Citadel Acquires Aschenbrenner AI Fund Assets After Leveraged Blowup

杠杆爆仓后，Citadel收购Aschenbrenner旗下AI基金资产 尽管对AI基础设施需求的长期宏观逻辑成立，但极度高企的杠杆与方差拖累仍迫使Situational Awareness立即清盘。 本期节目深入探讨了Leopold Aschenbrenner旗下专注AI的对冲基金Situational Awareness的戏剧性崩盘。该基金在对AI基础设施建立高杠杆头寸后迅速瓦解，随后被Citadel收购。主持人分析了该基金在高波动下走向覆灭的数学必然性，并将其与传统风险投资的风险特征进行了对比。随后话题转向法拉利纯电动车Luce提前完成2026年销售目标，剖析了收藏家心理以及法拉利相较于兰博基尼等竞争对手的电动化战略。最后，他们以一段轻松的讨论收尾，辟谣了关于其演播室标志性绿色美学风格的粉丝理论。 前 OpenAI 研究员 Leopold Aschenbrenner 创立的 AI 对冲基金 Situational Awareness 在重仓 AI 基础设施并做空软件股遭受巨额亏损后，被迫清盘其公开股票资产；该基金峰值资产管理规模曾达近 450 亿美元。 (0:01) Ken Griffin 旗下的 Citadel 已介入接盘 Situational Awareness 的公开市场资产，预计投资组合的实际回收价值约为账面价值的 50% 至 80%。 (0:01) Numerai 创始人 Richard Craib 分析指出，该基金 150% 的极高波动率使其归零破产的风险堪比抛硬币；与传统风投相比，公开市场杠杆基金在保证金催缴下仅需数天便会彻底溃败。 (4:40) 法拉利首款纯电动跑车 Luce（售价约 65 万美元）在发布两个月后已提前完成 2026 年约 500 辆的销售目标，其中中国市场的买家需求尤为强劲。 (10:24)

TBPN reports that Leopold Aschenbrenner's AI-focused hedge fund, Situational Awareness, unraveled under severe leverage and macro volatility, while Ferrari achieved its 2026 sales target ahead of schedule for its debut electric vehicle, the Luce. Situational Awareness Unwinds Concentrated AI Bets (0:01) Leopold Aschenbrenner’s hedge fund, Situational Awareness, was forced to liquidate its public stock portfolio following heavy losses on leveraged, concentrated bets. Based on Aschenbrenner's thesis that artificial intelligence requires an unprecedented expansion of compute, memory, and power, the fund took large long positions in infrastructure names such as Nebius, Sandisk, Micron, and CoreWeave, paired with a short position in Adobe. After reaching a peak reported assets under management near $45 billion around late June or early July, the trades collapsed amid recent declines in AI infrastructure, energy inflation, interest rate uncertainty, and shifting Federal Reserve expectations. Leverage, Margin Calls, and the Citadel Buyout (4:40) The fund's unwinding accelerated rapidly overnight. Bloomberg and the Financial Times reported at 9:25 PM that the fund was seeking emergency capital from limited partners, pitch decks describing the dip as a buying opportunity. By 6:05 AM, CNBC reported the public book was being unwound, and at 8:39 AM, The Wall Street Journal confirmed Ken Griffin’s Citadel had stepped in to acquire the portfolio. Unlike venture capital, where distressed firms slowly burn cash over years, leveraged public market vehicles face immediate margin liquidation. While social media speculation suggested a sale for pennies on the dollar, actual recovery was estimated at roughly 50% to 80% of book value. Mathematical Fragility and Risk of Ruin (4:40) Market observers emphasized that the fund's failure stemmed from extreme structural risk rather than an incorrect fundamental thesis on AI. Numerai founder Richard Craib highlighted that operating at 150% volatility incurs roughly 113% annually in variance drag alone, creating a coin-flip risk of total ruin over the fund's lifespan regardless of long-term AI progress. Co-hosts John Coogan and Jordi Hays noted that while Aschenbrenner previously served on the FTX Future Fund and resigned upon its 2022 collapse, public market leverage exposed his fund to rapid destruction once retail froth, interest rates, and macro headwinds converged. Ferrari Luce Hits 2026 Production Allocations Early (10:24) In automotive developments, Ferrari reached its 2026 sales volume target for the all-electric Luce within two months of its unveiling, driven largely by demand in China. Designed by former Apple design chief Jony Ive and priced at approximately 550,000 euros ($650,000 USD), the vehicle met its internal goal of roughly 500 units for 2026, advancing toward a cumulative target of 2,500 units by 2030. Ferrari directed dealerships not to condition high-demand supercar allocations, such as the SP3, SP4, or F80, on purchasing a Luce. Collector Strategy Versus Long-Term Brand Equity (15:35) The hosts debated the broader commercial rationale behind Ferrari's unconventional design choices for the Luce and models like the Purosangue. While one perspective holds that collectors purchase idiosyncratic or first-edition models simply to maintain standing on dealer lists for rare hypercars or to collect every variant the brand produces, another raises concern over brand equity. Unlike mass-market volume plays such as the Porsche Cayenne or Lamborghini Urus, eccentric, non-traditional Ferrari models run the risk of diluting brand prestige unless they achieve sufficient scale to cross-subsidize core performance programs. While Citadel's acquisition allowed markets to absorb Situational Awareness's liquidated positions without broader contagion, questions remain over whether ultra-luxury electric vehicles like the Luce will deliver lasting commercial returns or remain niche historical artifacts for high-end collectors.

本期节目聚焦由前 OpenAI 研究员 Leopold Aschenbrenner 创立的 AI 对冲基金 Situational Awareness 因高杠杆集中押注而遭遇爆仓清盘、最终被 Citadel 接管的始末，并探讨了法拉利首款由 Jony Ive 设计的纯电超跑 Luce 尽管饱受审美争议，却在中国市场推动下提前锁定 2026 年销售目标的商业逻辑与品牌争议。 Situational Awareness 巨额做多爆仓与 Citadel 接盘 (0:01) 由前 OpenAI 研究员 Leopold Aschenbrenner 掌舵的 AI 主题对冲基金 Situational Awareness 近日因杠杆头寸巨亏被迫清盘全部公开股票组合。该基金的资产管理规模在 6 月底至 7 月初曾达到约 450 亿美元的峰值，其策略高度集中于做多 Nebius、Sandisk、Micron 和 CoreWeave 等 AI 基础设施概念股，同时大举做空 Adobe 等传统软件龙头。随着过去数周 AI 基础设施板块大幅回调，叠加原油通胀、美联储降息预期落空以及美国房贷利率升至 6.66% 等宏观因素冲击，基金的杠杆多空交易全面崩溃。 事态发展极为迅速：前一日晚间 9 点 25 分，彭博社与《金融时报》披露该基金正向 LP 寻求紧急救助资金并宣称下跌是买入良机；次日清晨 6 点 05 分，CNBC 曝出基金被迫全额平仓；8 点 39 分，《华尔街日报》证实 Ken Griffin 旗下的 Citadel 已介入收购其全部公开上市资产。尽管市场传闻 Citadel 以一折以内的甩卖价抄底，但据估算实际资产回收率约为账面价值的 50% 至 80%。讽刺的是，在 Citadel 收购消息公布后，相关重仓股普遍录得两位数反弹，但近一个月内跌幅依然惨重。 波动率阻力与公开市场清算的毁灭数学 (4:40) 节目通过对比风险投资（VC）与公开市场对冲基金指出，杠杆是加速机构走向毁灭的核心杠杆。在传统 VC 模式中，失败企业如 Bird 可在两到三年内通过缩减开支与裁员维持运转，基金账面亏损亦可由 SpaceX 这类超级赢家随着时间抹平；然而公开市场的保证金追缴（Margin Call）容不得缓冲，数天之内就会触发流动性挤兑。曾任职于已倒闭的 FTX Future Fund 的 Aschenbrenner，虽然基于算力、内存与能源爆发性需求的宏观叙事极具前瞻性，却被市场短期波动彻底吞噬。 Numerai 创始人 Richard Craib 对该基金的风险架构提出了严厉批评。Craib 指出，在 150% 的年化波动率下，仅方差阻力（Variance Drag）每年就会损耗高达 113% 的收益，这导致基金在整个生命周期中的破产几率基本等同于抛硬币（约 50%）。甚至通过 AI 模型 Claude 计算后也显示，该策略走向覆灭不仅大概率发生，而且风险概率均等。这证明该基金的倒闭并非源于其 AI 长期预判的对错，而是被极端波动率、散户投机狂热与宏观利率收紧击垮了资金链。 法拉利纯电 Luce 提前完成 2026 年销量目标 (10:24) 在汽车产业方面，法拉利首款全电动超级跑车 Luce 在发布仅两个月后，便提前达成了 2026 年的全年销售配额。这款由前苹果首席设计官 Jony Ive 操刀设计的纯电车型售价高达 55 万欧元（约合 65 万美元），发布之初因极具突破性且非常规的造型在社交媒体上招致猛烈批评。然而据《金融时报》援引内部人士消息，得益于中国买家的强劲需求，法拉利设定的 2026 年近 500 台的交付配额已于 7 月初被抢购一空。 根据法拉利的长期规划，公司计划以每年约 500 台的节奏稳步推进，目标是到 2030 年累计生产 2500 台 Luce。为了安抚忠诚的传统内燃机收藏家社群，法拉利管理层明确向全球经销商下达指令，严禁将购买 Luce 作为进入 SP3、SP4 或 F80 等限量版顶级超跑候补名单的搭售前提，旨在避免伤害传统客户的核心忠诚度。 超级买家博弈心态与非传统车型的品牌分歧 (15:35) 节目深入剖析了 Luce 销售火爆背后的买家心理博弈。主持人指出，即便厂方明确禁止捆绑销售，深谙经销商潜规则的超级藏家们依然会主动买入冷门或边缘车型，以此提升自己在经销商客户档案中的积分排名，确保未来能优先锁定极为稀缺的限量版超级跑车配额。此外，对于追求集齐中置引擎、前置引擎、赛道版、F1 与电动版全谱系体验的骨灰级藏家而言，拥有法拉利历史上首款纯电车型具有无可替代的历史标本意义，甚至包含了押注该车若销量不佳后绝版停产的独特收藏价值。 在商业战略层面，主持团队就法拉利的非传统车型路线展开了辩论。兰博基尼 Urus 与保时捷 Cayenne 均是定位于大众日常驾驶的高产量车型，其核心目标是重塑销售规模曲线并创造充沛现金流，以此反哺顶级超跑的研发；相比之下，法拉利设计的 Luce 乃至 Purosangue 更偏向离经叛道与反传统的审美探索。有观点认为，如果这些非常规车型不能像 Cayenne 那样形成极高销量来反哺主干产品线，单纯推出日常化或怪异车型反而可能面临稀释法拉利顶级品牌资产的长期风险。 无论是激进押注 AI 算力革命的对冲基金因公开市场杠杆数学而遭清算，还是百年超跑品牌在争议声中依靠配额博弈提前完成电动化配额，均揭示了宏观愿景与资本底层运行规则之间的张力：前沿理念的落地不仅取决于方向的正确性，更受制于资产流动性承受力与买家群体的结构性博弈。

## Background 本期节目由科技商业播客 TBPN 于 2026 年 8 月 3 日发布在 YouTube 平台。节目由 John Coogan 与 Jordi Hays 共同主持，官方简介将其定位为浓缩每日直播精华的《Diet TBPN》，重点聚焦前 OpenAI 研究员利奥波德·阿申布伦纳旗下的 Situational Awareness 基金因杠杆爆仓被 Citadel 全面接盘的突发危机，以及前苹果设计师 Jony Ive 操刀的法拉利纯电超跑 Luce 提前售罄 2026 年配额的行业争议。 ## TL;DR 节目首先复盘了 AI 基础设施对冲基金 Situational Awareness 在遭受双向挤压后遭追缴保证金、最终公开资产被 Citadel 闪电接盘的全过程，深入剖析了二级市场高杠杆猝死与传统 VC 缓慢失血的残酷对比，并援引量化视角解释了极端波动率下的方差拖累陷阱。随后，话题转向汽车圈重磅动态，探讨了外观极具争议、单价高达 55 万欧元的法拉利纯电超跑 Luce 缘何能在两个月内提前锁定 2026 年配额，激辩这究竟是硬核藏家的孤品押注还是豪车经销商配额潜规则的延伸。最后，两位主持人轻松辟谣了演播室深绿视觉抄袭《蜘蛛侠》号角日报的网友调侃，分享了其源于 F1 赛车文化与复古美学的灵感故事。 ## Atmosphere & style 整体氛围兼具硅谷内行的敏锐风趣与华尔街交易台的冷峻犀利。主持人凭借在 Citadel 的过往渊源与对科技生态的熟稔，将惊心动魄的几十小时清盘危机讲得举重若轻，硬核的量化概率推演与“十亿美元 PDF”的戏谑调侃无缝交织。而在激辩法拉利 Luce 的品牌定位时，两人的观点交锋针锋相对、火花四溅，随后又在探讨深绿演播室美学与互怼看没看过漫威电影的松弛玩笑中收尾，听感节奏张弛有度，极具信息密度与陪伴感。 ## Bulletin - [00:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=1s) 前 OpenAI 研究员利奥波德·阿申布伦纳创办的 Situational Awareness 基金因 AI 基础设施板块暴跌遭遇保证金追缴，被迫清空全部公开股票，随后被肯·格里芬旗下的城堡投资（Citadel）全盘接盘。 - [06:36](https://www.youtube.com/watch?v=7Uz5py036MA&t=396s) 量化对冲基金 Numerai 创始人克雷布指出，Situational Awareness 投资组合年化波动率高达 150%，单是每年 113% 的方差拖累就让爆仓概率近乎抛硬币，结局与其对 AI 趋势预测的技术判断无关。 - [10:25](https://www.youtube.com/watch?v=7Uz5py036MA&t=625s) 法拉利由前苹果首席设计官乔尼·艾夫操刀的首款纯电超跑 Luce，虽在设计上饱受争议，但凭借中国市场的强劲需求，已在发布两个月后提前达成 2026 年销售目标（首年略低于 500 辆）。 - [12:29](https://www.youtube.com/watch?v=7Uz5py036MA&t=749s) 法拉利严禁经销商将单价 55 万欧元的 Luce 与稀缺限量款（如 SP3、F80）配额强行搭售，但市场普遍讨论买家是否仍为讨好车行或博取绝版历史孤品而买单。 ## Key takeaways - [00:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=1s) Situational Awareness 基金因建立高杠杆单向多空敞口（重仓 Nebius、SanDisk、Micron、CoreWeave 等硬件算力股，做空 Adobe 等应用层）遭遇挤压并被强平。 - [00:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=1s) AI 板块的估值回撤不仅关乎基本面疑虑，还交织着美联储利率预期、油价通胀阴云及房贷利率飙升至 6.66% 的宏观紧缩恐慌。 - [03:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=181s) 阿申布伦纳的宏大底层叙事源于其轰动业界的长文《Situational Awareness》，断言 AGI 将倒逼全球对芯片和算力基础设施进行海量资本开支。 - [03:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=181s) 主流财经媒体在 12 小时内接力报道了该基金的危机发酵：从前夜彭博社爆料紧急求援、FT 披露向 LP 发信推销，到清晨 CNBC 确认爆仓清算，最终 WSJ 敲定 Citadel 接盘。 - [04:40](https://www.youtube.com/watch?v=7Uz5py036MA&t=280s) 对冲基金由于存在杠杆和保证金追缴机制，崩溃往往在数十小时内发生；而纯股权 VC 生态即使遭遇项目失败，也可以通过削减成本和多周期平仓缓慢消化风险。 - [06:36](https://www.youtube.com/watch?v=7Uz5py036MA&t=396s) 宏观变量与散户模因狂热会彻底破坏第一性原理的技术推演，导致基于“算力增长倍数”计算的基础设施估值模型在短期内失效。 - [06:36](https://www.youtube.com/watch?v=7Uz5py036MA&t=396s) 该基金的 LP 多为身家数十亿美元的科技巨头创始人，仅将极小比例资产（1%-5%）作为超大号杠杆看多 AI 的期权彩票，整体风险承受能力极强。 - [09:16](https://www.youtube.com/watch?v=7Uz5py036MA&t=556s) 市场传闻 Citadel 以跳楼价（pennies on the dollar）接盘很可能被夸大，业内推测其实际成交价或达到账面价值的 50% 至 80%。 - [10:25](https://www.youtube.com/watch?v=7Uz5py036MA&t=625s) 法拉利纯电超跑 Luce 单价约 55 万欧元，计划在 2030 年前累计销售 2,500 辆（年均约 500 辆），首年配额在 7 月初即宣告售罄。 - [12:29](https://www.youtube.com/watch?v=7Uz5py036MA&t=749s) 围绕 Luce 的争议在于定位：支持者认为其作为高辨识度日常代步车具有历史孤品收藏价值，批评者则指出若无法成为 Cayenne 或 Urus 式的走量现金奶牛，可能会透支跑车品牌声誉。 - [17:20](https://www.youtube.com/watch?v=7Uz5py036MA&t=1040s) TBPN 演播室标志性的深绿色调视觉设计纯属偶然决定，灵感来源于 Pinterest、摄影作品及 Ralph Lauren 美学，并非故意模仿《蜘蛛侠》中的《号角日报》直播间。 ## Detailed walkthrough ### [00:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=1s) 算力杠杆遭遇强平：Citadel 抄底 Situational Awareness 基金 社交媒体时间线上昨夜今晨彻底陷入混乱。CNBC 在周四早间率先曝出重磅消息：由前 OpenAI 研究员利奥波德·阿申布伦纳（Leopold Aschenbrenner）掌舵的对冲基金 Situational Awareness，在 AI 基础设施板块遭受灾难性亏损后，已被迫清仓其全部公开股票投资组合。主经纪商（Prime Brokers）火急火燎地要求其筹措现金以满足追加保证金要求，而肯·格里芬（Ken Griffin）旗下的城堡投资（Citadel）——主持人 John Coogan 的老东家——则火速达成协议，全盘接盘该基金遭抛售的公开交易资产。 演播室里随即开启了一段关于老东家的调侃。Jordi 调侃 Coogan 当年在 Citadel 实习时，肯·格里芬是否也曾为了考验一个小实习生而故意掀起市场风暴，Coogan 笑着自嘲当时“哪怕能帮肯倒杯咖啡都算烧高香了”。不过，Citadel 本身的发展史就是一场惊心动魄的豪赌：在 2008 年金融危机期间，该基金净值曾腰斩 50%，陷入极度灰暗的时刻。Coogan 借此科普了经典的金钱数学逻辑——暴跌 50% 之后再涨 50% 绝不是回本，你必须实现 100% 的涨幅才能收复失地。 Situational Awareness 此前建立了极度集中的投资组合：重仓押注 Nebius、SanDisk、美光（Micron）和 CoreWeave 等 AI 算力与基础设施硬件公司，同时大举做空 Adobe 等传统应用层软件企业。然而，近几周 AI 算力基础设施概念股遭遇断崖式下挫，这一多空策略瞬间被双向挤压击溃。尽管由于美联储昨日维持利率不变、以及 Citadel 接盘消息刺激，相关个股在周四录得两位数的大幅反弹，但从月度周期看依然跌幅惨重。暴跌背后交织着多重宏观因素：对 AI 能否兑现商业价值的疑虑、开源大模型的冲击、油价攀升引发的通胀阴云，以及房贷利率飙升至 6.66% 一年多高位所带来的紧缩恐慌。 - **重仓标的**：多头集中于 Nebius、SanDisk、Micron、CoreWeave；空头押注 Adobe 等应用层 SaaS。 - **市场宏观**：周二 Citadel 曾预期美联储可能意外加息，引发市场提前抛售；周三决议维持利率不变，但房贷利率仍升至 6.66% 的高位。 --- ### [03:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=181s) 从“十亿美元 PDF”到惊魂 12 小时：媒体接力爆料时间线 利奥波德·阿申布伦纳之所以名声大噪，始于他在 2024 年发表的轰动性系列长文《Situational Awareness》（态势感知）。投资人 Will Manidis 曾形象地将其戏称为“价值十亿美元的 PDF”。阿申布伦纳在文中断言，超级人工智能（AGI）的飞速演进将倒逼全球对芯片、显存、电力和算力基础设施进行前所未有的海量资本投入。这一激进宏大的底层叙事直接构成了他创办对冲基金的投研基石。此外，其未婚妻正是 Anthropic 首席执行官 Dario Amodei 的幕僚长，这一背景让他在硅谷 AI 圈的地位格外瞩目。 然而，过去的半天时间内，主流财经媒体以极快的节奏报道了该基金的坍塌过程： ```mermaid flowchart LR accTitle: Situational Awareness 基金清盘时间线 accDescr: 展现从前一晚彭博社报道募资到次日早晨 Citadel 确认接盘的媒体报道时间轴。 subgraph Timeline ["危机发酵 12 小时"] A["前夜 21:25 (Bloomberg)<br/>曝出遭遇暴跌后紧急求援募资"] --> B["深夜 (FT)<br/>传出向 LP 发信推销抄底"] B --> C["清晨 06:05 (CNBC)<br/>确认遭追加保证金，被迫清仓公募股票"] C --> D["早间 08:39 (WSJ)<br/>Citadel 确认介入全面接盘"] end ``` 昨夜 21:25，彭博社率先披露 Situational Awareness 在 AI 板块遭遇血洗后正紧急寻求外部资本补充。随后《金融时报》跟进爆料，称有 LP 泄露了基金信件，信中阿申布伦纳仍在极力推销：“现在是绝佳的买入时机，宏大叙事坚不可摧，若要追加注资，当下就是最好窗口。”但在被追缴保证金的生死边缘，这种“低买高卖”的说辞显得苍白而残酷。紧接着在早晨 06:05，CNBC 投下震撼弹，证实阿申布伦纳在巨额亏损下已无法支撑，被迫全盘清算所有公开市场头寸；到了 08:39，《华尔街日报》最终敲定定局：Citadel 已正式介入接盘该清盘资产。 --- ### [04:40](https://www.youtube.com/watch?v=7Uz5py036MA&t=280s) 杠杆的致命速度：对比 VC 缓慢失血与对冲基金瞬间猝死 经历过 FTX 帝国坍塌和硅谷银行（SVB）闪崩的科技圈，再次在社交媒体上见证了极速覆灭的一幕。主持人们感叹，一旦体系中引入了高杠杆（Leverage），事态恶化的速度就会呈几何级爆发。回顾 FTX 和 SVB 的崩溃过程，最初只是社交媒体上的些许耳语与传闻，随后当事人发文声称“资金一切安好”，然而在短短几天之内，整个多米诺骨牌便彻底坍塌。 这与传统风险投资（VC）生态有着云泥之别。在没有杠杆的纯股权投资市场中，一家创业公司的消亡往往需要耗费两到三年甚至更长时间。以共享滑板车独角兽 Bird 和 Lime 为例，即便业务跌入谷底，因为账面上的融资全为实打实的股权现金，公司可以通过削减开支、裁员缩减规模，将原本 12 到 18 个月的跑道硬生生拉长到 24 个月。当发现融资无望时，再次缩减成本苟延残喘，Bird 最终彻底清算花掉了整整五年时间。 不仅如此，传统 VC 的投资组合即便在宏观上存在关联性，在实际清算时却可以解耦（Decorrelated）。风投机构账面上的某一处败局（如 Bird），往往能被同期的成功押注（如 SpaceX）所中和稀释，风投机构有数年时间在不同市场周期慢慢平仓，绝不会像带杠杆的公开市场对冲基金那样，在几十个小时内因为保证金爆仓而被迫“全部一次性出清”。 --- ### [06:36](https://www.youtube.com/watch?v=7Uz5py036MA&t=396s) 波动率拖累与爆仓概率：量化视角与 LP 的实际敞口 著名量化对冲基金 Numerai 创始人理查德·克雷布（Richard Craib）在社交媒体上发表了一段辛辣而深刻的评论。他直言，这类狂热信念驱动的基金能够存在于美国市场固然很酷，但“这一结局根本无关其对 AI 趋势预测的对错”。克雷布通过量化模型指出，当一个投资组合的年化波动率（Volatility）高达 150% 时，仅“方差拖累”（Variance Drag）一项每年就会吞噬高达 113% 的收益，在整个基金生命周期内其爆仓破产的概率（Risk of Ruin）基本上就是抛硬币。 Claude 模型计算得出的结论同样显示破产并非小概率事件，其风险几近五五开。阿申布伦纳遭遇的窘境在于，现实市场并非纯粹的技术推演——当油价、地缘政治、利率预期等宏观变量引发市场波动，叠加狂热散户涌入把 AI 概念股炒成模因股（Meme stocks）时，原本基于第一性原理计算“算力增长倍数（OOMs）”和“Token 边际成本”的技术推演体系便彻底失灵了。 面对社交媒体上四处狂欢的梗图和冷嘲热讽的“事后诸葛亮”，主持人们也客观剖析了该基金出资人（LP）的真实心态。出资支持 Situational Awareness 的许多顶层大佬本身就是身家数十亿美元的科技巨头创始人，他们已经在 AI 大模型实验室和核心生态中持有巨额现货资产。对他们而言，配置该基金可能仅仅占其总资产的 1% 到 5%，本质上是用极小的仓位买一张“超大号杠杆看多 AI（Giga Long）”的彩票。即便爆仓归零，在宏观资产配置上也算是一种务实的风险对冲，远未达到外界所渲染的倾家荡产。 ### [09:16](https://www.youtube.com/watch?v=7Uz5py036MA&t=556s) 对冲基金的极速回撤与清算争议 主持人在开头讨论了风险投资与二级市场基金的激进操作风格。对于某些基金经理而言，投资者的态度往往是放手让他们“放手去干、随心所欲（Go crazy. Do whatever you want）”，这也正是客户购买该产品所追求的超额敞口。然而，市场风云变幻的速度令人咋舌：某家管理规模（AUM）一度在 6 月底至 7 月初达到约 450 亿美元的基金，连为期三个月的回撤期都未能撑过，便迅速遭遇重创。主持人感叹世事变化之快，并联想到前 FTX 团队成员利奥波德·阿申布伦纳（Leopold Aschenbrenner），指出他在 FTX 崩塌时曾与 FTX Future Fund 团队一同立刻辞职，如今又遭遇了极为相似的剧情。 针对外界关于“投资组合被肯·格里芬（Ken Griffin，Citadel 掌门人）以几分钱的跳楼价甩卖收购（Pennies on the dollar）”的传言，主持人提出了质疑。通常人们听到“pennies on the dollar”会以为资产回收率只有 5% 左右，但他推测实际成交价可能高达账面价值（Book value）的 50% 甚至 80%，外界的悲观叙事很可能夸大了折价幅度。 ### [10:25](https://www.youtube.com/watch?v=7Uz5py036MA&t=625s) 法拉利纯电超跑 Luce 提前锁定 2026 年销售目标 随着话题转向汽车行业，主持人带来了《金融时报》（Financial Times）关于法拉利首款纯电动车型 Luce 的重磅报道。这款由前苹果首席设计官乔尼·艾夫（Jony Ive）操刀设计的电动超跑，尽管自发布以来因其打破常规的前卫造型遭到了投资者和车迷的强烈反对与嘲弄（甚至有前 CEO 和前设计主管公开批评“这根本不是一辆法拉利”，网友更是讥讽“用 ChatGPT 都能画得比这好”），但法拉利官方却宣布该车凭借来自中国市场的强劲需求，已提前达成 2026 年的销售目标，让所有唱衰者大跌眼镜。 知情人士透露，法拉利虽然未公开 Luce 的具体配额，但内部定下的首年销售目标是略低于 500 辆，并计划在 2030 年前的四至五年内累计售出 2,500 辆（即年均 500 辆左右）。该车单价高达 55 万欧元（约合 65 万美元）。令人意外的是，今年 7 月上旬——即引发巨大争议的发布会仅仅过去两个月后——首批目标配额就已宣告售罄。此外，法拉利向经销商下达了严令，禁止强迫传统燃油车收藏家购买这款电动车，明确表态不会把购买 Luce 作为获取限量版跑车（如 SP3、SP4 或特别版 F80）配额的强制搭售条件，将选择权完全交给客户。 - **车型与设计师**：Ferrari Luce，由 Jony Ive 操刀设计。 - **售价**：约 550,000 欧元（折合约 650,000 美元）。 - **销量目标与节奏**：首年（2026）目标略低于 500 辆，到 2030 年累计规划 2,500 辆，年均约 500 辆。 - **市场表现**：发布仅两个月（7 月初）即完成首年配额，中国买家需求尤其强劲。 - **销售政策**：法拉利严禁经销商将 Luce 与稀缺限量款配额（SP3、SP4、F80 系列）恶意绑定搭售。 ### [12:29](https://www.youtube.com/watch?v=7Uz5py036MA&t=749s) 深度激辩：配额潜规则、品牌历史还是日常超跑？ 在接下来的讨论中，两位主持人对 Luce 的市场反响展开了多角度交锋。在导播误切了一张无关车辆的乌龙照片后，Jordi 率先表达了质疑：虽然他极度赞赏该车的内饰做工，但实车上路视频看起来比静态官图还要怪异，令人摸不准目标受众。针对法拉利宣称的“不强迫配额”，Jordi 认为这纯属冠冕堂皇的说辞。在法拉利经销商的生态潜规则中，客户是否真金白银买下冷门车型，绝对会直接影响未来顶配神车的配额权重；现实中许多顶级收藏家哪怕不喜欢，也会捏着鼻子买下两台同款车以讨好车行、维持“忠实藏家”地位。因此，年销 500 辆对于法拉利这种级别的品牌来说只是最基本的保底数字，考虑到它作为日常代步电动车的全新定位，若没有吸引到大量非跑车用户增量，其实很难称得上巨大成功。 ```mermaid flowchart TD accTitle: 法拉利 Luce 购买动机与市场定位分歧 accDescr: 展示藏家心理与产品定位的多种不同维度 subgraph 藏家心理驱动 ["藏家购买动机"] A["经销商配额博弈<br/>（买 Luce 为未来抢购神车铺路）"] B["历史孤品押注<br/>（豪赌其成为法拉利绝版电动先驱）"] C["全谱系终极收藏<br/>（集齐前置、中置、SUV、F1全品类）"] end subgraph 产品定位争议 ["产品战略与定位争议"] D["日常化破局点<br/>（高辨识度、离经叛道且兼顾代步）"] E["品牌伤害风险<br/>（未能成为 Cayenne/Urus 式的走量爆款，恐稀释品牌价值）"] end A --> D B --> D C --> D D -.-> E ``` 对此，另一位主持人提出了三点针锋相对的反驳：第一，审美存在习惯效应，实车动态影像越看越顺眼，先锋前卫的外观正在被更多人接受；第二，藏家群体中有真正的“法拉利原教旨主义者”，他们即便不为配额，也会将 Luce 视作极具纪念意义的汽车工业切片——万一法拉利未来 30 年都不再生产纯电动车，Luce 就将成为绝无仅有的历史孤品；此外，还有硬核藏家热衷于集齐法拉利的所有形态（从自吸、中置、前置、Purosangue SUV 到 F1 赛车）。第三，法拉利制造 Luce 从来不是为了复刻保时捷 Cayenne 或兰博基尼 Urus 那样大批量走量换取现金流的商业策略，它更像是一款极度叛逆、高调且反传统的实验性作品，“既怪诞独特，又能兼顾日常代步”，这恰恰是它的最大卖点。不过，Jordi 坚持认为，如果一款背离传统的车不能像 Cayenne 或 Urus 那样为整个品牌的研发提供庞大现金奶牛支持，它对高端跑车品牌声誉的透支就得不偿失。 ### [17:20](https://www.youtube.com/watch?v=7Uz5py036MA&t=1040s) 场外趣闻：TBPN 演播室美学撞脸《蜘蛛侠》？ 在节目的最后环节，话题跳脱到了流行文化。主持人询问搭档 Jordi 是否看过电影《蜘蛛侠：英雄无归》（Spider-Man: No Way Home），而自称“最近看了《奥德赛》、已经进阶为严肃电影鉴赏家”的 Jordi 则斩钉截铁地表示完全没看过这类爆米花超英片。 主持人分享了社交平台 X 上的一个趣味讨论：网友 Rob Felt 在温习《蜘蛛侠：英雄无归》准备迎接新电影《崭新之日》（Brand New Day）时，猛然发现片中 J·乔纳·詹姆森（J. Jonah Jameson）所主持的《号角日报》（The Daily Bugle）网络直播间，其背景配色和视觉风格居然与 TBPN 的演播室布置如出一辙。网友甚至怀疑，身兼 TBPN 品牌视觉架构师的 Jordi 是否偷偷借鉴了电影里的置景。对此，两位主持人笑着辟谣：这纯属巧合，演播室的标志性绿色调完全是某天清晨两人一起健身撸铁时拍脑袋决定的灵感，与《蜘蛛侠》毫无瓜葛。 ### [18:50](https://www.youtube.com/watch?v=7Uz5py036MA&t=1130s) 演播室美学与绿色调：TBPN 视觉风格的灵感来源与收尾 两位主持人回顾了节目独树一帜的演播室视觉设计与品牌调性，特别是标志性深绿色的选定过程。其中一人回忆道，当时聊到用绿色时，对方兴奋地表示：“从来没人用过绿色。”尽管这一说法立刻遭到反驳——“这不对吧，Robinhood 就是绿色的，用绿色的多得是”，但对方依然坚称：“科技圈里从来没有人用过绿色。”不过在播客领域，这种墨绿色调确实算是一片“空白市场”（white space），因为在他们的印象中，确实没有其他哪档科技播客会采用这种暗绿色作为演播室的主背景。 为了打磨出满意的视觉呈现，团队在筹备期间大量翻阅了 Pinterest 寻找灵感参考，广泛涉猎了各类摄影作品、复古商品目录以及 Ralph Lauren 的经典型录。面对搭档关于“你当时究竟参考了哪些图片”的追问，两人互相开起玩笑——除了众所周知从 F1 赛车文化中汲取的元素之外，搭档调侃对方是不是完全凭借自己的大脑直接构想：“你难道只靠脑子？虽然平时也没见你往脑子里装多少东西”，随后更是补刀调侃其脑子里“肯定装不下电影”。在轻松幽默的互怼中，主持人向观众致意，提醒大家访问 tbpn.com 订阅每日新闻通讯（newsletter），并相约明天的节目再见。 - **品牌视觉差异化**：科技圈内虽有 Robinhood 等以绿色为代表品牌的先例，但在播客演播室设计中，深绿色背景构成了独特的视觉辨识度。 - **美学参考来源**：设计灵感主要来源于 Pinterest、各类摄影图集、复古目录、Ralph Lauren 风格美学以及 F1 赛车视觉元素。 ## Notable quotes > [00:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=1s) 哪怕能帮肯倒杯咖啡都算烧高香了 — John Coogan > [03:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=181s) 价值十亿美元的 PDF — Will Manidis > [03:01](https://www.youtube.com/watch?v=7Uz5py036MA&t=181s) 现在是绝佳的买入时机，宏大叙事坚不可摧，若要追加注资，当下就是最好窗口。 — Leopold Aschenbrenner > [04:40](https://www.youtube.com/watch?v=7Uz5py036MA&t=280s) 资金一切安好 > [06:36](https://www.youtube.com/watch?v=7Uz5py036MA&t=396s) 这一结局根本无关其对 AI 趋势预测的对错 — Richard Craib > [09:16](https://www.youtube.com/watch?v=7Uz5py036MA&t=556s) 放手去干、随心所欲（Go crazy. Do whatever you want） > [10:25](https://www.youtube.com/watch?v=7Uz5py036MA&t=625s) 这根本不是一辆法拉利 > [10:25](https://www.youtube.com/watch?v=7Uz5py036MA&t=625s) 用 ChatGPT 都能画得比这好 > [12:29](https://www.youtube.com/watch?v=7Uz5py036MA&t=749s) 既怪诞独特，又能兼顾日常代步 — John Coogan > [18:50](https://www.youtube.com/watch?v=7Uz5py036MA&t=1130s) 科技圈里从来没有人用过绿色。 — Jordi Hays > [18:50](https://www.youtube.com/watch?v=7Uz5py036MA&t=1130s) 你难道只靠脑子？虽然平时也没见你往脑子里装多少东西 — John Coogan ## Mentioned TBPN, John Coogan, Jordi Hays, Citadel, Ken Griffin, Leopold Aschenbrenner, Situational Awareness, OpenAI, Nebius, SanDisk, Micron, CoreWeave, Adobe, Will Manidis, Dario Amodei, Anthropic, CNBC, 彭博社 (Bloomberg), 金融时报 (Financial Times), 华尔街日报 (The Wall Street Journal), FTX, 硅谷银行 (SVB), Bird, Lime, SpaceX, Numerai, Richard Craib, Claude, 法拉利 (Ferrari), Ferrari Luce, Jony Ive, Ferrari SP3, Ferrari SP4, Ferrari F80, Ferrari Purosangue, 保时捷 Cayenne, 兰博基尼 Urus, 《蜘蛛侠：英雄无归》(Spider-Man: No Way Home), Rob Felt, J·乔纳·詹姆森 (J. Jonah Jameson), 《号角日报》(The Daily Bugle), Robinhood, Pinterest, Ralph Lauren ## Video description Sign up for TBPN’s daily newsletter at TBPN.com Diet TBPN delivers the best of today’s TBPN episode in 30 minutes. TBPN is a live tech talk show hosted by John Coogan and Jordi Hays, streaming weekdays 11–2 PT on X and YouTube, with each episode posted to podcast platforms right after. Described by The New York Times as “Silicon Valley’s newest obsession,” the show has recently featured Mark Zuckerberg, Sam Altman, Mark Cuban, and Satya Nadella. TBPN is made possible by: Ramp - https://ramp.com Public - https://public.com Cisco - https://www.cisco.com Console - https://www.console.com CrowdStrike - https://www.crowdstrike.com Figma - https://www.figma.com MongoDB - https://www.mongodb.com NYSE - https://www.nyse.com Railway - https://railway.com Shopify - https://www.shopify.com/ Follow TBPN: https://TBPN.com https://x.com/tbpn https://open.spotify.com/show/2L6WMqY3GUPCGBD0dX6p00?si=674252d53acf4231 https://podcasts.apple.com/us/podcast/technology-brothers/id1772360235 https://www.youtube.com/@TBPNLive [Source](https://www.youtube.com/watch?v=7Uz5py036MA)

[0:01] John Coogan: 哪怕能帮肯倒杯咖啡都算烧高香了 [3:01] Will Manidis: 价值十亿美元的 PDF [3:01] Leopold Aschenbrenner: 现在是绝佳的买入时机，宏大叙事坚不可摧，若要追加注资，当下就是最好窗口。 [4:40] 资金一切安好 [6:36] Richard Craib: 这一结局根本无关其对 AI 趋势预测的对错 [9:16] 放手去干、随心所欲（Go crazy. Do whatever you want） [10:25] 这根本不是一辆法拉利 [10:25] 用 ChatGPT 都能画得比这好 [12:29] John Coogan: 既怪诞独特，又能兼顾日常代步 [18:50] Jordi Hays: 科技圈里从来没有人用过绿色。 [18:50] John Coogan: 你难道只靠脑子？虽然平时也没见你往脑子里装多少东西

[4:40] Richard Craib: The outcome was never about being right or wrong on AI. At 150% vol, variance drag alone is 113% a year, and risk of ruin is roughly a coin flip over the fund's life. [4:40] Richard Craib: a child can do the math on a napkin Silicon Valley’s newest obsession [0:01] Will Manidis: the billion-dollar PDF [15:35] iconic, very iconoclastic, very contrarian... bold and weird and different [17:20] Jordi Hays: We should do green [17:20] Jordi Hays: no one in tech has ever used green before [17:20] John Coogan: Robin Hood is green

[0:01] Absolute chaos on the timeline last [0:04] night, this morning, around Leopold [0:06] Auerbach's hedge fund situational [0:09] awareness. They have been forced to [0:11] unwind their public stock portfolio [0:13] after steep losses on AI infrastructure [0:17] bets. CNBC reported this on Thursday, [0:20] today. [0:21] Uh [0:22] Prime brokers reportedly rushed to raise [0:24] cash to meet margin requirements while [0:26] Ken Griffin's Citadel, my former [0:29] employer, [0:30] >> [laughter] [0:30] >> uh reached a deal to purchase the fund's [0:32] publicly traded assets. [0:34] >> Did Ken ever do anything to you? [0:37] >> No. [0:37] >> Like this? [0:38] >> No. [0:38] >> [laughter] [0:39] >> What? No. [0:40] >> You were just an intern. [0:40] >> I was an intern. I would I I would have [0:42] loved to be getting him coffee or [0:44] something. [0:44] >> to maybe intentionally send the markets [0:46] into turmoil just to test you? [0:48] >> No, but I mean, honestly, the story of [0:50] Citadel is is crazy. I mean, after the [0:53] the housing crisis, the fund was down [0:55] 50% and it was a very very dark time. [0:58] There were building [0:59] >> Ken that at that time, Ken Griffin, what [1:02] he would become didn't exist. And [1:04] otherwise, he probably would have eaten [1:05] the young Ken Griffin alive. [1:07] >> [laughter] [1:07] >> Yeah, maybe. Maybe. [1:08] >> Much like it seems that Ken has done to [1:11] uh Leopold. [1:12] >> Yeah, maybe. Maybe. The other The other [1:14] frustrating thing is that they went down [1:15] 50%. I think the next year they went up [1:17] 50% and then [1:19] and they were, you know, this is the [1:20] classic, you know, explaining [1:22] fund math to people. Oh, you're back up [1:24] 50%. Great. You're back to where you [1:25] were? Nope. You need to go back up 100% [1:28] if you're down 50%. Of course. Anyway, [1:30] the fund had built concentrated [1:32] positions in AI infrastructure companies [1:34] including Nebius, Sandisk, Micron, and [1:37] Coreweave, while also betting against [1:39] software companies such as Adobe. Those [1:42] trades have as unraveled as AI [1:45] infrastructure stocks plunged in recent [1:47] weeks before rebounding sharply today. [1:49] Now, how much of the plunge is around [1:53] shaken faith in AI's ability to deliver [1:56] value, open source, or just oil [1:59] inflation. [2:01] The Fed's actions, we'll get into all of [2:03] this because there's a lot of moving [2:05] different moving pieces that led us to [2:07] where we are today. So, also relevant [2:10] here is from the TBPN newsletter. You [2:12] can go sign up at tbpng.com. And what a [2:14] bunch of people are pointing out [2:15] actually is the fact that on Tuesday it [2:17] was reported that Citadel expected a [2:19] surprise rate hike from the Fed meeting [2:23] that took place yesterday. Which did [2:26] coincide with more sell-off in the [2:27] market. So, the market has been selling [2:29] off based on what might happen at the [2:31] Fed. We reported on the the Fed news. [2:33] There were three Fed governors that said [2:36] we should raise rates, but the rate held [2:38] steady, but mortgage rates are high are [2:40] at over a one-year high at 6.66% [2:43] today. Very odd number. But yesterday [2:46] the Fed left the rate unchanged, and [2:49] today many of the stocks in Leopold's [2:50] portfolio are up double digits, and [2:52] we'll sort of go through them. They're [2:53] up today based on the news that Ken [2:55] Griffin is buying the portfolio, but [2:57] they are still down over the last month, [2:59] for example, in many cases. [3:01] Ash Brennen, a former OpenAI researcher, [3:03] rose to prominence after publishing his [3:04] 2024 essay series situational awareness, [3:08] the billion-dollar PDF, as Will Manidis [3:10] put it, I believe, which argued that [3:12] rapid AI progress would require an [3:14] enormous build-out of chips, memory, [3:17] power, and compute infrastructure. That [3:19] thesis became the foundation of his [3:21] investment strategy. After launching the [3:22] fund, he also engaged He is also engaged [3:25] to Anthropic CEO Dario Amodei's chief of [3:27] staff. The news is coming in hot and [3:29] fast on the story. Here's a timeline of [3:30] the most important headlines so far. So, [3:32] Bloomberg, 9:25 p.m. yesterday. I [3:35] remember I think you texted me this as I [3:37] was going to sleep. [3:39] We're like, "Whoa, this is big deal." I [3:41] wonder [3:42] how how crazy this will get over the [3:44] next few days. It got very crazy very [3:45] quickly. So, Bloomberg reported Leopold [3:48] Ash Brennen's situational awareness [3:49] seeks to raise capital after AI route. [3:52] There was an article in the Financial [3:53] Times as well last night just saying [3:55] that hey, there's some there's some [3:57] rumors that are leaking out from LPs [4:00] that they got a letter saying like, hey, [4:01] the market's down now's a good buying [4:03] opportunity. The thesis is as strong as [4:05] ever. If there was ever a time to put [4:07] more money into this fund, now's the [4:08] time. That can be good. You want to be [4:10] buying when things buy low, sell high, [4:12] right? But at the same time, if it if [4:14] it's to cover margin calls, it's if it's [4:16] because the fund's getting beat up, it's [4:17] a little bit rougher of a pitch. Then at [4:20] 6:05 a.m. CNBC announces that AI [4:23] investor Leopold Auerbach has been [4:25] forced to unwind all public stock [4:27] positions after steep losses according [4:29] to CNBC sources. And then the Wall [4:31] Street Journal reports at 8:39 a.m. that [4:34] Citadel has stepped in to buy [4:36] Situational Awareness's stock portfolio [4:38] after big losses in AI. And so [4:40] >> After living through FTX and SVB [4:43] and now this on the timeline. [4:46] Uh no. [4:47] >> You're just grizzled. [4:48] >> No, the the my key the key takeaway is [4:50] like when leverage is involved [4:53] >> Mhm. [4:53] >> things just move so so so fast, right? [4:56] You remember with FTX [4:58] >> Yep. [4:58] >> just kind of some rumblings [4:59] >> Yep. [5:00] >> a couple couple posts from SBF saying [5:02] like, we're fine, it's all good, and [5:04] then it was over. [5:05] >> And it is [5:05] >> And then the same thing with SVB, like [5:08] couple rumblings, maybe like a [5:11] couple weeks a week beforehand, few [5:13] posts here and there, and then it just [5:15] moved so so fast, right? Yeah, quite a [5:17] bit different than traditional venture [5:19] world where when a company's dying, it [5:22] dies over [5:23] two, three years, often sometimes more. [5:26] >> Yeah. Yeah. Yeah. Yeah. [5:27] I'm thinking of like I mean we've had [5:29] some of these companies on from the [5:30] private markets where they've gone [5:32] through big booms and busts like Bird [5:34] and then they built back and they've [5:35] turned around, but there's so much more [5:39] No, I don't know. Lime turned around, [5:41] but but but it took like an extra five [5:44] years for Bird to actually wind down, [5:46] and it's because there's no leverage in [5:47] the system. There's just a bunch of [5:49] dollars that sit there as equity, and [5:52] those get burned down, but every every [5:54] month if the business is deteriorating, [5:56] you're cutting costs, shrinking the [5:57] business, tightening things up, making [5:59] that 12 to 18 months last 24 months, and [6:01] then you wind and then you wind up, you [6:03] know, 24 months in and you're like, "Oh, [6:04] we're not going to be able to raise [6:05] again. Let's stretch this again." And [6:07] and right-size the business again. And [6:09] all of a sudden, it's so it takes like [6:11] years and years for these things to [6:12] unwind. Although they are correlated in [6:14] the venture world, they can be [6:15] decorrelated in the unwinding process, [6:18] and then there can be other things that [6:19] are outweighing the portfolio. So, every [6:21] VC that had Bird on their books probably [6:23] also had some SpaceX on their books or [6:25] something. And so, there's this [6:27] balancing effect, and it takes it takes [6:29] years for these things to balance out, [6:30] and they can be unwound at different [6:32] periods in the market as opposed to [6:34] everything needing to happen all at [6:35] once. So, I like this post from Richard [6:37] Craib, [6:38] one of my favorite investors. He runs [6:40] the quant hedge fund Numerai, and he [6:43] says, "I think it's cool that funds like [6:45] Situational Awareness can exist in [6:47] America, and that there's a market for [6:49] them, but the outcome was never about [6:50] being right or wrong on AI. At 150% vol, [6:55] variance drag alone is 113% [6:59] a year, and risk of ruin is roughly a [7:01] coin flip over the fund's life. A child [7:03] can do the math on a napkin. Claude did [7:06] it for him. AI says ruin wasn't [7:08] unlikely, it was roughly even money. So, [7:11] there's a 50/50% chance 50% chance that [7:14] that the fund sees you know, so many [7:17] losses that they have to that they have [7:18] to do this liquidation process. And [7:20] that's basically what happened. And it's [7:22] it must be so frustrating because this [7:24] is not really does not feel like, "Oh, [7:26] Leopold was wrong about AI and the AI [7:29] bailout." It's like, "Well, there's oil [7:31] and a war and interest rates and all [7:34] these other things going on that are [7:36] creating some jitters." And then also, [7:38] once the AI trade and the the [7:40] trade got so big, you wind up with like [7:44] this retail froth on top that makes [7:46] things even crazier. And then [7:49] >> All that he was benefiting from. [7:50] >> Benefiting from, but also it's much [7:52] harder to do sort of a first principles [7:54] analysis on what the psychology of a [7:57] frothy market will do as opposed to just [8:00] retreating to, "Okay, well, there's this [8:03] the model progress is progressing like [8:05] this and and token pricing is is is you [8:08] know, counting the ooms." Stop sort of [8:10] working when it's like, "Well, will this [8:13] particular stock become a meme stock, [8:15] right?" [8:15] >> Pull up this picture. There was a lot of [8:17] this going on [8:19] this morning. [8:19] >> But the memes are flying. This is truly [8:21] like [8:21] >> This was This was my This was my [8:24] the the first meme that popped into my [8:26] head of people saying [8:28] Oh, I don't know why the guy's head's [8:29] cut off. Are you guys okay? [laughter] [8:31] >> This was you. [8:33] >> No, this was this was just a lot of [8:34] people on the timeline being like, "I [8:35] knew he would blow up." [8:36] >> [laughter] [8:37] >> Yeah, yeah, yeah, yeah. [8:38] The armchair experts are out in full [8:41] force today and [8:42] in many ways we are among them. [8:44] >> So, I like to think about it like at [8:46] least some of the more high-profile LPs [8:48] >> Yeah. [8:49] >> that [8:49] in in situational awareness [8:51] >> a lot of them are like, you know, great [8:53] founders. You know, they you know, maybe [8:55] they have big big positions in in the [8:58] labs and all these different things and [9:00] it's quite possible that situational [9:02] awareness, at least when they invested, [9:04] was like 5% of their portfolio and [9:06] they're just thinking like [9:07] >> go giga long. Like go [9:10] >> Yeah, yeah. [9:10] >> of these people it might be like [9:12] >> one less than 1%, right? Whatever it is. [9:14] >> Have you ever [9:14] >> it's it's actually somewhat pragmatic [9:16] for them to just be like, "Yeah, go [9:18] crazy. Do whatever you want." [9:19] >> Yeah, that's the product. [9:20] >> That's the product. That's what I want [9:22] to buy. The rest of my portfolios fine. [9:25] You're going to have a a lot of AI [9:27] exposure whether you like it or not. But [9:29] yeah, it it's it it is actually crazy [9:33] that it didn't even take a three-month [9:36] drawdown, right? [9:37] >> Yeah. [9:37] >> Uh it was What was it? June June 1st, [9:41] they were at a 45 billion of AUM, [9:43] something like that was was the net. [9:44] >> end of June. [9:45] >> And end of June, so beginning of July. [9:48] And then how quickly how quickly things [9:51] can change and uh you know, poor poor [9:53] Leopold already went through this with [9:55] FTX. [9:56] I believe he uh he and the rest of the [9:59] the future FTX Future Fund team, I [10:02] believe resigned like right when [10:04] >> Oh, FTX collapsed. What's interesting is [10:06] people are people are uh framing this as [10:08] like they got pennies on the dollar or [10:10] or Ken Griffin bought the portfolio for [10:11] pennies on the dollar. And when I think [10:13] pennies on the dollar, I think like five [10:15] pennies per dollar, so like 5% recovery. [10:17] But it might be closer to like 50% of [10:20] book value. It might be 80% of book [10:22] value. I don't know. Uh there's news in [10:25] Ferrari world that the Ferrari EV, the [10:28] Luce, designed by Jony Ive, has already [10:32] hit the 2026 sales target. The haters [10:36] are in shambles. Everyone doubted that [10:39] this would sell, and the Italian car [10:41] maker reports strong demand from China [10:44] for electric model, derided for its [10:46] unconventional design. [10:48] Uh there's a whole bunch of interesting [10:50] tidbits in here uh in the Financial [10:52] Times article. Uh Ferrari has hit this [10:54] year's sales target for its first [10:55] electric vehicle on the back of strong [10:57] demand from China despite a polarizing [10:59] design that drew backlash from investors [11:01] and enthusiasts. Remember, even the [11:03] former CEO, former chief design officer, [11:06] former executive came out and said this [11:08] is not a Ferrari. There was a lot of uh [11:10] back and forth in the timeline. I could [11:12] have designed a better one with uh [11:14] ChatGPT. A lot of people threw out [11:15] different designs, uh but this one is [11:18] selling, at least according to the [11:19] Financial Times. The Italian group has [11:21] not disclosed the target for the Luce, [11:23] but two people with knowledge of the [11:24] matter said it had aimed to sell this [11:28] year just under 500 units of the EV. Not [11:31] a but their goal for by 2030 over the [11:35] next four years is to get to 2,500 [11:38] units. So, 500 a year for four or five [11:41] years. That's where they want to get [11:42] here, and they say they're on track. And [11:44] so, this is priced at 550,000 euros, [11:48] 650,000 USD, something like that. [11:51] One of the two people said the target [11:52] had been reached earlier in July, just [11:54] two months after its controversial [11:55] launch when critics on social media [11:57] derided its unconventional styling. [11:59] There's also an interesting line in here [12:01] that Ferrari says they gave strict [12:03] instructions to dealers not to force its [12:06] traditional petrol-loving collectors to [12:09] switch to electric cars. Said, "If you [12:11] want a Luce, we'll give you a Luce, but [12:12] we're not going to make you buy a Luce [12:14] in order to get in line for an SP3, SP4, [12:17] some special F80 thing that's more [12:19] limited. Just if you want it, it's here. [12:21] It's a choice." And a lot of people have [12:23] made [snorts] that choice according to [12:25] these insiders. What do you think? Give [12:27] me the pushback, and then we'll debate [12:28] it. [12:29] >> I was pausing because I wanted to start [12:31] with something nice. [12:33] Uh I think [12:35] So, the videos that I've seen of it on [12:37] the road, [12:38] it does look even more strange than in [12:41] the images. I'll say [12:42] >> Wait, what [laughter] is this photo? [12:43] >> That's not [12:44] What is that? [12:46] >> The team just [12:46] >> wrong [12:47] >> just accidentally put in some random [12:49] car. [12:50] >> Uh okay. So, so the car the car is [12:53] strange. I love the interior, but the [12:55] car overall is strange. It's still [12:57] unclear to me who it's really for, but [12:59] they are finding buyers. The idea that [13:01] that [13:03] buying the Luce like [13:05] Ferrari is saying, "We're not forcing [13:07] any dealer to push this car" or [13:10] whatever, but the idea that it's not [13:11] going to have some whether or not you [13:13] bought a Luce is going to have some [13:14] weight on your future allocations to me [13:17] is just insane. There's just no way that [13:19] that's true because every single dealer [13:21] is going to look at their client list, [13:23] they're going to look what cars have [13:25] they purchased. Like, we know multiple [13:28] Ferrari collectors that are buying two [13:30] separate cars that they don't even want [13:33] of the same style in order to gain [13:35] status within the dealership and show [13:37] that they're a proper proper collector [13:40] and and they're properly sort of [13:41] cherishing the brand. [13:42] >> Yeah. [13:43] >> And so, I'm not at all surprised that [13:44] they've sold 500-ish units. That's about [13:48] as many [13:48] >> Mhm. [13:49] >> uh as I would have expected for 2026. I [13:52] actually I guess if you asked me I maybe [13:54] would have thought they would have done [13:55] more. Like, to me this was a car that [13:57] was so different than the rest of their [13:58] cars. It serves a wildly different use [14:01] case. I would have expected their sales [14:03] targets to be quite a bit higher simply [14:05] because when it comes to their really [14:07] special cars, they make about 500 of [14:09] them. [14:09] >> Yep. [14:10] >> And so, I would have expected at least [14:12] 500 sales guaranteed and then you would [14:15] hope there was a bunch of incremental [14:17] buyers, people that are like, "Yeah, I [14:18] actually don't want a Ferrari sports [14:20] car, but I do want a daily and why not [14:23] go for a Ferrari daily, right?" So, you [14:26] would have thought that there would have [14:27] been like [14:28] 500 for the first year was like my very [14:31] base case [14:32] and I would have expected a bunch more [14:34] on top of that. So, I think they're [14:36] positioning this as a win. I think [14:38] people are going to love the car if you [14:40] ignore the price, but I don't think it's [14:42] the win that [14:43] >> The Pope [14:44] in the blue jacket looks good. [14:47] Okay, three points in response to yours [14:50] two. First point, the design's [14:52] absolutely growing on me. Like, watching [14:53] these videos here, it just looks way [14:56] better than when we first saw the the [14:58] first pictures and I don't know if it's [15:00] just distance and I'm becoming more [15:02] familiarized with it, but it looks a lot [15:05] better. Even the exterior. I've always [15:06] agreed on the interior. I think everyone [15:08] agrees on that, but the exterior is [15:09] looking better to me somehow. I don't [15:11] know if this is just like I'm getting [15:13] used to it. [15:14] Two, [15:16] yes, there are there are the 4D chess [15:19] Ferrari collectors who are saying, "I [15:21] know I'm not getting pressured, but I'm [15:23] buying one anyway because I think it'll [15:24] help me jump the line and I'm doing that [15:26] independently of any pressure that's [15:27] coming. Uh but there's also just [15:29] collectors that are like, this is going [15:31] to be a piece of Ferrari history [15:32] regardless of if they What if they never [15:35] make another EV again? What if the What [15:36] if the Purosangue is canceled next year [15:39] and Ferrari literally for 30 years never [15:41] makes an EV? This thing is it's [15:43] important historically. It's an [15:45] interesting thing to have in your [15:46] collection. And then there's also just [15:48] people that are like, I want I I'm a [15:51] true collector. I want every possible [15:53] Ferrari experience. Give me the SUV cuz [15:55] I want to see what that's like. Give me [15:56] the [15:56] uh the the mid-engine. Give me the front [15:58] engine. Give me the the the the [16:00] electric. Give me everything. Give me a [16:01] vintage. Give me a new one. Give me a [16:04] a road car. Give me a track car. Give me [16:06] I want I want a Ferrari F1 car. I want [16:09] all of the experiences because I just [16:11] want to experience everything Ferrari [16:13] cuz I'm that deep with the brand. And [16:14] then lastly, the question of, you know, [16:17] what was their goal? I don't see this as [16:20] their Urus. I don't see this as they [16:23] were trying to make a mass-market daily. [16:26] I think that they were trying to make a [16:27] very iconic, very iconoclastic, very uh [16:31] contrarian car that was bold and weird [16:35] and different and it happened to be [16:37] daily-able. And the fact that it is [16:39] daily-able is what is weird about it. [16:42] Like they're not known The brand is not [16:44] known for being able to be daily-ed and [16:47] yet they made one and that makes it [16:48] weirder. And and I don't think that they [16:51] were going after their This is something [16:53] that will be like a Lamborghini Urus, [16:56] which they can sell in mass volume and [16:58] completely change the profile of their [17:01] of their sales curve. [17:02] >> My view is that they I think those are [17:04] all great points. My view is that I [17:06] don't think they should have done a car [17:08] like this because I do think it hurts [17:09] the brand unless it was going to drive [17:11] so many sales that it could make the [17:14] rest of the cars that they make better, [17:16] right? Like what the Cayenne did for [17:18] Porsche, and what the Urus does for [17:19] Lamborghini. [17:20] >> I have one last post I want to go [17:21] through. Jordy, have you seen [17:24] Spider-Man: No Way Home? [17:25] >> Absolutely [17:27] no. Absolutely not. Should I? [17:29] >> I think I have. [17:30] >> into movies now. I saw The Odyssey, and [17:33] I appreciate film. [17:34] >> You're a film buff now. You're film [17:36] snob. [17:36] >> you whatever Yeah. [17:37] >> Honestly. [17:38] Uh no. Uh there was a question on the [17:41] timeline from Rob Felt rewatching He was [17:43] rewatching Spider-Man: No Way Home to [17:46] prep for Brand New Day, the new [17:48] Spider-Man movie. In Spider-Man: No Way [17:51] Home [17:52] the movie, uh the prequel to Brand New [17:54] Day, I think uh [17:56] uh Homecoming is in in this in this [17:58] series, right? There's a whole series of [18:00] new the latest round of Spider-Man with [18:02] Tom Holland, right? Is that it? Uh [18:05] Daily Bugle web show scene pops up, and [18:08] all I can think about now is a question. [18:11] Is the TBPN aesthetic inspired by J. [18:14] Jonah Jameson's web show? And if you [18:17] look at it, it does sort of look like [18:19] our show. And so it's a good question. [18:21] Did we see this film? Did Jordy, who is [18:24] the brand architect of the TBPN [18:26] aesthetic, watch Spider-Man: No Way Home [18:30] and say, "Ah, I like that color. I like [18:32] that design. Let's bring that into our [18:33] studio." And the answer? [18:36] No. [18:36] >> No. [18:37] >> No. Lots of other influences, but this [18:39] was actually not one of them. I I [18:41] believe I have seen this I like Yeah, we [18:43] like the color green. I remember Jordy [18:45] one morning uh we were working out, and [18:48] he's like, "We should do green." And I'm [18:50] like, "Okay, yeah, that sounds good. I [18:52] like green." And he's like, "No one's [18:54] done green." [18:56] And I'm like, "That's not true. Like [18:58] Robin Hood is green. There's plenty." [19:00] He's like, "No one in tech No one in [19:02] tech has ever used green before." [19:04] >> green. [19:04] >> I'm like, [19:06] "Uh it is a white space in the sense [19:07] that like yeah, I couldn't think of [19:09] another podcast with the green dark [19:11] green background." Uh and we And we did [19:13] find our own space. Uh We We We looked [19:15] at Pinterest a lot for different [19:17] references, some photos, some catalogs, [19:19] Ralph Lauren. [19:20] >> look at Pinterest. [19:20] >> What What What What What What images [19:22] were you pulling from? Cuz I know you [19:24] had some references. I mean, obviously [19:26] F1, but you just use your brain? [19:28] Just your brain? You don't let much go [19:29] in there, but [laughter] [19:32] >> Certainly not movies. [19:33] >> Certainly not movies. [19:35] Sign up for our newsletter at tbpn.com [19:37] and we will see you tomorrow. Goodbye.