capacity then please leave us a positive review on your favorite podcast platform And remember if you are a Bloomberg subscriber you can listen to all of our episodes absolutely ad All you need to do is find the Bloomberg channel on Apple Podcasts and follow the instructions there. Thanks for listening. Thank you. price point. And with their 360 degree guarantee, you can be 4imprint certain your order will show up just right right on time Explore more at 4imprint 4imprint 4 certain For Imprint For Certain When you running a business the best days are the ones where priorities stay on track For midsize and large companies risk can affect multiple parts of the organization at once from property and liability to cyber and regulatory challenges. At that level, managing risk becomes an ongoing discipline. At the Hartford, the focus is on helping businesses manage risk before it turns into something more disruptive. And when losses do happen, that work is paired with insurance coverage shaped by years of underwriting, risk engineering, and claims experience. Learn more at thehartford.com slash risk mitigation. Policies provided by Hartford Fire Insurance Company and its property and casualty affiliates, Hartford, Connecticut. What if you could get more from what you already do? As a Shell Fuel Rewards member, that's just a regular weekday afternoon. You get more rewards, more savings, and more special offers. earn rewards when shopping, dining out, and fueling up. And all of that becomes fuel savings when you're at the pump. Use the Shell app and enjoy life with more. Your nearest Shell station is less than five miles away.