Goldman's Solomon Rejects AI Job Apocalypse, Retains 2,500 Entry-Level Hires

高盛CEO所罗门驳斥AI失业潮论调，保留2500名初级员工招聘 高盛首席执行官大卫·所罗门（David Solomon）指出，人工智能将实现日常任务自动化并提高生产力，而不会引发大规模白领失业，这将使初级银行家能够更早专注于客户关系。 高盛首席执行官大卫·所罗门（David Solomon）认为，对AI引发白领就业末日的担忧言过其实。他强调技术将作为生产力倍增器，促使初级银行家更早转向建立客户关系，而非导致大规模裁员。他分享了高盛如何利用AI重构内部运营并清洗数十年积累的专有数据，同时强调人类的情商在高风险的重大交易中依然不可或缺。此外，所罗门还就主要资本市场活动发表了高层见解，包括Alphabet的巨额股权融资和SpaceX即将进行的IPO，同时评估了更广泛的市场估值、银行业网络安全风险以及AI在音乐创作中不断演变的作用。 所罗门驳斥了关于白领将面临灾难性失业末日的预测，坚称技术变革虽然会在短期内扰乱劳动力市场，但最终会推动更广泛的经济增长。 (5:00) AI自动化提供了一个契机，使初级金融人才得以摆脱繁琐的手工杂务，转而在职业生涯早期专注于客户互动与关系构建。 (11:00) 高盛的初级员工招聘规模正回归常态，每年约招收2500名全职新员工，低于疫情期间超3000人的峰值水平。 (15:00) AI即时提供数据可能会导致员工跳过深层的概念学习过程，这迫使金融机构调整对初级员工的指导与培训方式。 (18:15) 高盛的“1GS 3.0”项目展示了AI带来的显著生产力提升，将客户入职、KYC及反洗钱工作流程中的人工接触点从3800个精简至数百个。 (25:00)

Goldman Sachs CEO David Solomon argues that artificial intelligence will act as a long-term productivity driver rather than triggering a white-collar job apocalypse, shifting human capital from manual analytical grunt work back toward relationship-driven banking. Junior Talent and the Apprenticeship Dilemma (11:00) Solomon contends that automating repetitive tasks such as slide-deck assembly and basic data retrieval will allow junior bankers to engage directly with clients earlier in their careers. Goldman Sachs continues to hire around 2,500 interns and 2,500 full-time analysts annually, normalizing from pandemic-era peaks above 3,000. Solomon dismisses projections of massive entry-level job elimination, projecting only modest three-year contractions in junior hiring while operations and compliance workflows—such as client onboarding and anti-money laundering reviews—are streamlined from thousands of touchpoints to a few hundred under the firm's 1GS 3.0 initiative. However, Solomon highlights a fundamental institutional challenge: preserving human apprenticeship in an automated environment. In earlier eras, manually compiling stock comparisons using microfiche and graph paper forced junior analysts to absorb underlying financial mechanics. Because modern analytical tools deliver immediate answers, financial institutions must devise new training methods to ensure emerging professionals build conceptual market intuition without relying entirely on automated outputs. Irreplaceable Human EQ in High-Stakes Dealmaking (35:00) Despite technological advances, Solomon asserts that high-stakes advisory work, client coverage, and underwriting remain fundamentally dependent on emotional intelligence and interpersonal trust. Emphasizing communication habits, such as direct telephone calls having far greater commercial impact than text messages or emails, Solomon argues that distinct creative voice, empathy, and negotiation skills will command a premium as technical execution becomes commoditized. To demonstrate that advisory mandates rely on accumulated trust rather than transactional pitching, Solomon points to Goldman Sachs securing lead roles on mandates like the SpaceX IPO. Tracing the relationship back more than 15 years through coverage of SolarCity and Tesla, Solomon notes that multi-year commitment and ongoing executive dialogue—rather than short-term outreach or social media messaging—remain the primary mechanism for winning major corporate business. Financing Tech Infrastructure and Public Market Dynamics (45:00) The massive capital requirements of the artificial intelligence buildout are altering corporate financing strategies. Solomon points to Alphabet's follow-on equity offering—which Goldman Sachs handled exclusively for five months before expanding the syndicate—where the transaction was upsized from an initial $80–$84 billion target to between $85 and $90 billion. Solomon forecasts that other mega-cap tech companies will similarly tap equity markets to finance long-term, multi-year infrastructure commitments without overextending their leverage profiles. Concurrently, Solomon addresses the shift in public listing timelines, rejecting claims that liquid private markets are damaging capitalism. Due to regulatory burdens and ample late-stage private capital, growing enterprises regularly delay going public until their capital requirements exceed private market capacity, as seen with firms like SpaceX and Anthropic. Valuation Realities, Model Reliability, and Cybersecurity Risks (60:19) Addressing equity valuations, Solomon characterizes the current environment as a tech-driven "greed stage," yet distinguishes it from the 1999–2000 dot-com bubble. The top 10 S&P 500 companies currently trade at forward price-to-earnings multiples in the low 30s backed by substantial cash flow, while the remaining 490 companies trade at roughly 17x to 20x. For internal AI applications, Solomon stresses that model efficacy depends entirely on curated proprietary datasets—such as Goldman's 40-year SecDB history—to prevent the hallucinations typical of unverified public web data. From a risk standpoint, Solomon views cybersecurity as a key systemic vulnerability. While the eight largest U.S. banks have invested heavily to defend against sophisticated AI-assisted attacks, smaller and mid-sized regional lenders lack comparable defensive budgets. Solomon warns that an operational or cyber failure at a regional bank could trigger broader financial contagion similar to the panic following Silicon Valley Bank's collapse. While Solomon envisions productivity-led economic expansion over the next decade, key systemic and structural questions remain unresolved: how to compensate creative intellectual property utilized in AI training sets, how to shield vulnerable regional banks from systemic cyber shocks, and how to maintain the rigor of junior talent development when automated tools bypass foundational analytical work.

高盛集团（Goldman Sachs）董事长兼首席执行官David Solomon认为，关于人工智能将引发白领失业浩劫的论调被过度夸大；技术进步将通过自动化日常繁琐任务释放生产力，推动初级分析师重返以人际信任为核心的客户拓展，而金融业的长期核心竞争力依然建立在真实数据治理、人际沟通与审慎的资本配置之上。 初级人才重回人际网络与学徒制转型 (11:00) Solomon驳斥了关于白领将被大规模淘汰或依赖全民基本收入的极端预测，指出历史上的技术加速最终都会扩大经济总量与就业参与。针对有数据称初级岗位招聘下降16%的现象，他认为这仅反映了特定企业部门的统计偏差，并未涵盖更广泛的就业路径。高盛目前仍维持每年约2,500名实习生和2,500名全职新员工的入职规模，招聘人数从2021至2022年疫情期间超3,000人的峰值回归常态，未来三年可能会出现小幅收缩。 过去三四十年间，银行业过度依赖内部数据分析和幻灯片制作，削弱了年轻人的商业拓展能力。早期分析师需要花费6小时通过缩微胶片和坐标纸手动绘制股票走势图，虽然耗时，却通过强制性的手动研究加深了对底层市场的理解。如今AI工具瞬间生成答案，虽然对传统学徒培训机制构成了挑战，但也为年轻银行家打破了后台束缚。Solomon强调，直接通话的效果是短信或邮件的十倍，技术应充当“力量倍增器”，促使年轻员工更早走向前线与客户建立信任关系。 数据清洗、流程精简与人类判断的必要性 (30:00) 在衡量技术生产力方面，短期人均产出往往难以直接量化，但从长期资本效率看成效显著。高盛IPO时拥有约60亿美元资本和16,000名员工，如今资本规模达到1,100亿美元，员工逾45,000人，资本回报率大幅提升。目前高盛正推进“1GS 3.0”内部流程重组，将以往客户准入、KYC和反洗钱流程中需要3,800人处理的触点精简至数百人，大幅提升了运营效率。 Solomon强调，生成式AI的效果完全取决于数据质量。高盛正系统性清洗其专有交易系统SecDB积攒的40余年交易与投行业务数据。如果直接接入未经筛选的互联网数据，模型极易产生幻觉——正如他向AI询问高尔夫大师赛背靠背冠军记录时，AI曾遗漏Tiger Woods并给出错误答案。此外，投资银行与交易本质上是建立在情商（EQ）、创意和信任之上的领域，无论模型如何拟合脚本或方案，都无法替代承载人类情感经历的独特表达与现场交流。 资本市场结构转变与超级IPO的回归 (45:00) 在资本配置层面，科技巨头的AI基础设施建设催生了前所未有的融资规模。高盛此前作为唯一定价经办行运作了5个月，协助Alphabet完成了从最初计划的800亿至840亿美元上调至850亿至900亿美元的巨额增发。Solomon指出，超大市值科技公司着眼于5至10年的资本开支计划，在当前估值高企的环境下，发行股权相较于全靠举债能更稳妥地控制杠杆水平，预计将有更多科技龙头采取类似策略。 针对高盛牵头SpaceX以及近期获任Anthropic IPO的进展，Solomon澄清所谓通过社交平台私信争取业务的传闻，强调这是高盛团队逾15至20年深耕客户关系的积累。针对私有市场过热是否“破坏资本主义”的讨论，他指出企业推迟上市是过去25年监管负担与私有资金充裕共同作用的结果。然而，当巨头们的资本胃口达到仅靠私有资金无法审慎满足的极值时，重返公开市场便成为必然选择。 估值分化、系统性网络风险与创意局限 (60:19) Solomon在纽约经济俱乐部指出当前市场处于“贪婪阶段”而非“恐惧阶段”。尽管标普500指数前10大公司市值占比超过35%，但它们具备充沛的现金流和扎实的盈利支撑，远期市盈率处于30倍出头的水平，与互联网泡沫时期思科市盈率达40至50倍、1960年代宝丽来达80倍的极端估值存在本质区别。与此同时，指数其余490家公司的远期市盈率在17至20倍之间，随着技术应用提升其运营效率，未来盈利增长仍具吸引力。 在风险防控方面，Solomon警示了AI时代的网络安全失衡。全美排名前8的大型银行具备顶尖的防御资源，真正的系统性脆弱点在于中小型区域银行及地方公共基础设施（如地方供水系统）。正如硅谷银行引发的恐慌一样，中小机构若遭受网络攻击可能引发蔓延效应。在创意与音乐领域，AI虽通过数字化音频工作站降低了门槛，但依然面临未授权训练的知识产权侵权困境，且缺乏人类审美与情感引导的纯AI产物极易流于平庸。 Solomon总结指出，自2020年1月首个投资者日以来，高盛营收增长约65%，净利润增长140%至145%，股价突破1,000美元大关，证明了重构多元业务模式的有效性。然而，AI在金融与社会中的全面应用仍面临诸多悬而未决的挑战，包括中小金融机构的网络防御韧性、生成式内容的知识产权保护，以及在技术完全自动化任务后如何有效维持初级专业人才的基础知识沉淀与传统学徒培训机制。

## Background 本期播客由彭博社（Bloomberg）旗下的《Odd Lots》栏目于 2026 年 6 月 4 日发布。节目简介指出，在当前关于 AI 是否会带来白领大规模失业与投资回报不及预期的激烈辩论中，业务与用人结构多元的大型跨国银行是绝佳的观察棱镜。本期节目邀请到高盛集团（Goldman Sachs）董事长兼首席执行官 David Solomon，深入探讨 AI 对银行业务模式及用人编制的实际影响、华尔街资本市场现状、高盛牵头 SpaceX IPO 与 Alphabet 巨额股权融资的内幕，以及他在垃圾债时代的早期职业历程与对 AI 音乐制作的见解。 ## TL;DR 高盛 CEO David Solomon 在访谈中坚决驳斥了“AI 白领末日论”，认为技术将演示文稿制作和基础财务模型等繁重机械劳动彻底自动化，为初级分析师重返直接面对客户、建立真实商业信任提供了历史契机。他强调情商、主观直觉以及面对面建立的信任是算法无法跨越的人性护城河，高盛内部正通过“One GS 3.0”战略重塑中后台运营流程以释放人效，同时在前台与应届毕业生招募上依然维持庞大吸纳规模。在资本市场前沿，Solomon 披露了高盛独家操盘 Alphabet 近 900 亿美元历史级股权融资以应对长期 AI 算力开支的战略考量，并阐明了历经二十载深耕积累赢得 SpaceX IPO 主承销权的长期主义投行哲学。最后，他横向对比了当前美股头部公司的真实高盈利能力与过往泡沫周期的本质差异，并结合自身电子音乐制作爱好，剖析了 AI 虽降低了创作门槛，却面临着知识产权收益缺失以及缺乏人类“独特声音”的艺术平庸化困境。 ## Atmosphere & style 整场对话洋溢着华尔街顶级掌门人的从容、务实与松弛感，主持人的机敏打趣（如调侃所罗门是否在社交平台上私信马斯克揽客）与所罗门坦诚而自嘲的回应碰撞出活跃亲切的互动氛围。叙事在华尔街历史纵深与当下超级交易之间自如穿梭，既有早年依靠缩微胶片手绘图表、每天拨打上百通残酷冷线电话的生动复盘，也有执掌万亿级巨额增发与前沿模型安全审查时的沉着远见。整期节目剥离了空洞的末日恐慌与盲目的科技狂热，在宏大资本周期的冷静分析中，始终贯穿着对人际温情、深层信任与创造力本质的笃定关怀。 ## Bulletin - `05:00` 高盛董事长兼 CEO David Solomon 反驳“AI 引发白领大失业”的末日论，指出高盛目前仍招募约 2,500 名实习生及等量全职新人，AI 将初级员工从熬夜做 PPT 和繁琐建模中解放出来，使其能更早走向前端面对客户。 - `44:45` 高盛独家操盘 Alphabet 历史级后续股票增发，募资规模达 850 亿至 900 亿美元，标志着科技巨头在应对 5 至 10 年 AI 资本支出周期时，开始从纯债务融资转向股债平衡。 - `48:00` 高盛赢得 SpaceX IPO 主承销商席位，Solomon 澄清并非外界谣传的“X 平台私信抢单”，而是高盛团队自 SolarCity 与特斯拉早期起长达 15 至 20 年的长期服务与信任沉淀。 - `01:00:15` 针对当前美股集中度风险，Solomon 指出标普 500 前十大龙头的前瞻市盈率约为 30 倍出头且盈利强劲，远低于历史极端泡沫水平；其余 490 只成分股市盈率仅 17 至 20 倍，估值具长远吸引力。 ## Key takeaways - `05:00` AI 带来的生产力繁荣不会是平滑直线，但所谓的“这次不一样”的岗位末日论忽视了宏观劳动力市场在技术冲击下的自我重构与新业态孵化机制。 - `13:58` 投行初级岗位在过去几十年中异化为熬夜做 PPT 和核对财务模型，AI 自动化正在终结这种低效劳动，使金融人才培养机制重新回归直接人际沟通。 - `15:10` 所谓“入门级白领招聘下滑 16%”的统计口径过于狭隘，忽视了广泛的服务业、创业及技能型跨界发展，大学教育核心在于培养终身学习能力而非单纯挤入特定白领窄道。 - `20:00` AI 时代金融业面临的最大隐忧在于“学徒制”的传承断层：以往在胶片和坐标纸中费力寻解的过程帮助年轻人内化了商业逻辑，而即时生成的自动化答案可能削弱知识吸收深度。 - `25:00` 直接通话与面对面眼神交流的价值在信息泛滥时代被十倍放大，随着标准答案被机器商品化，建立人际信任的能力成为最高溢价的核心资产。 - `27:30` 高盛通过“One GS 3.0”推行端到端流程重组，将 KYC/AML 等合规准入涉及的人力从 3,800 人精简至数百人，以量化数据证明了运营后台的生产力飞跃。 - `31:30` AI 模型效能依赖数据集的纯净度：高盛拥有 40 年自主研发的 SecDB 证券数据库作为内部护城河，而在开放互联网上模型极易因数据杂质产生离谱错误，离不开人类判断力的持续纠偏。 - `35:00` 存在于人类大脑深处的隐性知识、主观审美与情商纽带永远无法被算法全盘提取，真正的核心决策依然取决于感性信任与独特的个人“声音”。 - `40:48` 自 2020 年初以来，高盛净利润累计增长 140% 至 145%，股价突破 1,000 美元历史大关，其战略重心在于通过重构流程释放产能以追求跨周期的长期盈利能力。 - `55:00` 企业在条件允许下应“尽可能晚上市”，当前超大规模独角兽集中启动 IPO 是因为其 AI 基础设施开支已达到私募市场难以完全承受的贪婪量级，必须借助公开市场资本。 - `01:04:47` 前沿 AI 模型引发的国家安全挑战重点在于跨机构的防御鸿沟：大型银行具备充沛的安防预算，而中小型金融机构与地方公用事业由于资源匮乏更容易成为系统性风险引爆点。 - `01:10:00` 数字音乐工作站与 AI 极大地民主化了艺术创作，但也带来了知识产权未受补偿的法律伦理危机，以及缺乏人类主观审美打磨导致的生成内容平庸化风险。 ## Detailed walkthrough ### `00:00` 华尔街大行视角的 AI 与白领就业：播客开场 节目的开场白中，主持人 Tracey Allaway 与 Joe Weisenthal 聊起了近期 Odd Lots 即将在香港 Soho House 举办的金融线下问答活动，随后切入本期主题：AI 对劳动力市场和企业运营模式的深远冲击。在探讨 AI 是否会带来大规模就业替代时，投资银行是一个绝佳的观察棱镜。一方面，华尔街投行拥有极为雄厚的技术预算，并且深度对接各类 AI 前沿企业，很多科技初创公司本身就是银行的核心投行客户；另一方面，银行内部的岗位谱系极为丰富多元： - 既有负责日常流程、重复性事务的后勤运营（Back-office）岗位； - 也有日夜赶工制作 PowerPoint 演示文稿和财务模型的初级分析师（Junior Analysts）； - 同时还包含大量依靠人脉网络、信任与情感纽带的资深投资银行家（Senior Investment Bankers）。 Joe 回忆起 2010 年代华尔街各大银行 CEO 纷纷自诩为“科技公司”、在办公区摆满乒乓球桌以招揽工程师的时代风潮。但他指出，在通用具身机器人成熟之前，AI 率先产生颠覆性冲击的正是白领阶层。然而，投行业务的核心往往建立在看似无法被算法替代的“客户关系”之上——正如华尔街常被调侃的“去高尔夫球场谈生意”，这种人际互动其实是业务运作的关键支柱。此前节目曾邀请过高盛首席技术官 Marco Argenti 分享高盛的技术部署，而本次则迎来了执掌全局的最高领导者：高盛集团董事长兼首席执行官 David Solomon。 ```mermaid flowchart TD accTitle: 投行岗位的 AI 冲击光谱 accDescr: 展示从容易被自动化替代的后端岗位到依赖深层人际信任的前端岗位的分布光谱 subgraph BackOffice["后勤与数据层"] A["流程化运营 (Back-office)"] B["数据录入与结算合规"] end subgraph MidTier["分析与内容制作层"] C["初级分析师 (Junior Analysts)"] D["PPT 报告制作与初级财务模型"] end subgraph SeniorTier["高阶关系与决策层"] E["高级投资银行家 (Senior Bankers)"] F["高尔夫/晚宴面对面人际信任"] G["复杂战略判断与高风险交易促成"] end BackOffice --> MidTier --> SeniorTier ``` --- ### `05:00` 破除“AI 岗位末日论”：David Solomon 的乐观主义与历史底色 David Solomon 近期在《纽约时报》（The New York Times）发表了一篇题为《我是高盛 CEO，AI 引发的就业末日论被过度夸大了》（"I'm the CEO of Goldman Sachs, the AI job apocalypse is overblown"）的专栏文章。在播客中，David 坦言自己并不想成为媒体专栏作家，在担任高盛最高负责人的八年时间里，这仅仅是他撰写的第二篇专栏文章。他的上一篇专栏还要追溯到 2019 年新冠疫情前，当时聚焦于能源供应链转型。他在那篇文章中虽然支持投资绿色清洁能源，但顶住舆论压力直言：石油与天然气在未来几十年内仍将是全球能源供应的关键支柱，高盛将在相当长的时间内继续为传统化石能源企业提供融资服务——这一事实后来完全得到了市场的验证。 谈及撰写这篇 AI 专栏的初衷，David 强调了三个核心判断： 1. **五至十年的宏观乐观主义**：他对 AI 抱有极度乐观的态度。AI 技术能够大幅赋能高盛这样的大型复杂机构，释放极大的生产力红利，进而推动经济增长、创造社会财富，并为整体经济生态带来正面外溢效应。 2. **转型过程的非线性与痛苦颠覆**：生产力繁荣不会是一条平滑的直线。正如人类历史上的每一次技术浪潮一样，颠覆必然发生，某些领域的岗位会被替代。大型跨国企业与政府必须协同制定策略，让技术转型以最平稳的方式落地，最终目标是提高全要素生产率，使更多社会成员参与并获益。 3. **否定“这次不一样”的末日论**：他反驳了“未来大家都不用工作、社会只能依赖全民基本收入（UBI）、面临毁灭性失业危机”的悲观叙事。根据他 40 余年的金融职业经验，历史上每一次技术革命都听过“这次不一样”的论调，但从来没有真正“不一样”过。此类极端预测脱离了宏观劳动力市场的底层细颗粒度现实，忽视了美国经济在过往技术冲击中自我调整、重构岗位结构并诞生全新业态的内生机制。 --- ### `07:30` 华尔街史前时代：1984 年的职业起点与金融业演变 为了深入探讨“AI 是否会消灭白领”，必须先剖析“投行岗位究竟包含了什么技能”。Joe 引用了 David 早年在贝尔斯登（Bear Stearns）晋升并最终跳槽高盛时的自述——“我是贝尔斯登的顶尖创收者（Big Producer），在那里建立了极为庞大的客户人脉，并把这些关系带到了高盛”——以此询问 David，在群雄逐鹿的投行环境里，究竟是什么个人特质和技能让他脱颖而出并最终登顶高盛。 David 风趣地回顾了自己 1984 年入行时的光景，戏称那是“金融业的史前时代”，调侃自己当年甚至还留着“一头飘逸的长发”。他指出，在探讨个人能力之前，必须理解当时的时代背景与金融生态的演进： - **长期滞胀与沉寂的金融市场**：从 1960 年代末到 1980 年代初的 15 年间，受持续恶性通胀影响，美联储利率一路飙升，华尔街投行业务在很长一段时间内相对处于萌芽与停滞状态。 - **1982 年 9 月 15 日的历史拐点**：这是全球金融市场历史上具有里程碑意义的一天，美国 10 年期国债收益率在那一天触及惊人的 15.9% 历史高点，随后开启了长达数十年的长期大牛市与债券收益率下行周期。 - **职业生涯起步**：在那个利率历史大顶不到两年后，David 来到位于华尔街 1 号（One Wall Street）的欧文信托公司（Irving Trust Company）开启了金融生涯。当时的初级人员同样需要接受严苛的财务分析与信贷评估培训，但当时的交易结构、工具复杂度与人际交往模式，与如今这个数据高度密集的全球化金融体系有着天壤之别。 ### `10:00` 投身垃圾债热潮与美林冷线电话的残酷洗礼 80 年代初期的金融界极具创业精神，几乎没有什么条条框框，到处都是可供年轻人闯荡的空白地带——尤其是当你避开当时最负盛名的“白手套”（white-shoe）老牌投行，比如高盛（Goldman Sachs）和摩根士丹利（Morgan Stanley）时，年轻人更是被极力鼓励去自立门户、开辟天地。当时大卫·所罗门（David Solomon）正在欧文信托（Irving Trust）任职。在那个几乎所有金融从业者都要读商学院的年代，他正筹备在 1985 年申请商学院，却意外拿到了德崇证券（Drexel Burnham Lambert）高收益债（即“垃圾债券”，junk bond）业务的工作邀约。垃圾债在 80 年代初是金融界几乎唯一迎来爆炸式增长、极具拓荒精神的领域。面对“大家都去读 MBA、唯独自己走偏锋”的风险，所罗门向父亲请教，父亲给了他一条简单而直接的建议：“去试两年，如果不成，两年后再去读商学院。” 德崇证券的企业文化极为激进、崇尚狼性，尽管其中一些负面因素最终导致其在 1990 年 2 月 14 日轰然倒闭，但在所罗门任职的四年半时间里，年仅 24 岁的他被赋予了极大的自由度与责任：“去建立人脉，把业务和钱带回公司。”那不是一个靠高尔夫球场社交的时代，他们就坐在办公桌前，不断拨打电话与人沟通。而在所罗门个人的职业成长中，最关键的一课其实发生在大学一年级结束后的那个夏天（1981 年）。当时他凭借高中女友父亲的引荐，在麦迪逊广场花园附近（宾夕法尼亚广场 10 号）的美林证券（Merrill Lynch）办公室获得了一份暑期实习。 这份实习给他的任务简单而残酷：办公室扔给他一份曼哈顿上东区的邮政编码名单（10028 和 10128），要求他每天打 100 个冷线销售电话（cold calls），询问对方是否愿意见美林的经纪人。整整 8 周时间，他日复一日地拨号，最终整个夏天只成功促成了 5 个人愿意交谈，最终完成 1 次线下拜访并开了 1 个新账户。第一周结束后所罗门就几近崩溃，向父亲倾诉自己干不下去，但父亲严厉告诫他：“既然接下了，就必须做完。”为了撑下去，所罗门把枯燥难耐的冷线电话变成了一场游戏。这项残酷的训练让他彻底克服了对与陌生人交流的恐惧。日后到了德崇证券，在丰厚的提成和激励机制刺激下，所罗门发现自己能够毫不怯场地下场联系任何人、留住客户、促成交易；后来转战贝尔斯登（Bear Stearns）这间信奉“猎物自食”（eat what you kill）文化的机构时，他迅速成长为顶级的业务骨干。 - 德崇证券的高收益债业务在 1980 年代初处于极具创业色彩的爆发阶段。 - 1981 年美林证券实习：主攻曼哈顿上东区（邮政编码 10028 与 10128），每天 100 个电话，连续 8 周，最终促成 5 次通话、1 次会面、开立 1 个账户。 - 德崇证券最终于 1990 年 2 月 14 日倒闭，随后所罗门加盟贝尔斯登。 ### `13:58` 赋能初级银行家：从深夜改 PPT 转向直面客户 所罗门从早年的推销历练中得出一个核心洞察：人与人之间面对面的真实沟通（human-to-human contact）永远不可替代。过去三四十年间，投行青年员工的培养路径发生了严重的偏移——初级分析师往往被困在办公室里通宵达旦地调整 PowerPoint 胶片或反复核对财务模型，远离了真实的商业和客户。所罗门认为，这一演变方向并不是好的转变。随着人工智能等新技术的加速普及，PPT 排版和数据处理等繁重劳动将被彻底自动化与商品化，不再需要十名分析师为了制作一份演示文稿熬通宵。 这不仅不会抹杀初级员工的价值，反而为金融业提供了一次将杠杆重新“回调”的历史契机。技术把年轻人才从琐碎低效的杂务中解放出来，让他们能够更早地走出办公室，直接与客户沟通、拓宽业务版图。所罗门强调，现在的年轻员工智识极高、能力出众，既然外部同样年纪（甚至不到 30 岁）的年轻人已经在独立创办高价值企业，高盛的年轻精英们也完全有能力尽早下场面对客户。 ```mermaid flowchart TD accTitle: Junior Banker Evolution accDescr: 过去三十年投行初级员工深陷自动化工作，AI介入后解放其重返客户沟通 subgraph Past["传统模式（过去30-40年）"] A["大量初级分析师"] --> B["繁重手工：排版 PPT / 摘录数据"] B --> C["熬夜赶工，远离客户真实沟通"] end subgraph Future["AI 时代新范式"] D["AI 工具与自动化平台"] --> E["瞬间生成模型初稿与简报"] E --> F["青年员工获技术杠杆赋能"] F --> G["走出办公室，直面客户拓展业务"] end ``` ### `15:10` 驳 16% 初级招聘下滑论：破除白领职业的狭隘叙事 当主持人提到斯坦福大学一项显示“入门级白领岗位招聘量下降 16%”的研究数据，并质问初级分析师职位是否正在被 AI 吞噬时，所罗门对该数据背后的狭隘假设提出了质疑。所罗门指出，这项研究所统计的仅仅是少数符合传统社会声望定义的白领精英行业，以此作为衡量整体青年就业健康度的标尺本身就存在视角偏差。如果一个文理学院的毕业生选择去当老师、开办连锁三明治店，或者从事手工业，根本不会被计入所谓的“16% 下滑”之中。 主持人在一旁打趣提到自己大学毕业后也做过“三明治生意”（实际上是在熟食店打工），所罗门则分享了他大学同学的真实案例：一位在纽约上州读物理科学的朋友毕业后选择做一名电工并自主创业，十到二十年后，凭借在电工业务中积攒的管理和商业技能，该同学成功横向跨界到更广阔的商业领域。所罗门强调，大学教育的核心在于成为终身学习者（learner through life）、开阔视野并结识朋友，而非单一地涌向某几类所谓的体面白领岗位。无论是开三明治店还是做电工，自食其力并养家糊口、追寻热爱的道路本就数不胜数。 - 斯坦福数据：特定入门级（entry-level）白领招聘下滑 16%，所罗门认为该统计口径局限，忽视了教育、服务、创业等广阔就业市场。 - 职业观倡导：教育的本质在于培养终身学习者，职业发展路径绝非仅有华尔街或咨询公司一条窄道。 ### `18:03` 高盛的招聘实态与技术杠杆的生产力演进 谈及高盛自身的实际人才吸纳规模，所罗门澄清事实与外界的悲观预期截然相反。在节目录制当下，高盛刚刚迎来了约 2,400 到 2,500 名暑期实习生；同年 7 月，还将迎来规模大致相同的一整批全职应届新员工。这些岗位广泛分布在面向客户的业务部门、中后台运营、工程与软件技术部门，甚至包括市场营销和广告等多样化职能，绝对没有出现大规模缩水。 所罗门提到，前几天晚上他刚在自己的公寓里与一群由风投支持的初创企业创始人共进晚餐，这群年轻创业者的平均年龄仅为 28 岁左右，极其精明干练。而高盛所招募的年轻分析师无论在智商还是潜力上，都与这群创业者不分伯仲。技术的飞跃不是为了裁撤他们，而是赋予他们更强大的杠杆。为了说明技术的杠杆效应，所罗门对比了自己初入行业时的极端低效：当年为了做一份普通股走势对比分析（common stock comparison），一名分析师得花整整 6 个小时——去翻阅缩微胶片（microfiche），查阅作为唯一历史股价源的《华尔街日报》，手绘在坐标纸上，再拍照制成幻灯片。在那样的时代，企业要想知道自己日内的即时股价，甚至不得不通过打电话或特定终端才能勉强获取。技术将分析师从长达 6 小时的机械劳动中解放出来，其根本目的不是消灭岗位，而是将顶尖智力解放到更高维度的决策与客户服务之中。 - 高盛当期新进人才规模：约 2,400–2,500 名实习生入职，7 月份迎来大致同等规模的常规模拟全职新人。 - 岗位多元覆盖：客户前端、技术工程、中台运营、营销与广告等多线业务。 - 历史效率对照：早年通过缩微胶片（microfiche）与坐标纸手绘制作一份股票对比分析需耗费整整 6 小时，现代信息技术已将其压缩至瞬间完成。 ### `20:00` 技术变迁与金融业“学徒制”的困境 David Solomon 回顾了自己刚入行时的金融业生态：当时客户若想知道某只股票的最新价格，必须打电话给投资银行家，而投行办公室里刚好有一台 Quotron 报价机，交易员能据此报出价格；普通人若没有 Quotron，就得找股票经纪人，或者走到街上的 Merrill Lynch、EF Hutton 或 Dean Witter 营业部门口盯着滚动行情纸带（ticker tape）等待报价，甚至只能等到次日翻看《华尔街日报》上的前日收盘数据。当年投行年轻分析师所做的大量工作，本质上是借助非常原始的技术手段，向企业运营者传递基本信息，整个行业的运转节奏极其缓慢。 过去四十余年里，技术工具突飞猛进，员工的生产力大幅提升，但高盛的整体雇员规模却并没有因此缩减。面对当下 AI 的迅猛发展，未来长期是否会导致人员总数根本性下降仍是未知数，但 Solomon 指出，高盛当前面临的最大挑战是“学徒制”（apprenticeship）的延续。过去分析师需要去查缩微胶片（microfiche）、在坐标纸上手绘股票走势图，并手动计算复利差距，在这一艰难获取答案的过程中，大脑深度吸收并消化了金融底层逻辑；如今这些答案可以瞬间生成，年轻人的大脑是否真正吸收了这些知识？如何在赋予年轻人强大数字化工具快速发挥影响力的同时，建立扎实的知识底座，是管理层正在深入思考的课题。 - **早年信息中介模式**：客户依靠投行的 Quotron 机器获取实时股票报价，个人投资者则依赖街头营业网点纸带或隔日印刷报纸。 - **学徒制的核心挑战**：自动化与即时答案剥夺了初级员工“费力寻找答案”时的深度思考与内化过程。 --- ### `25:00` 人际沟通的持久价值与科技周期的招聘修正 谈及从业技能的传承，Solomon 坚信直接拿起电话与人交谈的能力依然不可替代。他提到上个月在沃顿商学院毕业典礼上的演讲内容，当时他向毕业生强调“电话是世界上最伟大的技术之一，多用它”。在人际交往与商业拓展中，一通真实的电话沟通所带来的影响力，往往是发短信或电子邮件的十倍。主持人打趣回应虽然自己没有 Solomon 的手机号，但“如果是 David Solomon 发短信，我也完全接受”，引发现场笑声。 针对 ChatGPT 问世（2021/2022 年）至今的招聘结构变化，Solomon 提醒不能孤立地看待数据。2022 年全行业招聘规模触顶，本质上是疫情期间过度扩张导致的“人员臃肿高水位线”（high watermark of the bloat）。从 2022 年到 2026 年，各大机构的大量调整实际上是在修正 2020 至 2021 年间所犯下的招聘错误。若将视线拉长到十年维度： 1. 过去十年，工程技术类人才在高盛内部的相对占比显著超越了运营和传统客户支持类岗位； 2. 随着代码辅助工具与 AI 能力的大幅提升，未来的工程类编制增速可能会发生微调； 3. 应届生招聘人数从 2021 年高峰期的 3,000 多人回落至目前的 2,500 人左右（接近疫情前水平），预计未来三年可能会有温和微调，但依然会大规模吸收应届毕业生，绝不会出现白领断崖式缩减。 --- ### `27:30` 衡量生产力飞跃：“One Goldman Sachs 3.0”与人均产出 对于如何量化高盛内部真实的生产力提升，Solomon 区分了“业务运营流程重塑”与“传统前端投资银行”两个维度的衡量逻辑。在运营重构方面，高盛正在推行“One GS 3.0”战略，针对客户准入（client onboarding）、反洗钱（AML）以及了解你的客户（KYC）等繁复流程进行彻底改造。过去此类庞杂流程需要约 3,800 名员工不同程度地参与其中，而流程重塑完成后，将精简至仅需几百人处理，这类端到端（end-to-end）的项目能够提供明确且可量化的生产力提升数据。 而在投行等前端业务条线，衡量生产力则更加复杂。评估投资银行业务时，管理层主要追踪人均创收（revenue per head）、人均利润以及与之匹配的资本分配。过去十年间，投行部门的人均收入与利润均录得增长，人均效率显著提高；但与此同时，管理层也为该业务配置了更多资本以提供杠杆支持。技术的赋能叠加资本的杠杆效应，共同驱动了投资银行业务的人效跃迁。 ```mermaid flowchart TD accTitle: Goldman Sachs Productivity Measurement Framework accDescr: A diagram contrasting operational process reengineering with investment banking capital-leveraged productivity. subgraph Operations["运营流程重构 (One GS 3.0)"] A["原有 KYC / AML / 客户准入"] --> B["3,800 人不同程度参与节点"] B --> C["AI与数字化流程再造"] C --> D["精简至数百人处理"] D --> E["输出: 清晰可量化的工时与成本降幅"] end subgraph FrontOffice["前端投资银行条线"] F["年轻银行家 + 现代化数字工具"] --> G["人均创收与人均利润提升"] H["资产负债表与资本支持"] --> G G --> I["输出: 资本杠杆与技术协同的人效提升"] end ``` ### `30:00` 高盛的生产率跃迁与内部协同文化 谈及生产率的衡量，大卫·所罗门（David Solomon）指出，试图在按月或按年的微观维度上拆解单一工具对生产率的贡献是非常困难的。然而，如果以五年甚至几十年的跨度来审视，生产力的提升是显而易见的。回溯到高盛 27 年前上市时，公司仅拥有约 60 亿美元资本和约 1.6 万名员工；而如今，高盛已发展到拥有 4.5 万名员工和 1100 亿美元资本，并在这些庞大资本上持续创造出可观的回报。此外，高盛内部推行的“One Goldman Sachs”（1GS）1.0 与 2.0 战略强化了跨业务部门的协作，进一步驱动了收入增长。从长期来看，员工拥有了更强大的工具，能够承担更多工作，生产率的跃升毋庸置疑。 主持人随后抛出一个深层治理问题：AI 极度依赖汇聚海量非结构化数据，但在高盛或律所等精英专业服务机构中，常常存在利益对齐问题（alignment problem）——顶级业务高手（big producer）往往视自己的交易经验与人脉为私有护城河，以便未来跳槽时带走业务簿，因而可能不愿意将脑中的隐性知识全盘共享进公司的中心化知识库中。所罗门对这个问题的洞察表示赞赏，但他强调，高盛的文化恰恰相反，高盛从不奉行“个人明星制”（star system），而是长期将“分享与协作”深植于企业基因之中，因此高盛受到这种阻碍的程度远低于同行。 - **高盛上市时与现状对比**：资本从约 60 亿美元增至 1100 亿美元，员工人数从约 1.6 万人增至 4.5 万人。 - **内部协同机制**：通过“1GS”（One Goldman Sachs）1.0 和 2.0 框架打通各部门协作，消除信息壁垒与个人孤岛。 --- ### `31:30` 优质数据集与 SecDB：AI 面临的“垃圾进，垃圾出”挑战 所罗门将 AI 应用的有效性划分为三个逻辑层面。第一层是**干净的数据集（clean data set）**：当把大模型部署在高度整洁、结构完备的数据集上时，其产出的速度、能力与深度是令人惊叹的。高盛在这方面拥有无可比拟的护城河——早在 40 年前，高盛就自主研发了著名的证券数据库交易系统 **SecDB**（Securities Database），记录了过去 40 多年来所有交易与投行业务的详尽数据，这在华尔街几乎是独一无二的。目前高盛正全力投入，确保这些历史数据尽可能保持最纯净的状态。 ```mermaid flowchart TD accTitle: Data Quality and AI Output Architecture accDescr: 干净数据集与嘈杂互联网数据在大模型中的处理路径与结果差异对比 subgraph Internal["高盛内部专有环境"] D1["40 年 SecDB 交易数据 & 投行历史记录"] --> C1["内部数据清洗与标准化"] C1 --> M1["大语言模型"] M1 --> R1["高精度洞见 / 极速可靠执行"] end subgraph External["开放互联网环境"] D2["互联网 / 社交媒体 / 开放新闻"] --> C2["嘈杂非结构化数据源"] C2 --> M2["大语言模型"] M2 --> R2["错误事实 / 幻觉 / 答非所问"] R2 --> H1["必须介入人类判断纠错"] end ``` 然而第二层则是**不干净的数据集**：一旦把大模型推向开放互联网，模型就会频繁给出荒谬离谱的答案（"cockamamie answers"）。为了说明这一点，所罗门分享了一则亲身经历的趣事：几周前他去现场看美国高尔夫大师赛（The Masters），当时罗里·麦克罗伊（Rory McIlroy）在打完前两轮后领先 6 杆，所罗门便好奇历史上究竟有谁曾在大师赛背靠背连续夺冠。他记得泰格·伍兹（Tiger Woods）和杰克·尼克劳斯（Jack Nicklaus）做到了，但不确定还有谁，于是随手向一个 AI 模型提问。 - **AI 的漏洞回答**：模型明确回答“杰克·尼克劳斯和尼克·费度（Nick Faldo）”，却把泰格·伍兹漏掉了。 - **追问与辩解**：所罗门反问模型：“你算错了，难道泰格·伍兹没有连冠吗？”模型停顿一下后认错：“是的，您是对的，泰格·伍兹确实连续两年夺冠。”当所罗门质问它为什么会出错时，模型开始推脱，解释称自己引用的某些源数据存在偏差。 - **静态检索缺陷**：当所罗门追问：“那你下次搜索时会改变搜索逻辑吗？”模型却回答：“不，这不是我的工作机制。”所罗门直言，这就是典型的“垃圾进，垃圾出”（Garbage in, garbage out），模型去抓取社交媒体和网络碎片时必然伴随大量杂质，这凸显了人类判断力（Human judgment）在纠偏中的关键地位。 --- ### `35:00` 无法被算法抽取的隐性知识：情商、信任与人类声音 谈到第三层核心，即“如何把存在于人类大脑深处的隐性知识提取出来”时，所罗门的态度非常明确：**“我认为答案是永远不可能（never）”**。在他看来，这正是人类与机器之间不可逾越的护城河，特别是在高度依赖创造力和人际交往的投资银行领域。投资银行的核心不仅是数字计算，更是关于 EQ（情商）、建立信任、与客户面对面交流的人性艺术。机器可以成为极其出色的效率工具，但只要人类社会依然以人与人之间的信任为纽带，机器就绝不可能完全替代人类。真正的破局之道在于“借助工具放大自身的人性与连接力”。 为了生动诠释这一点，所罗门提到了自己当编剧的女儿。许多人热衷于讨论 AI 是否能写电影剧本，答案是肯定的，AI 当然能写出工整的剧本。但所罗门指出，那部剧本绝不可能拥有他女儿独有的“声音”（voice）： > “除非那个模型能够访问她曾经写下的每一个字、经历过的每一次人生体验、感受过的每一种情绪……但模型永远不可能拥有这些。在创作与写作中，独特的‘声音’至关重要。” 更关键的是现实中的人际化学反应。即使女儿完成了剧本并带到电影制片厂，制片人可能会挑剔说：“这个地方我不太喜欢，那个地方不对劲。”这种主观偏好纯粹是基于个人情感与直觉的感性反应，制片人绝不会让 AI 来代替自己做决定。投行业务同样如此：一个投行团队即便拿出再完美的数据分析报告去 Pitch 交易，客户的首席财务官（CFO）也可能仅因为一句“我就是不太喜欢那个家伙对我说话的方式”而否定整个合作。 这种存在于人脑中的、微妙的情感反应与直觉，正是无法被数据化的超级力量（superpower）。所罗门坚信，随着底层数据集越来越干净、工具越来越普及，人类所拥有的独特技能——公众演讲（public speaking）、真正有灵魂的写作、真诚的沟通以及如何投入时间与他人建立互信——不仅不会贬值，反而会变得“越来越有价值”。 --- ### `37:30` 商业赞助插播 音频末尾插入了两则商业赞助播报： 1. **4imprint 宣传品定制**：宣传其在各价格区间提供的定制礼品与周边解决方案，具备 360 度服务保障与准时交付承诺。 2. **Public.com 智能投资平台**：推介其最新上线的投资 AI Agents。用户可以用自然语言设定交易意图（如“当 VIX 达到 25 时买入标普 500 看跌期权”或“当现金超过 2 万美元时自动转入定向指数基金”），由 Agent 负责监控与执行；同时平台亦向活跃交易者开放完整的公开 API 读写权限。 ### `40:48` 高盛股价破千美元与所罗门的长期盈利增长哲学 对话从华尔街高管的人际交往风格与幽默感切入。主持人提及几个月前对前任首席执行官劳埃德·布兰克费恩（Lloyd Blankfein）的采访，称赞布兰克费恩极为风趣幽默；戴维·所罗门（David Solomon）自嘲地笑着承认自己也是个随和有趣的人，但在讲段子这方面“真的远没有劳埃德好笑”。随后，话题转到了高盛（Goldman Sachs）的股价表现——在高盛从当年报价机（Quotron）时代一路走来的漫长历程中，其股价近期突破了 1,000 美元大关，刷新历史最高纪录。 当被问及每天早晨醒来第一件事是否就是查看高盛股价时，所罗门坦言并不会这样，因为前一天收盘时他早就看过了，且隔夜通常不会发生剧烈变化。工作日里只要他在办公室奔走于各大会议之间，在回邮件和打电话的间隙，交易屏幕总会常开；但比起单纯盯盘自家股价，他更看重的是对整个市场脉搏的全面感知。作为一家全球投行掌门人，他必须对每日市场动向保持高度敏锐，甚至周末也要花大量时间复盘。所罗门强调，关注单日股价波动虽是人之常情，但他真正聚焦的是公司“年复一年”的长期盈利能力：只要脚踏实地进行业务投资、持续推动企业盈利增长，估值倍数（Multiple）固然会上下浮动，但股价自然会给出满意的反馈。 所罗门回顾了高盛自 2020 年 1 月举行首个“投资者日”（Investor Day）以来的具体成绩。过去数年间，高盛团队坚定不移地执行既定战略： - 营业收入自 2020 年初以来累计增长约 65%； - 净利润/盈利（Earnings）实现了惊人的 140% 至 145% 的增长； - 股价从所罗门执掌初期的约 200 美元一路飙升突破 1,000 美元大关； - 公司的业务结构更加多元化、稳固化，来自财富管理等领域的持久性收入（Durable revenue）占比提升，进而获得了市场给予的估值溢价倍数。 展望未来 5 到 10 年，所罗门对技术变革带来的潜力格外振奋。他指出，通过引入前沿技术彻底重构中后台与运营流程，能够将高盛最具生产力的人才从低效繁琐的事务中解放出来，投入更多时间服务客户，全面拓展财富管理等核心版图，为公司释放出更充沛的资本与产能以持续投资于增长。针对主持人提到“当前市场上很多股票走势都呈直线上涨”，所罗门从容回应道，关键在于横向相对表现（Relative performance）——高盛的综合回报率在整个周期中始终处于行业最顶尖（"Top of the heap"）的行列。 --- ### `44:45` 独家操盘 Alphabet 历史级巨额股权融资：科技巨头 AI 资本支出的转折点 随着讨论由公司市值延伸至资本市场动态，主持人抛出了一个重磅话题：在各大科技巨头纷纷依赖发债融资来支撑庞大的 AI 算力与基础设施建设之际，为何像 Alphabet（Google）这样资金充裕的万亿级巨头会破天荒选择进行如此大规模的二次股权增发与定向配售？ ```mermaid flowchart TD accTitle: Alphabet 融资结构转型逻辑 accDescr: 科技巨头从单纯依赖债务转为股债平衡以应对 5 至 10 年 AI 资本支出周期的决策路径。 subgraph Macro["市场环境与资本周期"] A["5 至 10 年超大规模 AI 资本支出计划"] B["公开股市估值倍数处于历史高位区间"] end subgraph Strategy["Alphabet 与高盛的资本研判"] C{"纯债务融资风险评估"} D["若遇经济下行或估值收缩\n资产负债表杠杆将严重承压"] E["高盛 ECM 团队独家筹划 5 个月\n制定长期股债平衡架构"] end subgraph Execution["交易执行与落地"] F["基准筹资 800 亿至 840 亿美元"] G["最终扩容并计入绿鞋机制\n累计募资达 850 亿至 900 亿美元"] H["确立全行业万亿 AI 资本周期的风向标"] end A --> C B --> E C -->|潜在下行风险| D D --> E E --> F F --> G G --> H ``` 所罗门披露了这笔震撼华尔街的交易内幕：高盛作为独家顾问与承销商，在这项极为机密的交易中独立操盘并运作了长达五个月之久，直到交易公布前数天才引入其他投行协同参与。该交易由高盛投资银行及股票资本市场部门（ECM）核心团队领导——包括金·波斯内特（Kim Posnett）、亚斯敏·霍达莫拉迪（Yasmin Khodamoradi）与戴维·路德维希（David Ludwig）等高管： - 这是全球资本市场历史上规模最大（或最具标杆意义）的二次后续股票增发（Secondary follow-on equity offering）之一； - 发行规模由原先的 800 亿美元一路追加扩容（Upsize）至 840 亿美元，若完全行使超额配售选择权（Green shoe，由于二级市场股价交易表现强劲，预期数日内将全部行使），最终募资总额将达到 850 亿至 900 亿美元的惊人量级； - 这一交易成为首个检验全球机构投资者对此类万亿级 AI 规模融资真实胃口的“具象化实体数据点”（Tangible data point）。 所罗门从投资银行家的视角剖析了 Alphabet 做出股权增发的战略逻辑。虽然科技巨头完全具备依靠债务发行的充足信用，但面对未来 5 到 10 年动辄千亿美元级别的超长周期 AI 资本开支，单纯依赖发债将埋下巨大的财务隐患。企业必须从长周期推演：一旦市场环境突变、估值倍数出现剧烈收缩，居高不下的财务杠杆将使公司陷入极其被动的险境。在当前股市流动性充裕、股票估值倍数处于高位的大好窗口期，通过发行股权补充底层缓冲垫是极富远见的风险管理手段。所罗门预言，Alphabet 只是推倒了第一块多米诺骨牌，未来将会有更多跨国巨头在平衡股债天平的过程中重回股权融资市场。 --- ### `48:00` 夺得 SpaceX IPO 主承销权：跨越二十载的深厚信任累积 紧接着 Alphabet 世纪交易的讨论，主持人祝贺高盛再度斩获另一项资本市场桂冠——赢得太空探索技术公司 SpaceX 首次公开募股（IPO）的独家主承销商/牵头投行席位。面对外界对于投行争夺此类超级大单时是否依赖“私信埃隆·马斯克（Elon Musk）”等社交公关手段的调侃，所罗门严肃且自豪地揭示了现代顶级投行赢得核心客户背后的真实逻辑。 所罗门明确指出，SpaceX 的主承销商任命绝非过去六个月里突击公关的结果，而是高盛团队在过去整整 20 年间持续深耕、不懈投入与建立互信的集大成体现。尽管最终的排他性决定是在近六个月内敲定的，且高盛投行核心合伙人团队——包括丹·迪斯（Dan Dees）、金·波斯内特（Kim Posnett）以及苏齐·谢尔（Suzie Scher）等人在此期间展现出了超凡的专业方案设计与执行力，但若没有长达二十年与马斯克及管理层的休戚与共与深度覆盖，根本不可能在这样竞争残酷的角逐中胜出。所罗门回忆，高盛最初与埃隆·马斯克建立业务对接时，负责覆盖马斯克的高盛银行家就已深刻敏锐地预见到，此人必将成为重塑时代的商业巨擘。真正的投行护城河不是单次的低价竞标或短期迎合，而是跨越技术周期与经济迷雾、陪伴世界级领军企业数十载沉淀下来的坚实信任。 ### `50:00` 高盛与马斯克的十五年长跑与投行业务本质 高盛之所以能够拿下埃隆·马斯克旗下核心企业的重磅业务，根基在于一段长达十五年以上的深厚关系。这段渊源最初始于当时高盛合伙人斯图尔特·伯恩斯坦（Stuart Bernstein）深耕绿色能源领域；彼时马斯克与其表兄弟共同创立了 SolarCity，高盛由此与马斯克建立了最初的业务对接。不久之后，高盛主导了特斯拉（Tesla）的首次公开募股（IPO），大卫·所罗门也是在 SolarCity 和特斯拉上市的前后首次结识了马斯克。到了 2012 至 2013 年左右，伯恩斯坦在高盛进入半退休状态，而刚刚在亚洲历练多年回美的丹·迪斯（Dan Dees，现任高盛全球银行与市场业务联席主管）驻扎西海岸并掌管 TMT 投行业务。迪斯接棒与马斯克建立起了极为紧密的私人互信与长期对话机制。 所罗门坦言，大客户关系的维系绝非平坦的单行道，期间双方也曾出现过步调不一致甚至脱节的时刻，但高盛作为一家机构，始终对马斯克、他的公司及其管理团队保持着坚定的长期承诺。在这桩合作中，SpaceX 的首席财务官布雷特·约翰逊（Bret Johnsen）等核心高管同样起到了至关重要的作用，高盛团队多年来与他们保持着高频互动。所罗门强调，投资银行的本质从来不是急功近利的“单笔交易导向（transactional）”，而是通过日复一日提供专业中肯的顾问意见、坚守长期主义来赢得客户的信任；唯有如此，当全球体量最庞大、最重要的交易浮出水面时，投行才有资格被客户选中成为第一排的护航者。 - **关键人物与节点**：斯图尔特·伯恩斯坦早期对接 SolarCity；特斯拉 IPO 奠定合作根基；丹·迪斯自 2012/2013 年起接棒深耕 TMT 客户关系。 - **高盛投行核心哲学**：不以单笔交易成败为目的，注重历经数年甚至数十年的信任沉淀，以换取全球重大历史性交易的主导权。 --- ### `51:40` 澄清传闻：用社交平台私信拉拢 SpaceX IPO？ 面对坊间盛传“所罗门通过 X 平台私信（DM）向马斯克推销 SpaceX IPO 承销业务”的花边新闻，所罗门在节目中直截了当地予以辟谣，称这完全是不实传闻，但他也详细拆解了事实是如何被外界以讹传讹的。实际情况是，马斯克在过去一年中的某个节点彻底停用了手机短信和电子邮件，并告知熟人如果想要联系他，唯一的渠道就是在 X 平台上发送直接私信（DM）。为了维持日常交流，所罗门自然也顺应了这一沟通习惯，转而在 X 上与马斯克探讨各类广泛话题。 当媒体曝出所谓“私信抢单”的报道后，所罗门还特意翻看了自己 X 账户过去半年的全部私信记录，并向高盛公关负责人托尼·弗拉托（Tony Fratto）核实确认。他从未在任何私信中向马斯克开口索要 IPO 业务；高盛能够牵头这项历史性交易，纯粹是因为投行团队脚踏实地的工作表现与长期陪伴。谈及目前担任主承销商的角色，所罗门表示高盛深感自豪，并将与其他合作投行通力协作，为 SpaceX 争取最佳的发行成果。 ```mermaid flowchart TD accTitle: 高盛赢得业务与沟通渠道辟谣 accDescr: 区分媒体谣传的私信推销与真实的高盛业务落地流程 subgraph 传闻假象 F1["外界报道/谣传"] --> F2["所罗门在 X 平台发私信抢单推销 IPO"] end subgraph 事实真相 R1["马斯克停用短信/邮件"] --> R2["日常联络切换至 X 平台 Direct Message"] R3["投行团队数年深耕"] --> R4["紧密服务以 Bret Johnsen 为首的管理层"] R2 -. "仅日常沟通，未谈及 IPO 揽客".-> R5["赢得 SpaceX IPO 主承销商席位"] R4 --> R5 end ``` --- ### `52:45` 超级 IPO 浪潮、未上市公司巨型化与资本主义之辨 随着 Anthropic 提交保密上市申请（播客录制现场甚至被彭博终端弹出的“Anthropic 选定摩根士丹利与高盛牵头 IPO”突发快讯短暂打断，主持人笑称“这就是今天的头条新闻”），市场普遍预计今年将成为“超级 IPO（Mega IPOs）”之年。主持人抛出一个深刻的制度性质疑：如今私募资本市场流动性极强，许多独角兽企业甚至有能力永远不上市；而像当年的微软或苹果，很早就在较小市值时公开上市，即便是普通的指数基金被动投资者也能充分享受其从标普 500 指数第 500 位一路飙升到前 20 位的巨大增值红利。如今巨头企业上市即巅峰，是否意味着公开市场的散户投资者被剥夺了早期红利？这种模式是否正在“玩坏资本主义”？ 所罗门给出了坚决的否定回答，直言“这绝不会破坏资本主义”，而且美国能孕育出全球规模最大、最重要的企业本身就是一件幸事。他指出，公众往往陷入了严重的幸存者偏差： > “大家总是津津乐道于微软上市时的早期红利，却没有人去谈论当年与微软同一年上市、最终价值归零（went to zero）的另外 900 家公司。” 所罗门解释，25 年前微软之所以在很小的体量就选择敲钟，是因为当时的金融市场根本不存在除公开股市之外的规模化资本渠道；企业如果不去公开市场，就无法融到生存所需的资金。而今天的市场与监管生态发生了翻天覆地的变化，不能简单地用过去的模板来套用当下的现实。 --- ### `55:00` 企业的资本饥渴与“尽可能晚上市”的忠告 所罗门回顾了自己的职业生涯，提到自己在 25 年前曾亲自执掌高盛的股票资本市场部（ECM）。几十年来，他给所有创始人和企业掌门人的忠告始终如一：**在条件允许的情况下，尽可能晚地走向公开上市**。在他看来，企业上市无非出于三个实质性诉求： 1. **急需资金（Need the capital）**：业务扩张需要公开市场的巨量资金回血； 2. **需要股票流动货币（Need the currency）**：用于展开大规模对外兼并收购（M&A）； 3. **解决退出诉求（Liquidity for early investors）**：为早期风险投资人与创始员工提供流动性变现渠道。 所罗门警示，除上述硬性诉求外的其他上市借口往往无足轻重。更关键的是，管理一家上市公司与管理一家私营公司的逻辑截然不同——公开市场伴随着季度业绩压力、严苛的外部审视以及完全不同的运营生命周期；一旦敲钟，掌舵者必须彻底改变运营方式。当前的高增长企业通过内生方式快速扩张，对兼并收购货币的需求并不迫切，加之私募股权市场长期资金充沛且变现手段多元，促使企业不断推迟上市。然而，如今这批科技巨头之所以集中重返 IPO 舞台，本质是因为它们发展到了极其惊人的体量，其对资本的需求是极其贪婪而庞大的（voracious capital needs），如果百分之百依赖私募市场且没有公开上市的股票工具，在财务和战略上都已不再审慎。 - **监管与市场摩擦**：现行监管体系与合规成本使得公开上市缺乏吸引力，虽然现任 SEC 主席正着手消除上市摩擦，但市场底层逻辑已变。 - **被动投资与小票折价**：在被动指数基金和 ETF 占据统治地位的今天，一家市值 50 亿美元、募资 8 亿美元的中小型企业在公开市场面临流动性折价，相比之下此前私募资本给出的溢价更有吸引力。 - **资本市场的周期性与终局**：私募资本与公开市场的折价空间存在周期性轮动；但从长期结构来看，未来公开市场的上市企业数量注定会更少，而留存与新上市的企业体量则会越来越庞大。 ### `01:00:15` 市场是处于贪婪还是恐惧？美股估值与历史泡沫的对比 在纽约经济俱乐部（Economic Club of New York）的发言中，大卫·所罗门（David Solomon）提到当前市场处于“贪婪”阶段而非“恐惧”阶段，引发了大量媒体头条。主持人指出当前资本充裕甚至显得疯狂，询问早在 1980 年代初就进入华尔街的所罗门，是否曾经历过类似的市场景象，当前周期在历史上有无先例可循。 所罗门澄清，有人对他说标普 500 指数（S&P 500）前十大公司的市值权重占到 30% 多是“史无前例的”，但事实并非如此。无论是在 1920 年代、1960 年代（如 Nifty Fifty 时代），还是 1990 年代末至 2000 年的互联网泡沫时期，龙头公司的市值集中度都曾达到 30% 多。人们因害怕错过（FOMO）计算、存储、芯片等领域的科技浪潮而蜂拥涌入少数几只股票，这确实是一种贪婪行为。 然而，所罗门强调当前的集中度在本质上与过去有巨大不同： - **前十大公司盈利能力极强**：当前标普 500 头部公司的盈利水平和现金流产生能力远胜以往同等市值的龙头企业。 - **前瞻市盈率（Forward P/E）对比**： - **当前前十大公司**：前瞻市盈率处于 30 多倍的低位区间。虽然位于历史分布的高位，但在其高成长和坚实利润支撑下并不算离谱； - **1990 年代末互联网泡沫**：龙头公司普遍在 40 倍甚至 50 倍以上交易，思科（Cisco）市值曾被推升至超过 6000 亿美元； - **历史极端案例**：1960 年代的宝丽来（Polaroid）估值高达 80 倍市盈率，1920 年代的美国无线电公司（RCA）市盈率甚至近乎无穷大。 - **其余 490 只成分股的吸引力**：标普 500 其余 490 家公司的前瞻市盈率大约在 17 到 20 倍之间（许多公司仅 17 倍）。如果这些企业能在未来五年内通过运用 AI 与新技术改善运营流程、提升效率并加速利润增长，17 倍的估值显得相当具吸引力。 所罗门总结道，市场在窄幅权重股的带动下确实跑得很快，而且由于基本面并未脱轨，这种涨势可能还会持续相当一段时间。但宏观层面上，地缘战争、石油冲击、供应链紧张以及粘性通胀等潜在风险依然存在，目前市场只是暂时选择将其忽略。 ```mermaid flowchart TD accTitle: Market Valuation and Breadth Comparison accDescr: S&P 500 top 10 companies compared against the remaining 490 companies on forward PE and earnings power subgraph Top10["标普 500 前 10 大龙头股 (~30%+ 市值)"] A1["前瞻市盈率：30 多倍出头"] A2["真实高盈利与强劲现金流"] A3["投资者因 FOMO 情绪扎堆抱团"] end subgraph Rest490["标普 500 其余 490 只成分股"] B1["前瞻市盈率：17 - 20 倍 (许多仅 17 倍)"] B2["潜在驱动力：借助 AI 提效、降本、加速盈利"] B3["估值具备长远吸引力"] end subgraph MacroRisks["宏观潜在变量 (目前被市场忽视)"] C1["地缘冲突与战争"] C2["石油冲击与供应链受阻"] C3["通胀韧性与粘性"] end ``` --- ### `01:04:47` 尖端前沿模型与国家关键基础设施的网络安全 高盛参与了 Anthropic 的 IPO 承销工作。主持人借机问及针对前沿模型安全测试的相关项目（如 Mythos、Project Glasswing 等），外界有人质疑向少数精选机构小范围开放前沿模型是因为“算力不够”，主持人探问高盛在测试中究竟发现了什么，网络安全团队是否因此夜不能寐。 所罗门从公开监管和行业协同的角度做出了回应： - **前沿模型威力惊人**：必须要求开发顶级模型的科技公司与政府机构及私营部门保持紧密协作，让外界充分理解模型能力，以保卫国家关键基础设施。虽然当前沟通测试的流程尚不完美，但一直在摸索迭代。 - **金融体系的防御差距与系统性风险**： - 过去十年中，全美排名前八的大银行在网络安全上的投入和防护水平，远非阿肯色州偏远地区排名第 2000 位的小型银行可比。 - 大银行虽然也随时面临风险、必须持续不断投资并严阵以待，但更大的系统性隐患在于中小型机构。正如硅谷银行（SVB）暴雷后迅速在全行业引发挤兑与恐慌一样，一旦某家中型银行遭遇毁灭性网络攻击，极易引发波及全行业的流动性和信心动荡。 - **跨行业的防御鸿沟**：这种脆弱性不仅限于金融领域。例如能源公用事业基础设施，最大的公用电力巨头防御成熟，而俄亥俄州阿克伦（Akron）的地方水务公司却往往缺乏应对尖端网络攻击的资源与技术。政府行政令推进模型安全审查和政企合作至关重要，整个社会必须想办法帮助那些资源不足的组织抵御风险。 --- ### `01:06:54` 音乐制作的民主化：从布鲁斯·斯普林斯汀到 AI 与编曲软件 在节目的最后部分，话题转向了所罗门的个人爱好——作为一名知名的电子音乐 DJ 与音乐制作人，他如何看待并使用 AI 生成音乐。 所罗门直言自己一直在亲自把玩和测试 AI 生成音乐。他认为这类技术最酷的地方在于**让创作渠道彻底民主化**，向那些过去永远没有机会制作音乐的人敞开了大门。为了说明这种技术变革的意义，他回顾了音乐制作的演变历史： - **传统模拟时代（以布鲁斯·斯普林斯汀为例）**： - 1978 年所罗门看了人生中第一场斯普林斯汀（Bruce Springsteen）的《Darkness on the Edge of Town》巡演。 - 在那个时代，创作者必须坐在钢琴前或抱着吉他，在纸上手写音符和歌词，然后召集真实的乐手（钢琴、管风琴、架子鼓、吉他、管乐、萨克斯等）齐聚录音棚，现场演奏录制，最后压制成黑胶唱片。如果不会演奏和手写乐理，普通人根本无法独立制作音乐。 - **数字音频工作站（DAW）时代（Logic Pro 与 Ableton）**： - 所罗门透露自己小学和高中吹了 12 年萨克斯管，懂得五线谱和乐理，大学毕业后当过业余鼓手，能打稳 2/4 拍基础节奏。 - 但在三四十年前，他完全不具备独立制作发行歌曲的条件。随着现代音乐制作软件（如 Logic Pro 和他常用的 Ableton）的普及，音乐制作被大幅平民化——编曲者可以在软件内自由调取 150 种不同的虚拟管乐音色，而不再需要雇佣一个管乐声部团队。AI 技术的加入正在将这扇创造力的大门推得更宽。 ### `01:10:00` 音乐制作的演变、AI 版权争议与人的“独特声线” 谈及音乐制作技术的变迁，大卫·所罗门（David Solomon）回顾了技术如何彻底重塑创作流程。在过去，制作一首曲目可能需要召集 15 位不同的乐手；而随着数字音频工作站（DAW）如 Logic Pro 和 Ableton 的普及，制作门槛发生了根本性改变。所罗门自嘲自己在这些软件平台上其实“相对手残”，虽然能跌跌撞撞地摸索，但仍需要专业人员从旁协助，因为“专家三秒钟就能搞定的事，我得花上十分钟”。但这类技术的核心价值在于解放了表达：创作者不再需要手写五线谱，只需在键盘上弹出一组和弦走向，用电脑试听，甚至只需轻点一下鼠标，就能将钢琴音色一键替换为管乐，迅速搭出极具吸引力的旋律主线（top line）。当今大众听到的很多流行音乐，正是通过这种方式制作出来的。 如今，新型 AI 应用正在将这一流程进一步自动化，但在所罗门看来，这项技术带来了两个必须正视的核心矛盾： 其一是版权与创作者报酬机制的缺失。现有的 AI 音乐应用本质上是在未经授权或未予补偿的情况下，大量抓取和使用人类艺术家的既有作品。如何将艺术家群体及其知识产权（IP）纳入讨论与利益分配机制，是一个亟待解决却尚未引起足够重视的关键议题。 其二是创作者不可替代的“个人声音”（voice）。所罗门将其与此前谈到女儿编剧的观点联系起来：无论是在金融界还是艺术界，真正打动人心的作品必须带有鲜明的主观意志。在音乐制作中，关键在于制作者明确表达“我喜欢这种听感，就这样做”或“我喜欢这种情绪氛围，咱们换这种方式试试”，随后将其推向市场检验受众反馈。虽然技术极大提升了生成速度，但如果缺乏人类在细节上的打磨、调优与克制，最终产出的往往只是平庸之作，而非具有持久共鸣与生命力的真正艺术。 - **数字音频工作站（DAW）的技术跃迁**：Logic Pro 与 Ableton 让非传统科班编曲者无需手写乐谱，即可通过快速切换虚拟乐器音色（如钢琴换铜管）搭建复杂旋律。 - **AI 音乐面临的双重挑战**： 1. **知识产权（IP）补偿危机**：AI 模型深度依赖人类既有版权内容训练，但原作者目前完全未获得相应报酬。 2. **审美平庸化（Mediocrity）风险**：AI 可以迅速生成素材，但缺少人类主观审美的微调与情感注入，生成内容难以产生深度共鸣。 --- ### `01:12:12` 访谈尾声：斯普林斯汀的“市场见底指标”与人际交往的稀缺溢价 在访谈结束后的复盘环节，主持人乔·威森塔尔（Joe Weisenthal）与特雷西·阿洛威（Tracy Alloway）交流了对大卫·所罗门的印象。特雷西对所罗门在音乐行业展现出的深入思考感到惊喜，原本她以为这位华尔街巨头只会把话题局限在银行业务。乔则特别惊叹于所罗门对历史事件的具体年份、交易细节以及关键合作者姓名的惊人记忆力，认为这展现了顶尖制作人和资深银行家共有的特质——准确铭记每个人的贡献并给予信用（credit）。 两人还幽默地复盘了所罗门提及的布鲁斯·斯普林斯汀（Bruce Springsteen）巡演轶事。乔指出，斯普林斯汀的经典专辑《Darkness on the Edge of Town》在 1970 年代末发行时恰逢美股历史大底，而该专辑在 2010 年再版重发时，又惊人地契合了次贷危机后的世代性低点。乔调侃道，这堪称金融市场最灵验的反向买入指标：“等斯普林斯汀第三次再版这张专辑时，我们就知道市场真正见底了。” 回归到 AI 对白领就业的冲击，乔与特雷西提炼出了贯穿本期节目的核心主题： ```mermaid flowchart TD accTitle: AI时代人际交往能力的溢价逻辑 accDescr: 展示知识大宗商品化与移动设备普及如何共同推高面对面人际沟通技能的稀缺溢价。 subgraph MacroTech["技术与知识层"] A["知识与经验高度大宗商品化<br>（AI 普及降低专业认知门槛）"] B["智能手机与移动互联网普及<br>（侵蚀年轻一代面对面交流习惯）"] end subgraph SkillShift["核心能力演变"] C["传统分析与案头技能贬值"] D["面对面沟通与眼神交流能力急剧稀缺"] end subgraph MarketOutcome["资本与职场新常态"] E["线下社交、信任建立与人脉拓展<br>成为最高溢价技能（Social Maxing）"] F["公开上市公司转向非公开私募股权融资<br>（如 Alphabet 巨额募资预示新趋势）"] end A --> C C --> E B --> D D --> E E --> F ``` - **知识商品化推高社交溢价**：当行业知识、研报撰写和代码生成都能被 AI 工具一键提供时，冷冰冰的标准答案便不再稀缺。能够直视对方眼睛沟通、建立信任并促成复杂交易的“人类温度”（human touch）反而成为最值钱的技能。 - **注意力危机加剧技能分化**：过去十五年智能手机的普及严重损害了许多年轻职场人进行面对面眼神交流的能力；技术造成的这种现实社交障碍，反倒让善于社交、敢于拿起电话直接对话的人享有更高的稀缺溢价。 - **资本市场新动态**：特雷西特别强调了所罗门关于“更多上市公司将开展私募股权融资”的判断。Alphabet 的巨额私募资本运作引发轰动，而所罗门预言这只是未来一连串类似交易的开端，预示着公开市场与私募股权界限的深度重塑。 ## Notable quotes > `05:00` 我是高盛 CEO，AI 引发的就业末日论被过度夸大了 — David Solomon > `07:30` 我是贝尔斯登的顶尖创收者（Big Producer），在那里建立了极为庞大的客户人脉，并把这些关系带到了高盛 — David Solomon > `10:00` 去试两年，如果不成，两年后再去读商学院。 — David Solomon 的父亲 > `25:00` 电话是世界上最伟大的技术之一，多用它 — David Solomon > `35:00` 我认为答案是永远不可能（never） — David Solomon > `35:00` 除非那个模型能够访问她曾经写下的每一个字、经历过的每一次人生体验、感受过的每一种情绪……但模型永远不可能拥有这些。在创作与写作中，独特的‘声音’至关重要。 — David Solomon > `52:45` 大家总是津津乐道于微软上市时的早期红利，却没有人去谈论当年与微软同一年上市、最终价值归零（went to zero）的另外 900 家公司。 — David Solomon > `55:00` 在条件允许的情况下，尽可能晚地走向公开上市 — David Solomon > `01:10:00` 专家三秒钟就能搞定的事，我得花上十分钟 — David Solomon > `01:12:12` 等斯普林斯汀第三次再版这张专辑时，我们就知道市场真正见底了。 — Joe Weisenthal ## Mentioned David Solomon, Goldman Sachs, Tracey Allaway, Joe Weisenthal, Marco Argenti, Bear Stearns, Irving Trust Company, Drexel Burnham Lambert, Merrill Lynch, EF Hutton, Dean Witter, Quotron, SecDB, The New York Times, Rory McIlroy, Tiger Woods, Jack Nicklaus, Nick Faldo, Lloyd Blankfein, Alphabet, Google, Kim Posnett, Yasmin Khodamoradi, David Ludwig, Elon Musk, SpaceX, Dan Dees, Suzie Scher, Stuart Bernstein, SolarCity, Tesla, Bret Johnsen, Tony Fratto, Anthropic, Morgan Stanley, Microsoft, Apple, SEC, Bruce Springsteen, Darkness on the Edge of Town, Logic Pro, Ableton, Silicon Valley Bank, S&P 500 ## Video description <p>There's a lot of debate about the future of AI &mdash; not just whether it will produce the returns investors are expecting, but also if AI will lead to mass worker displacement. Big banks are the perfect prism through which to explore some of these questions. Not only are they deploying AI very quickly, but they have a wide range of workers who are using the technology, from back-office employees to junior analysts to the most senior investment bankers. In this episode, we speak with David Solomon, chairman and CEO of Goldman Sachs, about the impact of AI on the banking business, and why he does not predict a major white collar wipeout. We talk about the outlook for headcount, current conditions in capital markets, and the bank's role in the upcoming SpaceX IPO and Alphabet's historic equity capital raise. He also tells us about his early career in junk bonds and (because of his love of electronic dance music) how AI is transforming music production.</p> <p>Read: <a href="https://bloom.bg/4uKjcE1">Anthropic Picks Morgan Stanley, Goldman Sachs to Lead IPO</a></p> <p><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/bf5b0ae1-c442-4354-8c4f-b45f00126aa1/audio.mp3?in_playlist=982f5071-765c-403d-969d-ae27003a8d83)

[5:00] David Solomon: 我是高盛 CEO，AI 引发的就业末日论被过度夸大了 [7:30] David Solomon: 我是贝尔斯登的顶尖创收者（Big Producer），在那里建立了极为庞大的客户人脉，并把这些关系带到了高盛 [10:00] David Solomon 的父亲: 去试两年，如果不成，两年后再去读商学院。 [25:00] David Solomon: 电话是世界上最伟大的技术之一，多用它 [35:00] David Solomon: 我认为答案是永远不可能（never） [35:00] David Solomon: 除非那个模型能够访问她曾经写下的每一个字、经历过的每一次人生体验、感受过的每一种情绪……但模型永远不可能拥有这些。在创作与写作中，独特的‘声音’至关重要。 [52:45] David Solomon: 大家总是津津乐道于微软上市时的早期红利，却没有人去谈论当年与微软同一年上市、最终价值归零（went to zero）的另外 900 家公司。 [55:00] David Solomon: 在条件允许的情况下，尽可能晚地走向公开上市 [70:00] David Solomon: 专家三秒钟就能搞定的事，我得花上十分钟 [72:12] Joe Weisenthal: 等斯普林斯汀第三次再版这张专辑时，我们就知道市场真正见底了。

[11:00] David Solomon: One of the great opportunities we have with this technology expansion is to get our young people out with clients, talking to clients, broadening our client footprint... This shifted over the last 30, 40 years. And it really, I don't think it shifted for the better. And now we're going to shift it back. [25:00] David Solomon: The telephone is one of the greatest pieces of technology in the world. Use it. [30:00] AI Model: No, that's not the way it works. [40:00] David Solomon: I'm really not as funny. I mean, he's much funnier than I am. [53:15] David Solomon: I don't think this is going to break capitalism.

[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. and it's going to be at Soho House. That's right. So if you would love to prove your knowledge of all things markets finance and economics if you want to meet myself and Tracy if you want to hang out with fellow listeners please, we'd love to see you at our Odd Lots Trivia Night on June 11th in Hong Kong at Soho House Hong Kong. And you can find tickets. Go to Bloomberg.com slash Odd Lots. You'll find a link there to our live event. And you can come purchase tickets. Just 400 Hong Kong dollars. I assume that's a good price. It's a bargain. and we hope to see you there. Bloomberg Audio Studios, Podcasts Radio News. Hello and welcome to another episode of the All Thoughts Podcast. I'm Tracey Allaway. And I'm Joe Weisenthal. So, Joe, when we talk about AI and its impact on jobs and businesses, one of the things we like to do or one of the prisms through which we like to explore AI is through banks. Absolutely. So you think about banks. They've got big tech budgets. They talking to AI companies all the time They their clients essentially So they deploying a lot of this technology And then they have this whole array of jobs right You have the classic sort of process workers in the back office You got the junior analysts who are churning out PowerPoints but you also have bankers in IB who are doing more I guess relationship business So you have this really good cross of the sort of white jobs that people think about when we think about AI disruption Do you remember in the 2010s when every bank CEO is like, you know, we actually like to think of ourselves as a tech company. I remember. That was a fun. Bring in the ping pong tables for the engineers. That's right. That was a fun little era. But no, you're absolutely right. And of course, you know, AI is largely expected, at least until we have much more advanced robotics, expect to affect white collar labor first. You mentioned the whole range. But there are aspects of when I think about being a banker working a bank that to your point really do feel like relationships and like there's the cliches like go out and golf. But I think they do actually golf a lot. I think that is part of the business. All right. Well, we'll find out. We have spoken to Marco Argenti at Goldman Sachs before he's the CTO about how the bank is actually deploying technology. But today we have an even more more perfect. We have the perfect that's possible. We're going to be speaking, of course, to you. [5:00] David Solomon, he's the CEO and chairman of Goldman Sachs, and he just penned an op-ed in the New York Times with the headline, I'm the CEO of Goldman Sachs, the AI job apocalypse is overblown. So truly the perfect guest to talk about all of this. David, thanks for coming on All Thoughts. Thank you for having me. I'm delighted to be with you guys. Congrats on the op-ed. You're in the media industry now. Well, I'm trying not to be in the media industry. I'm glad I wrote the op-ed. By the way, not my title, the New York Times title. But I'm glad I wrote the op-ed. Why did you write it? Well, I wrote the op-ed. I haven't written a lot of op-eds. I think in my eight years, I've written two. And the last one was back in 2019 before COVID. And it was around energy, where I was trying to talk about the fact that I thought there was going to be real change in the context of the energy supply chain. And it was very important that we made investments in things that were greener and we moved in that direction. But I also stated at the time, and took a little heat for this very emphatically that oil and gas was going to be a significant part of the energy supply chain for decades to come. Good call. Goldman Sachs would be financing fossil fuel companies for a long, long time. And you know in looking at the lens now I think that it is very important that we have a thoughtful and detailed discussion about the change that going on And let just start if you don mind if I can step back I am hugely hugely optimistic about this technology and the impact that this technology can have on large-scale enterprises like Goldman Sachs and the productivity gains that it can create, which I do believe can increase economic growth, create more prosperity and more impact on our economic system broadly in a very, very positive way. I am hugely optimistic. And I think about things in five to 10 year time bites, like what should we expect to see over the next 10 years? That's the way I think about things when I make a statement like that. At the same point, I'm extraordinarily cognizant about the fact that it's not going to be a straight line. There's going to be disruption. There are going to be jobs displaced, as there have been with any other technology acceleration over the course of history. And we as an important company and governments also do to play a role in thinking about what's the best way to make this go as well as possible. But with the ultimate goal of having the technology make our society more productive and therefore allow for more economic growth and more participation by everyone. And I was listening to a variety of things that were being said and they didn really make sense to me And I thought it was important for Goldman Sachs to get into the discussion And so there are things that are always different But I generally you know if you want to talk about kind of fundamental principles that I have after 40 some years of doing this it never different this time meaning I don't think we're going to wake up in a world where nobody works and there has to be universal basic income and we have massive unemployment. I just don't think that's the way our economy works. And when people start talking about those things or postulating about those things, I don't think they're being thoughtful about really looking in a granular way at the labor force and how job creation works in our economy and how technology disruption in the past has changed and shifted the nature or the makeup of labor and the economy in the U.S. so i think to even begin the discussion of like will ai cause a white collar wipeout or as you did say in the op-ed like some some areas will really change and headcounts will change but to begin this conversation you have to know like well what is a job what does someone actually do at their job and so i don't maybe this is a chance i'll let you like butter you up let you brag about yourself for a little bit i saw a quote i was doing some prep and you're talking about early in your career and going from Bear Stearns to Goldman And you said I was a big producer at Bear Stearns I built a lot of relationships there and brought a lot of relationships to the firm What made you good at your job What was like there were a lot of people there at Bear at the time You eventually became the CEO of Goldman Sachs. What was it that like the skill that you've had or that successful people in that role had that made them do their jobs well? Well, finance has evolved. I started in 1984. I recently referred to that as the prehistoric days of finance, when, by the way, I had long flowing hair. It was just a very different world. And so let's just start with the fact that finance was a relatively nascent industry and also had been relatively stagnant for 15 years because we really went through a period from the late 60s to the early 1980s where interest rates were going straight up. September 15, 1982 is a very, very important day in the history of financial markets because that is the day that the 10-year treasury hit 15.9%. And it also happens to be a little less than two years before. I landed at the Irving Trust Company at One Wall Street to start my career. And at the time, you got trained to do analysis, but the business [10:00] was very entrepreneurial. There was very little structure and there was so much open running room to go out and participate that young people, especially when you got away from the most white shoe firms, which at the time were Goldman Sachs and Morgan Stanley, were really encouraged to go make their way. And so I was working at Irving Trust and this was a time when almost everybody in finance went to business school. I was applying to business school in 1985, and I got a job offer to work in the jump bond business at Drexel Barnum. And I decided to take it because it appeared lucrative. And my dad- Jump bonds were like the only thing taking off, basically, in finance in the early 80s, right? It was definitely an area of finance that was accelerating, and it was very entrepreneurial. and I said to my dad, this is taking a risk because everybody else is going to business school. And my dad said, go try it for two years. If it doesn't work, you go to business school in two years. It was a pretty simple piece of advice. But I went to Drexel Burnham. It was very entrepreneurial. And there were good things and bad things about the culture at Drexel And some of the bad things are the reason why in 1990 on February 14 1990 Drexel Burnham went out of business as many other Wall Street firms have in the history of Wall Street But for the four and a half years that I was there, basically, I was told, go find a way to build relationships and make money. Go find a way to build – I was 24 years old. Go bring business into the firm. Did you golf? We sat at a desk. I golfed a little bit, but we sat on a desk, and we called people and talked to people. And it was a time when it was a time in the world when if you picked up the phone and called someone, you would get them on the phone and you would talk to them. And if you were an effective salesperson, you could do a lot of business. So to some extent, you were like good at sales. You were good at making that call. One of the great experiences I have, there's an experience I've talked about a little bit, but I haven't talked about a lot. But a seminal experience in my personal development was the summer after my freshman year in college. I got an internship in a Merrill Lynch office. That happened to be the office that was, I think it was like 10 Penn Plaza. It was by Madison Square Garden. And I got the internship because my high school girlfriend's father ran that Merrill Lynch office. That always helps. And he said why don you come and do an internship for the summer And I said what would you do And he said you call on clients I really had no idea what he was talking about So I show up at this Maryland office and they plop me down and they give me a zip code list Okay, basically, and it was the zip code was 10028 and 10128, basically the Upper East Side of Manhattan. This is 1981, okay? They give me a zip code list and they say, make a hundred phone calls a day. and ask people if they're interested in talking to a Merrill Lynch broker. And during the course of the summer, okay, for eight weeks, I made 100 phone calls a day. I got five people to talk to a Merrill Lynch broker, and we went and had one meeting and opened one account. Oh, my God. Okay. It was brutal. Brutal. And after the first week, I said to my dad, I was like, I don't know that I can do this for the whole summer. And he said, yes, you can. You took it on. You got to finish it. And I started thinking about, and this is the seminal thing that I'm trying to point to, I started to think about how am I going to make this more fun? Because it is very difficult to cold call. Now, this was a time when people would answer the telephone, and I just started making it a game. But what did it teach me It taught me to pick up the phone and talk to anybody And when I got to Drexel Burnham and here there were real financial incentives for picking up the phone and talking to people. I mean, they weren't paying me anything. And they just wanted me to do this. And I was doing it to learn. You put financial incentives in place, I found I could call anybody. I could get people on the phone. I could keep people on the phone. I could go see people. I could relate to people, even though I was very young. And I could bring business into the firm. And very quickly, I became good at that. And when I arrived at Bear Stearns after Drexel Burnham went out of business, I very quickly became a very significant producer and a culture that was kind of eat what you kill. And it was a great experience. I don't think that's the way, I don't think it was the perfect culture. And culture has evolved, but I do think there's something here. And one of the things we're talking about a lot at Goldman, I think it's really important. I still think human to human contact matters a lot. And I think one of the great opportunities we have with this technology expansion is to get our young people out with clients, talking to clients, broadening our client footprint. They're incredibly capable. They're incredibly capable. This shifted over the last 30, 40 years. And it really, I don't think it shifted for the better. And now we're going to shift it back. And I think it's a real opportunity. Wait, say more about this, because in your op-ed, you talk about the Stanford data showing [15:00] like a 16% decline in entry-level hiring. So this is something that's happening right now. And presumably, a lot of the work that younger analysts at Goldman Sachs do right now could be commoditized through AI, right? If I'm building a PowerPoint presentation, I no longer need like 10 analysts to stay up all night to do that. I can just generate it through a platform. So how do you see the role of junior analysts? And then how do you, I guess, train the next generation of banking talent at Colby? I think that's a really, really good question, and I want to get to that, but I just want to comment on one of the things that you said that I think is a really interesting thing for us to think about, because I did quote that statistic, but I think it's important for us to think about and see that statistic for what it is, okay? Because when I graduated from college, it was very, very hard to get a job, but not everybody wanted to get a job, in the white-collar professions that are defined by the 16%. But I'd also argue even today, if you graduate from a liberal arts college and you want to go be a school teacher, that not counted in the 16 There not a decline What they done is they taken a handful of industries that they looking at that fit this kind of social narrative of what the right kind of job is when you coming out of college, okay, which by the way, I think is a very faulty lens to look through. And that group is declining by 16%, okay? But if you want, what if you want to go buy, what if you graduate from college and you say, you know what, I really think I can do something really cool and entrepreneurial by making the world's best sub sandwich. I'm going to go open a store and start a sub-business and try to grow a sub-business. And over the next however many decades, you grow a huge sub-business. Joe used to be in the sandwich business. Were you in the sandwich business? The sandwich business. After college. That's a nice way of saying I had a job at a deli. There are so many different things you can do. I've got great friends I went to college with. I have one friend that I went to college with that basically graduated from college. He had studied science, you know, physical sciences. And he decided he wanted to be an electrician. And he had his own electrical business in upstate New York where I went to college. But then ultimately, he got pulled into other kinds of businesses where the skills he had developed over 10 or 20 years building his electrician business could be put into other things. And his career evolved. And so there a narrative about jobs and entry level jobs And I don even know what it all means What I know is you want to be a learner through life You go to school to get educated and to broaden yourself and to meet people Not everybody should go to a traditional school. There should be lots of different paths. But the goal is to start on the journey of figuring out what you're passionate about, what you enjoy, how you're going to make money, how you're going to support yourself, how you're going to support, hopefully, the family that you have. And there are lots of different paths. So lots and lots of different paths. OK, but assume that we're talking about someone who wants to get into Goldman out of college. Yeah. And I don't mean to I don't mean to pontificate. No, no, no, it's a really important thing because this goes to the op ed and kind of the labor force. There's a lot of the labor force that has nothing to do with what we're talking about. And I think that's something that we've got to keep coming back to in the discussion. So, you know, coming to Goldman at this point in time, we hire we just right now. Okay, we just have 2,500, it might be 2,400 interns starting. Wow. And in July, we have approximately the same number of permanent new hires that are starting. The first thing is there are a bunch of them that are in quote client jobs and roles There are a bunch of them that are in operational roles There are a bunch of them that are in technology and engineering roles There a wide wide range of jobs I mean there are a couple of them that are working in marketing, okay, and advertising. I mean, there's a wide, wide range of jobs. All the experiences are different. All the opportunity sets are different. When I was talking to you about out with clients, I was talking about the people that are in the client-facing roles. And one of the things I truly believe in, look, I see this. I sat with a group of venture backed founders at dinner in my apartment the other night. The average age was like 28. So impressive. So smart. By the way, the people who are hiring Goldman Sachs are so impressive. So smart. They're peers with these people that are going and starting businesses. They can't go out and talk to clients and have an impact. And so with the technology, it makes it more leveraging. And I think you guys saw it because I've used this example of when i started doing a common stock comparison took six hours yeah go microfiche wall street journal the only source of historical stock prices put it on graph paper take a picture of it if you really think about that example what that example is actually saying is if you were a company and you wanted to know where your stock was trading during the day yeah you had two choices you could call it with three choices i guess technically [20:00] You could call your investment banker and say, hey, where's my stock trading right now? And the fact that we had a Quotron in our office, okay, we could tell them. They didn't know. They didn't have a Quotron. Or you could call your stockbroker if you had one or go outside your office and find a Merrill Lynch or an EF Hutton or a Dean Witter office on the street and look at the ticker tape and wait until your stock would buy on the ticker tape. Okay. Or you could wait until the next day and open up the Wall Street Journal and see where your stock traded that day. So we were offering a huge value service by simply having this technological machine that allowed us to tell them where their stock was trading. So if you think about what young people were doing when I started, a lot of what we were doing was taking very primitive technology and giving information to people that were trying to run their businesses. And the pace of everything was very, very slow. We've now, over 40 some years, accelerated that enormously. And so the productivity level of all these people's going up, up and up and up. We haven't had fewer people. Now, maybe this technology is accelerating in a way where over time we will actually fundamentally have fewer people. But what I would argue is the real challenge for us is we got to find ways to apprentice them and teach them a variety of things that they not going to intuitively learn because they don actually have to work as hard to get the answer When I had to go to the microfiche and look at the different stocks and see how they were trading actually put it on graph paper and absorb the differences and do the math to explain what the percentage compounding differences were, I was learning something that now, if you ask for it, you get it instantaneously. Has your brain really absorbed what's actually happening? And so the challenge is, how do we apprentice and give people the base of knowledge, but how do we also allow these tools to allow them to get out into the world so that they can really have a bigger impact faster? And we don't have the answers to this yet, but we're really thinking about it. Thank you. so you can find the right fit 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 really watching the bottom line you 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 times, 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. For certain. Support for the 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 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 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 by Public Advisors, LLC, SEC Registered Advisor. Complete disclosures available at public.com slash disclosures. Let's talk about health care for a second. It doesn't always work the way people expect it to. If you've ever waited on a prescription refill or had a hard time getting the care you needed, you know the feeling. The system should just work better for everyone. That's exactly what the people at Optum are trying to do every day. They're a health care company linking patient care and pharmacy services and using data and technology to drive the whole system so care is connected, not complicated. for patients and providers. Things like making it easier to get care that looks at the whole person, from primary care doctors to mental health support and even in-home care, and then using technology to make sure they all work together. Technology designed to help doctors spend less time on busy work and more time with their patients. And those prescriptions? Optum is working to bring costs down, save patients money, [25:00] and make it easier to get refills. Little by little, Optum is helping make healthcare work as one for everyone. Head to business.optum.com to see how. Two sort of short job-related questions. First one, you can always say yes or no. It's like those skills that you had, you were very comfortable picking up the phone and so forth and getting maybe people to answer. Today, is there a role for that person still? That same basis? Definitely. Okay. That's what I figured. By the way, I don't think that's going to change. You guys can go look. I gave a commencement speech at Wharton last month, and one of the lines in the commencement speech was, the phone, the telephone is one of the greatest pieces of technology in the world. Use it. A telephone call to someone is 10 times more valuable than a text or an email. My daughter says that's an unverified statistic, but I know that's true. And it is. If I call you personally on the phone, it's going to have a completely different impact than if I sent you a text or an email. But I don't know your number. I would also accept a text from David Solomon. Just as the follow up. And this is more of a sort of specific. That's not going away. I mean, I really think, Joe, that's not going to go away. Of that mix of new hires the range of jobs including engineering jobs including Can you identify is there a change different between today and 2026 versus say the pre GPT era 2021 Well in terms of like have you seen a difference yet just in terms of like the allocations of anything with any hiring decisions? Subtle changes, but you can't look at it in a vacuum. So you said to me, I can be very detailed. Yeah, please. You know, I think with all this stuff, it's nuanced details matter. If you look over the last 10 years, I'd say what's happened is the amount of engineering talent we had at the firm grew materially on a relative basis to operational talent or client talent. Okay. There are a variety of things that happened because of the pandemic that bloated or inflated hiring across all large enterprises. Yeah. And so if you say the cinch check GPT, like if you take 2022, that was like the high watermark of the bloat. This is the problem. The high watermark of the bloat. Historians will be vexed by the confounding variables of 2022 for a very long time. Absolutely. So if you ask me a question, I'm like, okay, 2022. Yes, but that's the wrong starting point. Because if you look at 2022 a bunch of what we done from 2022 to 2025 and now in 2026 was correcting the stupid things that we did in 2020 and 2021 And so I think you've got to look at it through a 10-year lens. And what I would say is, at this point, if you look at a 10-year lens, it shifted to more heavily weighted engineering. My guess is that's going to shift differently going forward, given the power of these tools and our ability to code. And you're going to see nuanced changes that probably to some degree reduce the number of people that we start with over the next few years, but probably not what you and I would call dramatically. So I don't want to be predictive because if I throw out a number, I hear everybody's going to start holding me to the number. But I just told you we were running it at 2,500, which, by the way, looks similar to what we were running at pre-COVID. but lower than the 3,000 plus we were running in 2021. Okay. My guess is it's gonna be in the next three years that will contract a little, but not in a way when you and I would say, wow, this is a huge, huge change. We're still gonna hire a lot of people out of school. I mean aside from potential headcount reductions and like cost savings how do you actually measure productivity gains at Goldman That is a great question Tracy I can tell you some examples of things we doing where I know we'll be able to quantify some significant productivity gains, but they're more in the category of remaking operating processes. So what we're doing under what I know you guys have read about 1GS 3.0. So if we redo our client onboarding, anti-money laundering, KYC processes, and we had a process that had 3,800 people touching it, not 100% of their job touching it. And now we'll have a process with a few hundred people touching it. That's a productivity. When it's done, I can measure the productivity gain of that process. Okay. That's different than saying to me, okay, you've got a bunch of people who work in investment banking. Okay. How much more productive are they? Now, one measure is I can look at the revenues we have in investment banking and the earnings in investment banking and the number of investment bankers we have and the amount of capital we allocate to that business over the last 10 years. And you know what? Our people are more productive because we're making on a per person basis, more revenue, more profit on a per person basis. But we definitely are allocating more capital to the business and the capital is helping and creating leverage. But there's no question. [30:00] and they have better tools. They can do more. We've found ways for them to do more. We also have collaborative power going on inside the firm in the context of 1GS, 1.0 and 2.0 that are adding to revenue growth. It's very hard to pull it apart. But I think every five years we can directionally say, is it more productive? Yes, it is. Because you go back when the firm went public, the firm had $6 billion of capital, I believe, and we had 16,000 people, something like that, order of magnitude. Today, we have 45,000 people and $110 billion of capital. The firm is much more productive against producing reasonable returns on that capital than it was 27 years ago. I mean, that's clear, but that's a 27-year look, not a year-to-year-to-year look. Here's a question. AI works when there's a lot of data, pooling together a lot of unstructured data. When I think of a firm like a Goldman Sachs or maybe a law firm, et cetera, There's always a little bit of an, I think, I suspect, an alignment problem where maybe the big producer doesn't want to share all their data with the rest of the firm, doesn't want everything in their head just to be in some big knowledge pool that everyone has access to You might want to be able to take your book of business to a rival shop every once in a while That probably always been the case But I think it probably a bigger deal with AI because AI could in theory really leverage all of that latent knowledge that exists in a lot of people heads How do you think about getting a firm wide understanding of what's truly known and understood within many people's heads so that you can really maximize this technology at a time when, again, it might make sense for at least some relationship players to keep a little bit of knowledge inside their own head? Well, it's an insightful question. And I start with the fact that one of the things I'm really proud of about Goldman Sachs is Goldman Sachs is less what you described. Yeah, let's say the other one. I think that we really have, for a long time, it's been part of the culture, but we really have a culture of sharing and collaboration. It is not a star system. People share, share, share. It's just such an ethos at Goldman Sachs that I think we have less of this problem than other organizations have. But what's interesting to me about your question, and it makes me really think about it, is it's something that I deeply believe that you started with, and I'd frame it in three buckets. this technology is incredibly powerful when you have a clean data set put these models against a clean data set it is extraordinary like it is mind to me what you can do the speed with which you can do it the information that you can accumulate and how powerful that is when you have a clean data set. It wouldn't surprise you. Goldman Sachs has an extraordinarily amazing data set that we're working very hard on to make as clean and possible on everything we've ever done from a trading and investment banking perspective over the last 40 plus years. We happened to build a system 40 years ago called SecDB that was a trading system. So we have data from all our trading that goes back long before most other firms kept data on this stuff. And so we're working very hard on that. But these tools, extraordinary when you have a clean data set. When you don't have a clean data set and you send them out into the world, you get really, really cockamamie answers. Let me give you a simple example. just tell a story from a few weeks ago that was another one of these data points for me and really understanding how this all works. It's pretty simple. I was down at the Masters and Rory McIlroy was up by six strokes after two days. And I was like, holy cow, he's probably going to win twice in a row. How many people have won the Masters twice in a row? I know Tiger Woods has. Okay. I think Jack Nicklaus has but I don know if there I can remember if there anybody else So I asked one of the models hey how many people have won the Masters twice in a row And the answer comes back Jack Nicklaus and Nick Faldo So I was like, oh yeah, Nick Faldo did win. But it didn't say Tiger Woods. Oh, interesting. So I said to the model, I said to the model, hey, I don't think you're right. Didn't Tiger Woods win the Masters two years in a row? Model says, wait one second, comes back. Yes, you're right. Tiger Woods did win the Masters twice in a row. Why did you get that wrong? I said to the model. Model comes back, well, I use this source. I did this. I did this source. The bottom line is not a clean data set. Yeah. Not a clean data set. Garbage in, garbage out. Now, human judgment had to look at that and say, no, no, no, I don't think that's right. Are we going to get to a place where the data sets get better and better and better? Yes. But as long as one of the functions of the model is they go out into the internet and they go out into social media and they go out into the media and they go out into the world there's going to be a lot of garbage in garbage out the models can get better and better and better but that's going to be it's going to be really interesting to see over time how the models get better and better at distinguishing and one of the things that's interesting is i then asked the model hey okay will you search this differently the next time because wherever you went got it wrong and the model said no no that's not the way it works now one day i think it will work [35:00] way and it will search it differently. But that's just interesting to me about what we are. Here's the third thing. Okay. You were talking about what's inside people's heads. Yeah. I personally believe the answer to this is never. And this is one of the magic things about humanity. This is one of the magic things about creative things. And by the way, part of investment banking and client relationship is creative. It's EQ. It's emotional intelligence around spending time with people, how to connect with people, how to build trust with people. Okay. Machines can be really, really good. Tools can be really, really good. But unless you're in the camp that humanity is not going to be about trust and relationships, I think this is like a never thing. And so the bridge is how to use the tools. Okay. And with the tools, leverage more of your humanity and your connectivity in a special way to be more impactful with people. Now, here's an example that I think is an easier example to kind of look at. I have a daughter who's a screenwriter. I've talked to her extensively about this, because people talk when somebody writes about their voice. Okay? And one of the reasons why, if you really understand the process of writing a movie, okay sure can AI write a script Absolutely But the first thing is that script if my daughter the screenwriter it is not in her voice unless that model has access to every single thing she ever written, every single experience she's had in her life, every single emotions he's ever felt. Never going to have that. And so voice matters when you're writing, you're doing something creative. Now, here's the thing. She puts her voice on whatever she writes. Now she goes to the studio the producer says you know i don't like this i don't like that that's a personal feeling he's not asking the model to say whether i like it or i don't like it he's reacting to it emotionally because it's creative it's like relationships i don't like this by the way somebody came in and pitched an investment banking deal and the cf said you know there's just something i just didn't like the way that guy talked to me okay that's it that's a human emotion yeah that matters that's not going to change and so this concept of getting things out of people's head what you're really talking about is human intelligence, humanity, EQ, and I think we're going to figure out how to use these models more powerfully. The data sets are going to get cleaner and cleaner. Yes, maybe we'll expand better into the second and clean up when you go out in the world, but when it comes to the way human beings interact I actually think this is a superpower that going to get more valuable and more valuable and more valuable And it something we want I think people should think about People should try to learn I think public speaking still matters I think writing is still going to matter. I think how you communicate and talk to people is going to matter, how you spend time with people, how you believe. All this stuff's not going away. Thank you. 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 times, 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 4imprint, 4certain. Support for the 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. Brokerage services by Open to the Public Investing, Inc., Member FINRA, and SIPC. Advisory services by Public Advisors, LLC, SEC Registered Advisor. Complete disclosures available. [40:00] at public.com slash disclosures. Let's talk about healthcare for a second. It doesn't always work the way people expect it to. If you've ever waited on a prescription refill or had a hard time getting the care you needed, you know the feeling. The system should just work better for everyone. That's exactly what the people at Optum are trying to do every day. They're a healthcare company linking patient care and pharmacy services and using data and technology to drive the whole system so care is connected, not complicated for patients and providers. Things like making it easier to get care that looks at the whole person, from primary care doctors to mental health support and even in-home care, and then using technology to make sure they all work together. Technology designed to help doctors spend less time on busy work and more time with their patients. And those prescriptions? Optum is working to bring costs down, save patients money, and make it easier to get refills. Little by little, Optum is helping make healthcare work as one for everyone. Head to business.optum.com to see how. and Sachs have extraordinary interpersonal skills. I believe it. We interviewed Lloyd Blankfein a couple months ago, and Lloyd, like, he's a funny guy. He's super funny. Yeah, he's really funny. Super funny. You're a fun guy, too. Yes, you too. I'm not as funny. I'm really not as funny. I mean, he's much funnier than I am. Well, okay, you mentioned looking up share prices earlier on the Quotron machines, and speaking of share price, your stock recently crossed $1,000, Goldman stock. A all-time high. An all-time high. So congratulations. Thank you very much on that. My question is, how often do you look at the share price of Goldman Sachs? Is that like the first thing you look at when you wake up in the morning? Well, I don't. It's not the first thing I look at when I wake up in the morning because I looked at it when I closed the day before. And it doesn't it generally doesn't change overnight. Look, I during the week, I would say I look at the stock price most days during the week, but it depends on what I'm doing. If I'm in the office, if it happens to be a day when I'm in the office going from meeting to meeting and in between meetings, you know, I jump on my desktop and I'm responding to emails or making telephone calls, the screen's up and I'm looking at our stock price, but I'm looking at a bunch. I'm trying to look at what's going on in the market, have an absorption as to what's going on in the market that day. If it's a day where I'm in a bunch of meetings and I'm on the road, I might not look at the stock price at all but probably at the end of the day I look and say hey what happened today And by the way it not just our stock price It actually what happened in the market today For me to be effective I have to be super in touch with what happening in the market every day And I spend time on the weekend also catching up on it. You know, from my perspective about our stock price, of course, it's human nature. I'm going to look and see. But what I really care about is kind of year to year to year. And what I really believe deeply, and it's been one of the things our leadership team has really been focused on for the last eight years is if you grow the earnings of the firm, if you invest in growth in the firm and you grow the earnings of the firm, the stock will do just fine. The multiple will move around. But if you grow the earnings of the firm, the stock will do just fine. And so we've been focused on investing in the business and growing the earnings of the firm. And we've grown the earnings of the firm material. We had our first investor day in January 2020, and we put out a plan and we've been executing against that plan. And since then, we've grown the revenues of the firm 65-ish percent. We've grown the earnings of the firm about 140 to 145%. And what do you know? The stocks reacted really well. And we've gotten a little bit of a better multiple because it's a much bigger, more diverse, durable business. The durable revenue nature of the business is better. So we're getting a better multiple on what we're doing. But I'm super excited because I now look ahead for the next five to 10 years. And I look at this technology and our ability to remake operating processes and take the most productive people and have them spend more time with clients and broaden our client footprint in areas like wealth management etc I'm like, wow, we can really grow the firm. We can really grow the firm and grow the earnings further because we're going to have more capacity to invest in growth. And so that's super exciting. And so it's been a journey. I mean, when I started, the stock was around 200. And now the stock's over 1,000. By the way, it could go back, but it's not going back to 200. But the stock can back off, the multiple come down, our earnings could slow if we went into a tough economic environment. But fundamentally, the firm is bigger, broader, more diverse, more durable. And I think on a super course for more earnings growth and more performance for our shareholders over the next five years. So the performance is objectively strong, but no offense, you can see a lot of charts right now in the market that are going straight up into the right. Absolutely. The important thing, and I know you know this, Joe, the important thing is to look at the relative performance over the time we're talking about is top of the heap. Totally. And we're going to keep driving to try to deliver that. But I got a sort of a capital markets question for you that sort of maybe as a banker, you might have some thoughts on. So connecting it to the AI conversation, what was your thought when you saw, or maybe you had something involved in it, Google making an equity raise to continue... [45:00] That was interesting because there's been a lot of debt financing from the big tech companies for the AI buildout, which we could have done a whole episode just on that. What do you make of a company that big doing an equity raise at these levels and even doing a private placement? When I asked you at GPT, and it could be wrong, as your example with Tiger Woods noted, when I asked about comparables, like what other big companies did these private raises? They're almost all crisis-era Berkshire investments, including one in Goldman. But what did you make of that, Google doing an equity raise to finance rather than a debt raise? Well, there are a couple of things to think about here. But first of all, I know a little bit about this. There are some things I can talk about and some things I can't. But I think you know that the firm ran this whole process. The firm was the only bank that was involved in this for the last five months up until a couple of days ago when other banks were brought in to participate. The firm ran this over the last five months. This was Goldman Sachs' deal on the team led by Yasmin Kupal and Kim Posnett and David Ludwig, you know, as our equity capital markets business did an extraordinary job. This is one of the largest, if not the largest, secondary follow-on equity offering ever. Yeah. And we ultimately if the green shoes exercise we going to raise 85 and 90 It was up this just morning from 80 to 84 Yeah we upsized it and the green shoe my guess is given the way the stock trading will probably be exercised in the coming days And so it raised $90 billion. And so it's really extraordinary. And by the way, I think one of the interesting things to think about, this is the first concrete tangible data point on investor demand at this scale. And we have some other things coming up at this kind of scale. And so this is a very important tangible investor data point. I think one of the things that's very helpful to the firm is sitting in the middle of this. We got a lot of information on how investors are thinking about all this and the others that I think is super, super interesting for us. But it's unprecedented in its scale. And it's reflective of the fact that we now have some of the largest companies in the world that have capital plans over the next five years that are very, very significant. And they can fund a lot of it with debt. but I think you're going to see and Alphabet is the first one to make this decision. I think you're going to see a bunch of them fund a bunch of it with debt but also raise more equity because they're starting to think about, okay, what does this look like? It's one thing, what does it look like this year but what does it look like over the next five years? How much capital do we need How much debt Equity markets are good Our multiples are pretty high given the way the market responding at the moment We got to think about our leverage going forward the capital available And so the analysis is and this is not all public but in specifics but I'll talk generally, a company has to look and say, not just what's our one, two, or three year capital plan, what do we think our capital plan is over the next five to 10 years? How's that going to change our leverage? What if there was a change in the market multiple, and our leverage is up? How are we going to feel about that? And it requires kind of a long-term conviction on what you're doing from a capital deployment standpoint and making the decision as to what makes you more comfortable in kind of the leverage and balance between debt and equity. I think you're going to see more companies issue equity because capital is available and you want to be cautious about this. If you get it wrong, if you totally rely on debt and you get it wrong, you will really regret if you wind up having a downside scenario that's tougher than you expect. So I think you can see more companies think that through. And Alphabet's the first company to step forward. Our team spent a lot of time over the last five months really helping them think through these kinds of issues. I think they were very smart to really think it all through. Well, speaking of equity raisings, you got the lead on the SpaceX IPO. So again congrats on that Again our team really the team on that really really extraordinary Dan Dees Kim Posnett I mean just really Susie Schur really really extraordinary partners of the firm who have just done an extraordinary job over years. I mean, over years. I imagine this was a very competitive process, right? And you were talking a lot about the importance of relationships earlier. Using SpaceX as a concrete example, like what does that relationship actually look like? Is it true that you are sliding into Elon Musk's DMs to pitch yourself? What is it that you offer to a potential client that other banks can't offer? Well, I mean, there are things that I'm willing to say and try to give you some perspective on it. But the most important point that I could amplify about this is we didn't win this in the last six months. We won this over the last 20 years. Wow. The decision wasn't made. The decision wasn't made until the last six months. And there's a lot of things that the team did over the last six months that contributed to the ultimate result, which, by the way, now we have to execute. But this is a cumulative effort of lots and lots of people over 20 years. And we first as a firm, we first met Elon Musk and the person that first was responsible for covering him was going to be a big deal. [50:00] Stuart Bernstein was doing a bunch of stuff around kind of green energy. And he and his brother had SolarCity. And that's where we first connected with Elon. And then we took Tesla public when Tesla went public. Sometime around SolarCity and Tesla, I first met Elon. And so I've known Elon for more than 15 years. Dan Dees, Stuart Bernstein kind of semi-retired from the firm in the early part of the last decade. So let's say 2012, 2013. and Dan Dees, who now runs our global banking and markets business, was an investment banker who was just coming back to the U.S. after having been in Asia for a long time and he was basing out of the West Coast and running our TMT banking business and he started building a personal relationship with Elon and Dan has built a really wonderful dialogue with Elon over a long time. But it's not a straight line and there are times where we're doing things and we're not completely in sync. But at the end of the day, as a firm, We've been long-term committed to Elon, his companies, his teams. By the way, in this situation, Brett Johnson, the CEO, is a hugely, hugely important guy. And we had people spending time with him for years and years and years And that what our business is Our business is building trusting relationships over time with people giving the right advice taking a long view not being transactional, and trying to earn trust. So when the biggest, most important things happen in the world, we have a better chance of being asked to provide services. And that's kind of what we do did you send him a dm so with respect to the dm it is unfortunately it's untrue that i pitched elon via dm about the deal okay what is true and this is the way facts get confused elon some point in the last year stopped taking texts on his phone or emails and started telling people that he knew that if you wanted to get in touch with him direct message him on x and so because I have a relationship where I reach out to Elon and we exchange messages about a whole variety of different things. I stopped sending him texts on his phone and started direct messaging him when I wanted to. This is how Joe communicates. This is entirely when Tracy is a warrior. When I want to communicate, I started communicating. So I actually, I went back actually and looked at the chain you know in my ex count over the last six months just to make sure I wasn wrong about this when I saw it in the press and I said to Tony I don think this is true And I never I never I never asked Elon in a text message about about the IPO We focused on his team. We focused on the work that we had to do. And we're really proud and pleased to be leading this. And we're going to do everything we can working with the other banks to get the best result that we can for SpaceX. So it looks like this year is going to be a year of mega IPOs. We had the private confidential filing from Anthropic. We'll probably get a couple others, et cetera. However, the much bigger story is still generally private capital markets have gotten so liquid that some of these very successful companies, they could, in theory, maybe never go public if they didn't want to. Some people think this is bad for capitalism, that a bunch of big companies are going to come in and that unlike the Microsofts and the Apples and so forth, where the broad public really got to participate, even if you were just an index fund investor. And a lot of the upside, these companies are going to be coming at some of the biggest market cap companies in the index. And the public investor will not have caught any of the ride from the 500th biggest in the S&P 500 to the 20th, et cetera. Does this break capitalism? Is this bad for the US or is this fine Well I going to give you a conclusion first Okay We can try to dig in a little bit to the details which I know you guys love to do But I just start I don think this is going to break capitalism OK All right. And I think it's and I think it's good for the red headline. There goes your headline. I think it's good. I think it's good for the US that we have the biggest. Oh, by the way, it's important. Breaking hot, hot headline. Just moved on the terminal. Anthropics had to tap Morgan Stanley, Goldman Sachs to lead IPO. This just came out. OK. So there you go. OK. Anyway, keep going. So I think this is good for the United States to have the biggest, most important companies in the world. When you talk about what's happening, there are a lot of complex things happening about why we have fewer public companies today than we had 25 years ago. Yeah. And this is a long term journey around policy decisions and market structure over a very long period of time. the reason that companies like microsoft went public when they did it the size that they did was there was no capital available to them other than going into the public market right right but it's not fair to say that investors got to participate you know in everything because it's not like there's more access well no one talks about the 900 companies that went to zero that went public in the same year as microsoft that's i mean that's so and by the way that's another thing most companies don't survive so it's it's just not that simple i do think we've created a regulatory structure and a market structure. [55:00] that really makes it unattractive to go public until you have to. The reason these companies are going public now is because they have to. They have capital needs that are so voracious that it is not prudent for them to try to do 100% of it in the private market and not have a public currency. It's just not prudent. And that's why I think you're going to see a bunch of these companies go public because they actually need the capital. It's interesting. I've advised companies for years. And at 1.25 years ago, I actually ran the equity capital markets business at Goldman Sachs. That was a job I had earlier in my career in Goldman Sachs. And I used to say to companies, there are a few reasons to go public. You need the capital. You need the currency. You have to create liquidity for early investors. But all the other stuff that people talk about is really not critical. And you have to recognize that if you're running a company successfully and you don't need those things, when you take it public, you will run it differently. You will run it differently. It is different running a public company than a private company. There are different pressures. There's a different life cycle of how you have to respond. You will run it differently. And so my advice has always been wait as long as you possibly can to go public okay if you giving advice to companies and I think companies have recognized the capital has been available privately at most scale the liquidity has been available in a variety of different ways at most scale. The currency thing is tricky. But generally speaking, with these growth companies, they don't need the currency because they're growing, they're not doing M&A. And so I think you're going to see a bunch of these companies go public because they need the capital because they've reached the scale and they have a demand for capital where they need the capital. I think the current SEC chairman is working on a variety of things that hopefully will take some of the friction out of going public, which will help a little bit. But at the end of the day, given the way markets work and indexation and passive funds and ETFs, a $5 billion market cap company with an $800 million IPO, the discount associated with getting that going has been less attractive to investors. And part of the reason is public capital has been available without a discount. But that, by the way, that could be a cyclical thing. In other words, we could see periods of time where private capital will have a bigger discount and the public market will be more efficient And then you see more of those companies come to the public market But I think fundamentally we going to operate with fewer public companies and the companies are going to be bigger I mean that just that the way the market structures evolve Thank you. Thank you Or imprint for certain Support for the show comes from public If you 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 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. Let's talk about healthcare for a second. It doesn't always work the way people expect it to. If you've ever waited on a prescription refill or had a hard time getting the care you need, [60:00] You know the feeling. The system should just work better for everyone. That's exactly what the people at Optum are trying to do every day. They're a health care company linking patient care and pharmacy services and using data and technology to drive the whole system so care is connected, not complicated, for patients and providers. Things like making it easier to get care that looks at the whole person, from primary care doctors to mental health support and even in-home care, and then using technology to make sure they all work together. Technology designed to help doctors spend less time on busy work and more time with their patients. In those prescriptions, Optum is working to bring costs down, save patients money, and make it easier to get refills. Little by little, Optum is helping make healthcare work as one for everyone. Head to business.optum.com to see how. Okay, so speaking of capital being available, which is kind of putting it mildly, the markets right now seem kind of crazy to me. And you were at the Economic Club of New York yesterday talking about markets being in more of a greed stage than a fear stage. Based on your Wall Street experience and again you mentioned you started in the early 1980s have you ever seen anything like this Does the current time period I don know have a historical analogy for you Sure There are there In fact somebody said to me recently this is unprecedented They were talking about 10 companies having 30 some percent of 30 some percent of the S&P waiting on a market cap base. And I was like, no, it's not unprecedented. How about the 1920s? How about the 1960s? How about the late 1990s to 2000? I made the comment yesterday at the Economic Club because I was asked about the way the the stock market's running. And if you actually look at the whole segment, I said a lot more than that. And I actually said before I said it, I said, look, I'm going to say something. I know it's going to create a lot of headlines. But I do think it should be said because I do think I'm seeing some of that behavior that I've seen before. And people are worried about missing out on this technology boom, whether it's compute, whether it's storage, whether it's chips, etc. And so people are crowding in. And that's a greed thing because they want to participate in their fear of missing out. What I should have said, okay, at that moment was all of that. But then I should have said, but, you know, it's kind of interesting because if you look at the S 500 the top 10 companies which by the way are mid to high 30s market cap and those other three periods the top 10 companies were also in the 30s And one of the things I have pointed out to people is these top 10 companies actually have more earnings than the companies did. They generate a lot of cash. They generate a lot of cash. So that's something to observe. But what I should have said is the market multiple, the earnings multiple on the top 10 companies in the S&P, these are approximate numbers. So don't hold them to me directly. I actually looked at this this morning because I thought this would come up. And so hold it as approximate. The top 10 companies are trading on forward earnings in the low 30s. In the late 90s, 2000 internet boom, those companies were trading at more like high 40s to 50. You think about Cisco when its market cap was over 600 billion. So these multiples, yeah, are high, but the growth's pretty good. The earnings are real. Those numbers are on the high end of the distribution for sure, but not crazy. The other 490, okay, are trading between kind of 17 to 20, which by the way, 20 high side, when you add it all together, it's kind of 22 times forward. A lot of it's at 17. I actually think the 17 sounds pretty attractive. If you think that all these companies are going to be able to use technology not all of them will execute successfully but be able to use technology to improve operating processes and create more efficiency and therefore faster earnings growth and more investment in their business You should see over the next five years improvement in the earnings growth of that other 490 I'm talking on average. So is the market running? Yes. Are people crowding into a handful of stocks because they're fearing of missing out? Yes. That's what I meant about the greed versus fear. but this could go for quite some time because it's narrow and on a historical context it doesn't look like polaroid at 80 times earnings in the 1960s it doesn't look like rca at an infinite amount of time times earnings in the 1920s so it's i'm not sure is the bottom line i'm not sure but there's no question for a narrow group of stocks the market's really running and we've got a lot of what's interesting to me is we have a lot of things going on in the macro with the war with the oil shock, supply chains, inflation getting stickier that are going to have an effect. And, you know, the market may absorb all that differently than it's absorbing it at the moment. At the moment, it's kind of brushing it aside. Things can always get crazier, as Joe likes to say. That's right. All right. You know, like I said, you're going to be involved in the Anthropic IPO. Congrats on that. Just between the three of us here. Just between the three of us here. I know you're part of Mythos and [65:00] Project Glasswing. And there's this cynical take. It's like, oh, they just released it to a small group because they actually didn't have enough compute, whatever. What have you seen? And is it really like, is it keeping your cybersecurity people up at night? Have you found a bunch of bugs that you didn't know about? Tell us. What can you tell us about that? Here's what I can say publicly. I think it's very important. These models are incredibly powerful. And I think it's very, very important that these companies that are producing these incredibly powerful models, work collaboratively with the government and with the private sector to give people opportunities to understand what these models can do and to try to continue to secure our critical infrastructure as effectively as we can. Do I think that the process around this was perfect? No, but we're learning as we go. And I think there are a lot of things in the process that have been good and have been helpful. It wouldn't surprise you that just using financial services, for example, is the eight largest banks are significantly ahead on cyber investment and cyber protection over the last decade than bank number 2000 that sits in the middle of the country in Arkansas Okay And the risk in financial services it doesn mean that there can be issues for the biggest banks and the biggest banks need to constantly constantly constantly constantly think about this We invest a lot in this and we all over this That doesn't mean we can't have a problem. We could. But the risk is that a medium sized bank has a problem and it creates fear in a broader group of medium sized banks. And you don't have to look that far. If you look at SVB and how SVB had a problem and it created fear in a variety of banks, you could see a cyber attack at a midsize bank creating ripples through the banking system, which will be dislocating. And so I think it's really important that we figure out how to help the broader business community get more invested. And by the way, that's something that's got to happen both by the private sector investing more, but also the government finding ways to help. So, you know, I do think that the administration's executive order to get these models is important. There's got to be coordination between the government and business on this. And I think the process will continue to get better. There are going to be more models. There's going to be more process. And, you know, we've obviously had a team that's learning, looking. You know, we learn things, making adjustments, and we're going to continue to do that. But that has been a part of what we do This might be accelerating a little bit but it not that different from what we been doing for a long time Just maybe the pace of some of the change on what these models can do is going a little bit faster And so we got to be more nimble and we have the resources to do that. What I worry about is all the companies that don't have the resources to do it. And it's, by the way, it's not just financial infrastructure. Think about energy infrastructure. The biggest utility company is going to be way ahead of, you know, the local water company in Akron, Ohio, right? And so everybody's vulnerable. And so we've got to figure out how to help people and bring people along. I want to pivot. I guess we're going to run out of time soon. So I want to pivot slightly to a very important question, which is your own use of AI. Joe and I know that you're into music. You like to DJ famously. Have you been experimenting with AI generated music? Yeah, I mean, I've been I've been playing with it. And what I'd say is what I think is cool about it. As someone who's produced a bunch of music, and I want to come back to that because I think there's an important kind of lesson to learn about history in terms of music production to where it is. Good. Let's go. I think it's super cool, but what it does is it opens up the creative channel to people that would never have an opportunity to produce music. Okay. Well, it's just a little bit of history about music production because by the way, 30 or 40 years ago, I could never have produced music. Now it so happens I played an instrument when I was in elementary school and high school I played the saxophone I can read music I understand music theory because I played an instrument for 12 years When I was out of college I would call myself a recreational drummer I mean, I could sit on a drum set and carry a 2-4 beat. So I understand bass music theory. But I've been a rock and roll guy forever. I love Bruce Springsteen. I saw my first Bruce Springsteen show in 1978, the Darkness on the Edge of Town tour. When he wrote that album, and by the way, the albums that came before had a bunch of them that came after. He sat in a studio or he sat at a piano or he sat with a guitar. He wrote notes down on paper. He wrote lyrics down on paper. He played around with it. Then he invited human beings into the studio with him that played piano, organ, drums, guitar, horns, saxophone, all different things. And they sat there and they made the music. and somebody recorded it and it got pressed onto a vinyl album. Okay. A lot of the music that we listen to today has been made on a computer on platforms called Logic Pro or Ableton, which is actually how I work to produce the music that I produced. And so if you want horns, you can choose from 150 different horns. Okay. You don't need... [70:00] 15 different musicians. Okay. Nobody's talked about the difference, how technology evolved. And by the way, what it did is because Logic Pro and Ableton exist, and I'm relatively dysfunctional on the platforms, you know, I can muddle my way around, but I need an expert to help me because they can do in three seconds what takes me 10 minutes. But what it allowed is for somebody to say to me, hey, I want this kind of a sound. Okay. I didn't now have to write the musical notes down. okay you could basically basically take out a piano and say this chord sound this chord progression try it on the computer you could say i want this to be done in horns instead of in a piano click it's now horns and you can basically put a top line together that's very interesting with a computer a lot of the music that's produced today yeah that we listen to is produced that way okay Now what you've got is you've got an application that's basically doing that. Okay. And I think there are a couple of big issues that need to be wrestled through with this. The first is these AI apps are using other people's content in some way. Right. And the artists aren getting in any way compensated for that And I think you got to think about how you going to bring the artist community and their ip into the discussion of this over time and you know it not gotten a lot of attention but i think it a very very important a very very important thing that has to be dealt with and secondly this goes back to the whole discussion i had about voice when i was talking about my daughter and screenwriting okay the stuff i produced okay it's my voice i said i like i like the way this sounds to me so do it this way i like the way this feels so let's try it this way okay and then you put it out there and you see whether or not anybody likes it obviously the technology allows you to do it faster but without a human being really then massaging it and tempering it okay i think you get you're going to get at the moment certainly mediocrity and as opposed to really, really interesting, resonant stuff. All right. David Solomon, clearly you've given the music industry quite a lot of thought. Thank you so much for coming on All Thoughts. Really appreciate it. Well, thank you for having me, guys. And I appreciate the conversation. Yeah, it was fantastic. I appreciate you so much. Appreciate it. Thanks. Absolutely. We'll do it again. We will. Joe that was a lot of fun That was a lot of fun A lot of fun I kind of surprised how much thought David puts into the music industry to be honest Yeah, I loved that you asked that last question, and I liked how thoroughly he answered it. I wasn't sure if that would be the type of thing where he was like, well, I'm going to focus on the banking. You know the thing that really shows, and there's a number of themes within the conversation that are very important to think about. but I mean, he's really good with dates and names. And as someone who is not good with names specifically, and I'm okay with dates, I asked him that question. I was like, well, what do you have that like, okay, there are other bankers and you're like a producer. You get some of those, you get that sense, right? Someone who's like that good at like crediting the right people, knows the name, knows the days, knows what they contributed. I kind of wish I had that skill a bit more. Yeah, you can always tell a good all thoughts guest by like how many specific dates and data points. Absolutely. I know a good data point, by the way. Just the one. There's one data point. So in the music question, and he mentioned having gone to see Bruce Springsteen's Darkness on the Edge of Town tour that album came out really close to the bottom in the 1970s But this is why it interesting because plenty of albums came out Bruce bruce springsteen re darkness on the edge of town in 2010 so twice that album has been associated with sort of like generational historic lows oh that's funny okay so okay we'll know the crash is coming when bruce springsteen re-releases this for the third time the next time there's going to be at some point there'll be another crash when he re-releases it that's the bottom yeah the buy signal all right all right uh important data point here but no thinking about the ai job situation this is a theme that keeps coming up which is if knowledge is commodified if knowledge and experience gets commodified on a single platform then the human interaction social maxing looks maxing whatever you want to call it seems to be getting more important i would actually even say there's another thing it's absolutely true and i would say there's probably something else which that, okay, everyone gets this sort of commodified AI answer. So, okay, then you can infer that the human touch will be more important. But then you add in the fact that the last 15 years of mobile phones has destroyed so many people's ability to make eye contact, that it probably be even more in a premium, even setting aside AI, because the skills are probably just going to get rarer as so many people lose their minds to the phone and the internet. All right. So everyone, remember to look into each... [75:00] when you're speaking with someone and I guess pick up the phone for David's advice. I also just thought it was really interesting that he thinks we're going to see more public companies doing private equity offerings. That alphabet headline, the Google headline, that was a big deal and I think caught a lot of people by surprise. So the fact that it's like, this is the first of multiple he expects. It's a very interesting trend to watch. For sure. All right. Shall we leave it there? Yeah, let's leave it there. This has been another episode of the All Thoughts podcast. I'm Tracy Allaway. You can follow me at Tracy Allaway. And I'm Joe Weisenthal. You can follow me at The Stalwart. Follow our producers, Carmen Rodriguez at Carmen Armand, Dashiell Bennett at Dashbot, Kale Brooks at Kale Brooks, and Kevin Lozano at Kevin Lloyd Lozano. And for more All Thoughts content, go to Bloomberg.com slash All Thoughts. 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 slash All Thoughts. And if you enjoy Odd Lots, if you like it when we talk to the CEOs of big banks, 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 free. All you need to do is find the Bloomberg channel on Apple Podcasts and follow the instructions there. Thanks for listening. Thank you 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. Healthcare doesn't always work great. If you've ever waited on a refill or couldn't schedule an appointment, you get it. That's the kind of stuff Optum is changing. They're using data and technology to integrate patient care, pharmacy, and everything else. So healthcare is connected, not complicated. What's that look like? Cheaper prescriptions that are easier to get and care that looks at the whole person how you need it. Optum is helping make health care work as one for everyone. Learn more at business.optum.com. As industries evolve faster than ever, companies need an environment that accelerates strategic growth, and Michigan delivers on that promise. From emerging startups to global enterprises, Michigan offers what executives value most. A resilient, innovative ecosystem, diverse communities that attract top talent, and a quality of life that supports work-life balance. With our unified Team Michigan approach, businesses scale faster and compete at the highest level. Michigan, pure opportunity. Seize your opportunity at michiganbusiness.org.