Rocket Lab Buys Iridium for $8 Billion to Challenge SpaceX Starlink

Rocket Lab斥资80亿美元收购Iridium挑战SpaceX星链 Rocket Lab利用其600亿美元估值收购Iridium现役卫星网络及频谱访问权，绕过多年的星座研发周期，打造直面SpaceX竞争的垂直整合竞争对手。 本期节目涵盖多项重大科技与商业动态，首先探讨Rocket Lab斥资80亿美元收购卫星运营商Iridium以推进垂直整合，向SpaceX发起挑战。主持人随后讨论了Comcast计划从NBCUniversal进行结构性分拆、Blue Origin发射台事故后的加速复飞方案，以及预测市场的监管税务现实与假设性AI避税方案的对比。最后，节目回顾了《华尔街日报》读者丰富的科技愿望清单，并分析了Derek Thompson关于AI工具如何助推高利润“超级个体”（solopreneur）迈入黄金时代的相关文章。 Rocket Lab 宣布达成以 80 亿美元现金加股票收购卫星运营商 Iridium Communications 的协议，通过扩大垂直整合直接挑战 SpaceX 的 Starlink 卫星网络。 (0:02) Comcast 正在彻底逆转其长期战略，计划推行结构性分拆，将核心网络连接业务与 NBCUniversal 的媒体资产剥离为两家独立公司。 (6:37) 针对有提议利用 AI Agent 和 C 类公司降低高收入医生税率的避税方案，分析指出根据美国国内税收法典第 61 条，个人服务报酬必须计入总收入，该方案从法律层面直接失效。 (9:14) Meta 正在评估涉足预测市场的可行性，但外界对于其会提供真钱投注还是仅将预测数据整合进 Instagram 及 LLaMA 模型仍存分歧。 (13:22) 《华尔街日报》读者科技心愿单引发讨论，涵盖太阳能汽车、微型核聚变反应堆以及车际通信等领域；主播同时提出激进预测，认为自动驾驶技术将如同大语言模型一样迅速在各大车企间大宗商品化。 (16:51)

TBPN reviews major shifts across the tech and commercial landscape, highlighting Rocket Lab's acquisition of Iridium to compete directly with SpaceX, the unraveling of telecom-media mergers, and the emergence of an AI-driven boom in solo entrepreneurship. Rocket Lab's Push Toward Full Space Integration (0:02) Rocket Lab agreed to acquire satellite operator Iridium Communications in an $8 billion cash-and-stock deal, paying a 20% premium over Iridium’s prior closing price. The transaction brings Iridium's operational fleet of 66 low Earth orbit satellites under Rocket Lab, advancing the company's ambition to shift from a small-satellite launch provider into an integrated space prime alongside its Electron launch operations and the development of its reusable Neutron rocket. Media and Telecom Strategy Unwinding (6:37) Comcast plans to structurally separate its core connectivity infrastructure from NBCUniversal, reversing a decade-long strategy that began with purchasing a 51% controlling stake in NBCUniversal in 2011 and acquiring the remaining 49% in 2013. The planned split aligns with earlier retreats across the telecom sector, including Verizon selling Yahoo and AOL to Apollo in 2021 and AT&T spinning off WarnerMedia in 2022 after finding limited synergies between network distribution and entertainment production. Prediction Markets and Financial Structuring (13:22) Debate over Meta's potential foray into prediction markets centers on whether the platform intends to facilitate real-money transactions or restrict use to data partnerships, AI model feeds, and user engagement. Media and financial entities such as CNN and the NYSE have previously integrated prediction market data without operating wagering venues. Autonomous Systems and Consumer Tech Wishlists (16:51) A review of reader submissions to The Wall Street Journal highlights consumer demand for automation, ranging from inter-vehicle communication standards to eliminate accordion traffic delays to compact fusion micro-reactors and domestic cooking robotics. While surface ideas like solar panels on car roofs face fundamental energy-density limits, autonomous vehicle capabilities are predicted by the hosts to commoditize rapidly across manufacturers rather than serving as a lasting proprietary moat. The Rise of AI-Enabled Solo Entrepreneurs (29:10) Analysis by journalist Derek Thompson outlines an emerging cohort of solopreneurs leveraging AI software, consulting, and design tools to scale multi-million dollar businesses faster than prior cohorts recorded by payment processors. Thompson contrasts this data-backed trend against both technological alarmism and dismissive claims, while cautioning that a widespread shift from W-2 employment toward pass-through business entities may compress federal income tax revenues. The broader economic effects of micro-startups and changing corporate structures remain unsettled, particularly whether tax bases and shared commercial infrastructure can adapt to decentralized, software-augmented work.

本期节目重点梳理了商业航天与电信行业的资产重组逻辑，并探讨了人工智能如何重塑个人创业经济与税务、衍生品监管及消费科技的未来预期。 商业航天整合与电信分立 (0:02) Rocket Lab 宣布以 80 亿美元的现金加股票交易收购老牌卫星运营商 Iridium Communications，溢价达 20%。Iridium 拥有由 66 颗在轨近地轨道卫星组成的机队，长期为航海、采矿及美国政府机构提供通信服务。Rocket Lab 自 2021 年 8 月通过 SPAC 上市（当时估值 41 亿美元，每股 10 美元）后，其股价从 2024 年 7 月的 4.80 美元上涨至 11 美元，总市值达 600 亿美元。CEO Peter Beck 指出，真正的商业价值在于空间应用，这需要发射权、卫星制造规模和稀缺无线电频谱的结合；此次收购旨在快速建立垂直整合能力，直接在轨道网络领域对抗拥有约 10,000 颗卫星的 SpaceX Starlink。 与航天领域推进纵向一体化相反，传统电信巨头正加速剥离媒体资产。Comcast 宣布计划分拆为两家独立公司：一家专注于宽带和网络连接基础业务，另一家由 NBCUniversal 组成独立内容公司。这一决定逆转了其 2011 年以 51% 控股权起步、2013 年全资收购 NBCUniversal 的十年战略。类似的分拆在电信业屡见不鲜，如 Verizon 在 2021 年向 Apollo 出售包括 AOL 和 Yahoo 的媒体部门，AT&T 则于 2022 年分拆 Warner Media 并退出 DirecTV，反映出内容生产与管道分发之间的长期协同效应未达预期。 发射场快速恢复与自动驾驶普及预期 (8:13) 在发射与运载工具方面，Blue Origin 在 New Glenn 火箭遭遇发射台爆炸事故后调整了恢复策略。CEO Dave Limp 表示，公司计划在一年内重返飞行，推翻了外界关于发射台需重建两年的传言。Blue Origin 并未原样修复旧设施，而是加速将其原为 9x4 New Glenn 开发的水平/垂直混合操作概念提前落地，利用现有基础设施绕过专用的运输起竖机，变相压缩了技术升级周期。 在地面智能硬件方面，节目主播在体验 Rivian 演示车后提出，自动驾驶技术的商品化速度正比肩大型语言模型。主播认为，完全自动驾驶技术正在迅速转变为各家汽车制造商的基线通用功能，未来不会长期作为个别龙头企业的独占护城河。此外，对于《华尔街日报》读者提出的车对车通信防拥堵构想，主播指出该机制在 20 年内具备高度可行性，其核心在于全行业能否建立统一的数据通信协议以消除人为刹车造成的连锁停顿。 税法界限与预测市场衍生品属性 (9:14) 节目通过讨论针对高收入群体的避税构想，解析了美国税法的刚性约束。针对利用 AI 批量设立特拉华州 C 类公司、将医疗挂号费包装成公司资产以图适用长期资本利得税率的方案，ChatGPT 的合规检测将其定性为非法的滥用型避税安排。根据美国《国内税收法典》（IRC）第 61 条，所有个人服务报酬及服务费用均属于总收入，并按普通收入税率征税，无论外部如何包裹公司法人外壳均无法改变其收入性质。 关于预测市场的商业模式与监管定位，节目讨论了 Meta 可能与现有市场供应商展开的合作。分析指出，此类合作可能仅限于数据授权以接入 LLaMA 5 或搜索结果，而非直接在社交软件内搭建受监管的金融支付闭环。从监管层面看，商品期货交易委员会（CFTC）已将特定事件合约定义为以未来结果为标的的衍生品，并给予了 Kalshi 合约市场地位。若此类合约最终明确适用税法第 1256 条，交易者将通过年度盯市会计制度享受 60% 长期资本利得和 40% 短期资本利得的混合优惠税率，这对预测市场的金融合法化至关重要。 AI 超级个体崛起与宏观外溢效应 (29:10) 记者 Derek Thompson 的分析指出，当前关于人工智能的讨论被两极化掩盖：一方是无视充分就业现状、宣扬所有人失业的末日论者，另一方则是全盘否定 AI 提效价值的怀疑论者。实际经济数据显示，依托 AI 编写代码、进行设计与咨询的单人微型初创企业正迎来黄金时代，这类独立创业者在 Stripe 等支付平台上跨入数百万美元收入门槛的速度已超越以往任何群体。 然而，这种向“一人公司”转变的趋势伴随着宏观隐忧与实体商业的局限。Thompson 指出，大量全职 W-2 雇员转型为穿透税收实体（pass-through entities），可能会持续侵蚀联邦政府的传统个人所得税税基，给财政收入带来摩擦。与此同时，主播认为个人创业潮并不会提振联合办公空间行业，因为这类商业模式始终面临个人对专属私密空间的需求瓶颈，极限制约了共享办公形态的长期经济效益。 民间前沿科技诉求的工程现实检验 (16:51) 针对《华尔街日报》读者收集的技术愿望清单，节目对一系列概念进行了工程可行性辨析。针对车顶集成太阳能板（Solar on Wheels）的想法，讨论指出当前太阳能电池的光电转换能量密度极低，普通车顶面板面积不足以在短时间内提供实用级续航，目前仅限于科研实验或军方特种挑战项目。而针对通过表面接触测温的“智能煎铲”，工程分析指出在外部高温焦糊层阻隔下，非侵入式读取牛排内部核心温度在物理上极难实现，除非在餐具内集成工业级 X 光或核磁扫描装置。 相比之下，部分前沿与消费硬件构想则具备现实创业基础。西盛顿州核聚变初创企业 Avalanche 研发的电池尺寸紧凑型微型聚变反应堆，展现了未来无需频繁补充燃料即可驱动特种航空器与航天器的原子朋克前景；而具备采购、整理食材并完成烹饪全流程的家用双足机器人厨师，被视为 20 年内具备落地可行性的软硬件研发方向。此外，针对消解噪音的艺术化育儿隔音板以及将研发资源倾斜至癌症与抗衰老药物的诉求，反映出市场对提升长期生产力与生活质量的核心关切。 本期内容反映出各技术领域正在回归理性审视：航天与流媒体通过并购或拆分纠偏资产布局，AI 落地加速了微型企业的诞生却也对税制产生冲击，而诸多消费级技术构想则面临基础物理规律与监管制度的双重校准。

## Background 该视频由科技脱口秀频道 TBPN 于 2026 年 7 月 23 日发布在 YouTube 平台，为旗下 30 分钟精简版节目《Diet TBPN》。根据官方简介，节目由 John Coogan 与 Jordi Hays 主持，本期重点梳理了 Rocket Lab 正面硬刚 SpaceX 的重磅收购、Meta 涉足预测市场的传闻以及全美科技愿望清单等前沿科技动态。 ## TL;DR 节目核心聚焦于 Rocket Lab 斥资 80 亿美元收购 Iridium，实现垂直整合以挑战 SpaceX 的战略布局，并回顾了 Comcast 计划拆分 NBCUniversal 带来的电信与内容大退潮。随后，主持人们以极客脑洞探讨了借助 Agentic AI 与预测市场为医生避税的荒诞构想，并分析了 Meta 试水预测市场的深层意图。紧接着，节目盘点了《华尔街日报》读者征集中的各类未来科技愿景，涵盖车车协同防堵车、弱视音乐家读谱辅助、全自动机器人主厨及婴儿房隔音方案。最后，主播结合德里克·汤普森的观点，深入剖析了 AI 时代“单人超级个体”与高营收微型企业的爆发式增长。 ## Atmosphere & style 整期节目节奏明快、谈笑风生，充斥着典型硅谷科技圈的极客幽默与机智讽刺。两位主持人配合默契且互动频繁，能在严肃的数十亿美元硬科技并购分析与荒诞不经的脑洞点子（如把看病包装成收购特拉华空壳公司）之间无缝切换。收听体验既有硬核商业洞察的干货快感，又宛如置身一场好友间肆意调侃、金句频出的科技茶话会，充满戏剧张力与感染力。 ## Bulletin - [00:02](https://www.youtube.com/watch?v=uQawEBikW8U&t=2s) Rocket Lab 宣布以 80 亿美元现金加股票收购老牌卫星运营商 Iridium，溢价约 20%，借助后者在轨星座与稀缺频谱资源全面挺进 SpaceX 的腹地并实现太空应用业务闭环。 - [06:36](https://www.youtube.com/watch?v=uQawEBikW8U&t=396s) Comcast 计划拆分 NBCUniversal，结束长达十余年的“内容+网络分发”整合实验，再次印证传统电信巨头跨界掌控好莱坞媒体的协同模式走向瓦解。 - [29:09](https://www.youtube.com/watch?v=uQawEBikW8U&t=1749s) 德里克·汤普森与 Stripe 数据揭示，AI 杠杆正在引爆“高营收微型企业”浪潮，单人团队达到数百万美元营收的速度创下历史新高，但同时也带来个体孤独感加剧与税制冲击等隐忧。 - [13:19](https://www.youtube.com/watch?v=uQawEBikW8U&t=799s) 针对 Meta 涉足预测市场的传闻，分析认为其更有可能仅限于内容合作、搜索数据授权或广告引流，而非自行搭建真钱衍生品清算与金融交易平台。 ## Key takeaways - [00:02](https://www.youtube.com/watch?v=uQawEBikW8U&t=2s) Rocket Lab 斥资 80 亿美元收购 Iridium Communications，单笔交易金额为其 2021 年 SPAC 上市估值（41 亿美元）的两倍。 - [02:59](https://www.youtube.com/watch?v=uQawEBikW8U&t=179s) Peter Beck 提出“太空应用方程式”，指出航天核心价值在终端应用，收购 Iridium 能直接绕过耗费数年的频谱申请、卫星发射与获客周期，快速获得正向现金流。 - [06:36](https://www.youtube.com/watch?v=uQawEBikW8U&t=396s) Comcast 拟分拆 NBCUniversal，与此前 Verizon 折价甩卖雅虎/AOL、AT&T 剥离 DirecTV 和时代华纳如出一辙，标志着电信巨头内容并购整合路线的全面退潮。 - [08:14](https://www.youtube.com/watch?v=uQawEBikW8U&t=494s) 蓝色起源 New Glenn 发射台爆炸后重整思路，采用水平-垂直混合运行概念统一两个工位标准，预计年内重返飞行且无需重新制造起竖车。 - [09:57](https://www.youtube.com/watch?v=uQawEBikW8U&t=597s) 借助 Agentic AI 批量注册特拉华 C 公司将医生就诊包装为股权收购以规避普通所得税的构想，在税法（IRC 61）实质课税原则下被认定为违规的滥用型避税工具。 - [12:12](https://www.youtube.com/watch?v=uQawEBikW8U&t=732s) 若预测市场事件合约被监管机构归入《国内税收法典》第 1256 条合约，其收益可适用 60/40 资本利得优惠混合税率，并带来用衍生品对赌重构日常服务交付的极客想象。 - [13:19](https://www.youtube.com/watch?v=uQawEBikW8U&t=799s) Meta 因长期未能打通原生金融与支付闭环，且面临严格金融监管红线，直接下场运营真实货币预测交易的可能性极低。 - [19:29](https://www.youtube.com/watch?v=uQawEBikW8U&t=1169s) 大众对未来 20 年科技创新的诉求展现出强烈的平民视角，涵盖车车通信协同消除交通“手风琴效应”、弱视乐手专用乐谱滚动设备等实际痛点。 - [23:34](https://www.youtube.com/watch?v=uQawEBikW8U&t=1414s) 居家降噪与育儿隔音具备明确的商业转化潜力，例如将专业吸音板定制为美观的手绘儿童房装饰画布以满足家庭美学与隔音的双重诉求。 - [27:38](https://www.youtube.com/watch?v=uQawEBikW8U&t=1658s) 自动驾驶技术正迅速走向大宗商品化（Commoditizing），其扩散逻辑与大语言模型类似，未来各家车企均能普及，而非长期被少数几家寡头垄断。 - [29:09](https://www.youtube.com/watch?v=uQawEBikW8U&t=1749s) AI 正在推动“超级个体”与单人初创企业成为高产出经济主体，当前 AI 与就业的讨论在很大程度上陷入了失业末日论与无用虚无主义的两极偏见。 - [31:11](https://www.youtube.com/watch?v=uQawEBikW8U&t=1871s) 即便单人创业者激增，共享办公空间仍难言真正利好，因为知识工作者最终追求的依然是具备私密性与专属土地/空间的独立环境。 ## Detailed walkthrough ### [00:02](https://www.youtube.com/watch?v=uQawEBikW8U&t=2s) Rocket Lab 斥资 80 亿美元收购 Iridium，正面硬刚 SpaceX 节目开场谈到了科技圈热议的话题：Rocket Lab 正式杀入 SpaceX 的腹地。TBPN 的 Brandon Guerell 在日常通讯中对此做了详细梳理。发射服务与太空系统提供商 Rocket Lab 达成协议，将以现金加股票的方式收购老牌卫星制造与运营商 Iridium Communications，交易对 Iridium 的估值高达 80 亿美元，相较其上周五收盘价溢价约 20%。Iridium 作为近地轨道（LEO）卫星通信的开山鼻祖，早在 30 年前就发射了首颗卫星，当年其标志性的厚重“卫星电话”不仅天线巨大，还经常被大家拿来开玩笑，称“那天线要是离脑门太近感觉都能被辐射干掉”。如今 Iridium 运营着一支由 66 颗较大型老款卫星组成的网络，专门为远洋船舶、矿区、美国政府机构以及各类企业级客户提供关键连接。 这一收购标志着 Rocket Lab 垂直整合战略的重大飞跃。多年来，Rocket Lab 已经从一家小众的轻型卫星发射商，蜕变为集卫星设计、航天核心零部件制造、服务主流国防承包商于一身的太空综合巨头。其主力火箭 Electron 已成为美国年发射量仅次于 SpaceX 的二号运载工具，同时他们还在全力推进可复用中型运载火箭 Neutron 的研发。尽管 Jeff Bezos 的蓝色起源随时可能来争夺这枚“银牌”，但通过吞并 Iridium，Rocket Lab 摆明了要正面抗衡拥有上万颗 Starlink 卫星的 SpaceX。 主播回顾了 Rocket Lab 堪称惊艳的资本市场走势：2021 年 8 月 25 日，Rocket Lab 通过 SPAC 以 41 亿美元估值（约每股 10 美元）登陆资本市场，随后整整三年股价一直在发行价下方徘徊。知名投资人 Delian Asparouhov 曾提及，2024 年 7 月 1 日其股价曾低至 4.80 美元左右。直到 2024 年 9 月，随着营收快速兑现，市场对其进行了全面重估（re-rating）。如今这家市值已冲至 600 亿美元的太空新贵，股价回升至 11 美元上方，更有底气直接吃下价值 80 亿美元的 Iridium，完成规模翻倍的大并购。 - 交易架构：Rocket Lab 以 80 亿美元现金加股票收购 Iridium Communications，溢价 20%。 - Iridium 现有资产：拥有 66 颗在轨运行的低轨通信卫星，专注海事、采矿、国防及企业级市场。 - 历史估值对比：Rocket Lab 2021 年通过 SPAC 上市估值为 41 亿美元，低迷期股价曾下探至约 4.80 美元，本次单笔收购规模已达其最初 IPO 估值的两倍。 --- ### [02:59](https://www.youtube.com/watch?v=uQawEBikW8U&t=179s) Peter Beck 的“西装梗”与太空应用方程式 ```mermaid flowchart TD accTitle: 太空应用方程式与收购捷径 accDescr: 展示 Rocket Lab 原有能力、太空业务高门槛以及通过收购 Iridium 实现业务闭环的路径。 subgraph RocketLab["Rocket Lab 自有核心能力"] A["自主发射能力<br/>(Electron / Neutron)"] B["规模化航天器制造<br/>(Space Systems)"] end subgraph Barriers["商业化三大高壁垒"] C["极度稀缺的频谱资源<br/>(Spectrum)"] D["漫长的基建与发射周期<br/>(Time to First Dollar)"] E["漫长的获客与现金流建立<br/>(Sustained Cash Flow)"] end subgraph Solution["收购 Iridium 的捷径效果"] F["现成稀缺频谱 + 运营网络"] G["成熟盈利模式 + 数百万客户"] H["1 + 1 = 3: 垂直整合的太空应用超级巨头"] end RocketLab --> Barriers Barriers -.->|"耗时过长，寻求快捷路径"| Solution F & G --> H ``` 随后节目切入了 Rocket Lab 创始人兼 CEO Peter Beck 的官宣视频片段。视频一开场就充满幽默戏剧性，Peter Beck 穿着一件略显紧绷的深蓝西服出现在白板前，调侃道："上一次我穿这件蓝夹克还是在宣布 Neutron 火箭的时候，所以我想这代表今天又有大事要宣布了。" 现场主播忍不住大笑，吐槽他套上西装时动作僵硬，最后才揭晓伏笔——原来是因为西装买小了，正片结束还不忘补一句“我得去买件新夹克了”。 在这场即兴“白板教学”中，Peter Beck 阐述了他推导的“太空应用方程式”（The Space Application Equation）。他指出，航天产业最核心的价值并非发射本身，而是在于终端应用（Applications）。要真正吃下这块蛋糕，通常需要跨越重重难关： 1. 必须拥有不受限制的自主进入太空通道（自主发射能力）； 2. 具备规模化建造航天器的工程能力； 3. 获得极度稀缺、几乎不可能申请到的无线电频谱（Spectrum）； 4. 熬过卫星研制、测试、排期发射直到赚得第一块钱的漫长研发周期； 5. 熬过建立客户群、实现正向现金流并获取稳定订阅收入的漫长商业周期。 Peter Beck 直言自己是个“极度缺乏耐心的人”，不愿再按部就班熬过长达数年的折磨，而收购 Iridium 就是他们找到的捷径。这笔交易堪称“1 + 1 = 3”：Rocket Lab 拥有强大的发射与卫星制造能力，而 Iridium 拥有现成在轨的卫星星座、全球极度抢手的黄金通信频谱、数百万稳定付费客户以及成熟健康的盈利现金流。两相结合，使 Rocket Lab 瞬间蜕变为全产业链一体化运营的太空超级巨头，不仅能继续运营现有网络，更能基于新网络开拓全新太空服务。 - 核心收购逻辑：跳过长达数年的频谱申请、组网发射及商业获客周期，快速获得稳定正向现金流。 - 视频结尾彩蛋：Peter Beck 解释自己穿西装之所以笨手笨脚，是因为这件衣服实在太小了，录完得赶紧去换一件。 --- ### [06:36](https://www.youtube.com/watch?v=uQawEBikW8U&t=396s) 时代退潮：Comcast 计划拆分 NBCUniversal 紧接着，主播转向了传媒与电信行业的重磅重组。《华尔街日报》披露，Comcast 历时十余年探索的“内容 + 渠道分发”大一统商业实验正走向终点，公司计划将业务一分为二：Comcast 母体重新回归纯粹的电信宽带与网络连接业务，而 NBCUniversal 则分拆为独立实体，全力押注内容娱乐。 回顾这段并购史：2011 年，拥有约 3000 万美国家庭宽带与电视用户的有线巨头 Comcast，先是从通用电气（GE）手中买下了 NBCUniversal 51% 的控股权。主播调侃道，自己向来着迷于这种“重工业公司跨界掌控好莱坞媒体”的荒诞戏码。2013 年，Comcast 收购了剩余的 49% 股权实现全资控股，随后大举开疆拓土——收购欧洲广播巨头 Sky、动画巨头梦工厂（DreamWorks Animation），并在流媒体大战中孵化了 Peacock。然而如今，电信网络与娱乐内容的协同神话破灭，只能各奔前程。 主播指出，过去十年间电信巨头“疯狂并购买内容、最终惨淡剥离”的循环屡见不鲜： - **Verizon**：曾耗费重金吞并 AOL 与雅虎（Yahoo），将其重组为 Verizon Media，最终在 2021 年不得不打折转手甩卖给私募巨头 Apollo； - **AT&T**：先后斥巨资接盘卫星电视服务商 DirecTV 与传媒航母时代华纳（Time Warner），结果债务缠身，先是彻底清仓割肉 DirecTV，又在 2022 年将时代华纳拆分出逃，与 Discovery 合并为 Warner Bros. Discovery。 --- ### [08:14](https://www.youtube.com/watch?v=uQawEBikW8U&t=494s) 蓝色起源发射台修复进展，与医生的“避税狂想” 太空领域的另一则利好来自蓝色起源（Blue Origin）。在此前 New Glenn 火箭遭遇令人沮丧的发射台爆炸事故后，外界曾悲观盛传该发射台至少需要两年时间才能修复重建。但蓝色起源 CEO Dave Limp 带来了最新转机：团队今年内就将重返飞行。他们没有选择原地原样复建老旧发射台，而是借机直接采用为 9x4 New Glenn 火箭研发的“水平-垂直混合运行概念”（horizontal-vertical hybrid conops）。这一方案复用了现有基础设施，跳过了全新起竖运输车（transporter erector）的漫长制造，并在两个工位之间实现了统一的标准化运作。主播打趣称：“他们这根本不是出了事故，纯粹是拿爆炸当旧设施的拆迁工程（demo）用了。” 在节目的最后，主播抛出了一个无厘头却又极具痛点的商业点子：“当医生最大的问题是什么？是熬夜倒班吗？不，是税！”主播分析道，顶尖外科医生即便靠救死扶伤赚取高薪，其收入在税法上仍被归类为普通所得（ordinary income），面临极高的边际税率。主播提到自己身边甚至有位医生朋友在筹备新诊所的空档期跑去操盘了一家对冲基金，只为了能享受低税率的资本利得（capital gains）。另一位主播随即吐槽表示这听起来极其吓人：“如果知道负责给我主刀做手术的医生私下里还在兼职跑去搞对冲基金，我躺在手术台上恐怕会非常担忧。” ### [09:57](https://www.youtube.com/watch?v=uQawEBikW8U&t=597s) AI 智能体避税狂想：把看病包装成收购特拉华 C 公司 如果一个外科医生在给病人做手术的同时还在兼职管理对冲基金，任何躺在手术台上的人恐怕都会感到毛骨悚然。但从现实角度来看，医生们挽救生命、收入丰厚，却必须按普通所得税率（Ordinary Income）缴纳高额税款，而无法享受更优惠的资本利得税率（Capital Gains）。针对这一“痛点”，节目抛出了一个脑洞大开的创业方案：借助 Agentic AI（智能体 AI）的力量，为每位医生预先批量注册成千上万家特拉华州 C 类公司（Delaware C Corp）。 按照这套荒诞的构想，每次病人预约就医需要付款时，并不直接向医生支付诊金，而是收购医生名下的一家特拉华 C 公司。该公司资产负债表上唯一的资产，就是一张医生必须为持股人提供医疗服务的欠条（IOU）。由于医生早已持有这批公司超过一年，股份套现后便能按长期资本利得申报税款；而患者在离开诊所时只需签署一份文件，由 AI 智能体自动执行公司清算并变现资产，甚至还能将该医疗服务资产进行折旧抵税。 面对这套号称“天衣无缝”的商业模式，搭档当场吐槽这纯粹是用垂直领域 AI 智能体在搞“高科技逃税”。连 ChatGPT 对此方案给出的裁决也是“一出生就注定夭折”：这根本不是什么精妙的税务筹划，而是附带了医疗法规连带责任的滥用型避税工具（Abusive Tax Shelter）。根据美国《国内税收法典》第 61 条（IRC 61），任何为个人劳务支付的报酬与费用在本质上都属于总收入，仅仅用刚出炉的空壳 C 公司对账单进行一层包装，在法律实质上完全无法改变其劳务所得的性质。 ```mermaid flowchart TD accTitle: 荒谬的 AI 智能体看病避税闭环 accDescr: 展示将医疗服务包装为特拉华 C 类公司股权收购并试图获取长期资本利得的荒诞流程 subgraph Setup ["前期准备"] A["Agentic AI 批量孵化"] --> B["医生名下持有超 1 年的特拉华 C 公司池"] end subgraph Transaction ["就医交易"] B --> C["资产负债表载入：医生医疗服务欠条 (IOU)"] C --> D["患者看病时全资收购该 C 公司"] end subgraph Settlement ["退出与清算"] D --> E["患者签字：AI 自动清算公司并折旧资产"] E --> F["医生按长期资本利得申报税费 (被 IRS 否决)"] end ``` ### [12:12](https://www.youtube.com/watch?v=uQawEBikW8U&t=732s) 用预测市场重构一切交易：从医疗账单到“蓝莓外卖” 既然空壳公司方案走不通，两位主持人随即把脑筋动到了预测市场上：能否把看病过程包装成一场对赌？例如每次就诊时，患者开设一个盘口，以接近 0% 的极低概率下注“医生这次治不好我”，而医生则押注“自己一定能治好”。如果治疗顺利完成，医生将赢得全部奖金并实现全额报酬；若发生医疗事故，患者则不承担费用，整个流程甚至还天然自带了医疗事故险。 这番玩笑引发了关于预测市场利润税务性质的严肃探讨。目前美国商品期货交易委员会（CFTC）监管的事件合约（如 Kalshi）在税法上的定性尚存争议。由于 CFTC 将其定义为收益取决于特定事件的衍生品合约，而非单纯的赌博，因此业界有相当强有力的论据认为其可归类为《国内税收法典》第 1256 条合约（Section 1256 Contracts）。这意味着无论持仓时间长短，盈亏都可能按年度盯市（Mark-to-Market），并享有 60% 长期资本利得与 40% 短期资本利得的混合优惠税率。 这种用金融衍生品重构日常生活的想法，精准呼应了去年科技圈流传的一篇经典热帖——“预测市场终将取代万事万物”。在这套极其极客的世界观里，现实生活中的服务交付可以直接转化为盘口博弈：如果你想吃蓝莓，只需在市场上建立一个“绝不会有人给我送蓝莓”的合约并下注 15 美元反方；另一边的外卖员为了赢走这 15 美元，就会立刻下注正方并迅速把一盒新鲜蓝莓送到你家门口。 - 预测市场盈利目前的税务分类尚未完全定论，但 Kalshi 作为指定合约市场（DCM），其事件合约极具被视为金融衍生品的潜力。 - 若被认定为第 1256 条合约，投资者将获得 60/40 资本利得税率分割的显著优势，无需满足一年的持有期门槛。 - 借由事件对赌完成实物交割（如买蓝莓），生动展现了金融化与科技极客对预测市场通用化的终极想象。 ### [13:19](https://www.youtube.com/watch?v=uQawEBikW8U&t=799s) Meta 会做真钱预测市场吗？社交影响力还是金融野心 关于 Meta 即将涉足预测市场的传闻，节目组内部展开了激烈交锋：扎克伯格究竟是想搞一个涉及真实真金白银的交易平台，还是仅用于社交影响力与声誉展示的游戏化功能？支持真钱交易的一方认为，Meta 已经主动接触了多家主流预测市场供应商，试图建立深度合作伙伴或供应商关系，而在该行业中，这类合作的先例通常是由外部平台提供底层清算与交易基础设施。 但反方观点认为，“建立合作”在科技巨头的语境下有完全不同的解读维度。Meta 与预测市场的合作，很可能仅仅停留在广告赞助、为预测市场提供折扣推广，或者购买数据授权。例如当用户在 Instagram 搜索框中提问时，底层的大语言模型（如 Llama）或搜索功能可以合法引用预测市场的实时概率数据，而不必承担侵权风险；亦或是将用户导流至第三方平台的拉新漏斗，Meta 绝无可能为了自己做竞品，却主动找竞争对手引流并拱手交出核心用户。 从历史维度审视，主流媒体巨头如 CNN、CNBC 以及纽约证券交易所（NYSE）都曾与预测市场建立过合作，但从未有任何一家要求用户直接在自家客户端内绑定信用卡去下注。更关键的是，Meta 二十年来屡次试图建立用户的原生金融账户体系（例如在 Instagram 或 Facebook 中保存支付信息实现一键购衣结账），但始终未能真正成气候。对于庞大如 Meta 这样的社交帝国而言，贸然跨入受严密监管的金融交易与博彩红线领域，在商业逻辑上依然充满了巨大的问号。 ### [16:51](https://www.youtube.com/watch?v=uQawEBikW8U&t=1011s) 《华尔街日报》20年科技愿望单：车顶太阳能与“原子朋克”核动力 节目最后将视线转向《华尔街日报》向读者发起的一项未来征集：你在未来 20 年内最希望看到什么科技创新？读者给出的答案五花八门。来自得克萨斯州欧文市的玛丽·吉莱斯皮（Mary Gillespie）提出了“移动太阳能”的设想，她渴望未来的汽车能在车顶全面集成高效太阳能电池板，无论行驶还是停泊时都能源源不断地为动力电池补能。 这一想法在概念上并不新鲜，现实中已有不少试验车尝试过车顶贴太阳能板，但由于目前光伏转换效率与能量密度的限制，寻常车顶微弱的功率可能需要连续晒上整整五个月才能充满一辆电动汽车。早年 DARPA 挑战赛中的特种车辆或是实验性太阳能无人机，必须依靠极其庞大展弦比的机翼表面积才能实现近乎无限续航的飞行。不过主持人表示，读者真正呼吁的是光伏能量密度的革命性跃升，在未来 20 年的科技长河中，这并非完全不可能实现。 顺着能源与出行的讨论，话题自然延伸到了更具科幻色彩的终极方案——核动力汽车。这是一种带有浓郁“原子朋克”（Atom Punk）美学、令人梦回冷战时期《辐射》设定的复古未来主义构想。虽然核裂变难以小型化上路，但可控核聚变却展现出了曙光。例如位于华盛顿州的小型聚变初创公司 Avalanche Energy，正在研发体积仅相当于大型蓄电池尺寸的小型聚变反应堆，未来有望直接塞入汽车、无人机甚至航天器中长期巡航。在此基础上，如果再辅以让各车自动驾驶系统直接互联互通的通信技术，不仅驾驶安全能大幅提升，令无数司机抓狂的“手风琴效应”（Accordion Traffic Jam，即幽灵堵车中车流如手风琴般一紧一松的拉扯现象）也将彻底成为历史遗迹。 ### [19:29](https://www.youtube.com/watch?v=uQawEBikW8U&t=1169s) 车车互联协同与视障音乐家的“乐谱视觉” 在探讨普通大众对未来科技的设想时，主持人们注意到人们的痛点往往比硅谷从业者想象的更接地气。来自阿拉巴马州伯明翰（Birmingham, Alabama）的摩根·克莱顿（Morgan Clayton）提出了一个针对堵车走走停停的解决方案：他希望所有汽车的自动驾驶系统能够直接互相通信、实现数据协同。在没有通信延迟的前提下，所有车辆能同步加速和减速，彻底消除因人类反应迟缓而导致的交通“手风琴效应”。主持人们认为这在技术上完全可行，核心难点在于需要跨车企统一标准。这种来自民间的真实需求让硅谷那套“将模型推理成本降到十分之一”的日常叙事显得有些缺乏生活想象力。 紧接着是来自犹他州阿尔卑斯（Alpine, Utah）的玛丽·吉福德（Mary Gifford）提出的“音乐视觉”（Musical Vision）设想。她呼吁发明一种能够帮助视力受损音乐家看谱的专用设备。玛丽提到，社会上有很多帮助视障人士阅读普通印刷文字的技术，但钢琴家等乐手需要在弹奏的同时实时阅读乐谱。她的父亲晚年视力极差，看谱时眼睛几乎要贴在谱架两英寸以内，只能每次死记硬背一两个乐句然后再练习，老人家甚至在94岁高龄时全凭记忆完成了人生最后一场独奏音乐会。 面对这个感人而极具挑战的诉求，主持人们展开了既敬佩又带着科技调侃的脑洞风暴： - **物理放大方案**：有人半开玩笑地提出，难道不能直接用卡车尺寸的打印机把乐谱印成海报大小，或者戴高度数眼镜？但问题在于严重弱视患者即便戴镜也必须近距离辨别。 - **动态数字乐谱**：利用放在琴架上的 iPad，开发专门的无障碍应用，一次只滚动显示一到两个特大音符，或者通过语音播报音符序列辅助记忆。 - **硬核具身智能**：调侃是否能造一个人形机器人，像操纵木偶一样直接抓住演奏者的手，在琴键上精准砸出旋律。 --- ### [22:03](https://www.youtube.com/watch?v=uQawEBikW8U&t=1323s) 智能锅铲与全自动机器人主厨 科技愿望清单迅速从艺术领域转向了厨房烹饪。保罗·布兰科（Paul Blanco）提出想要一把“智能锅铲”——它不需要像传统探针温度计那样刺入肉排内部，仅需平贴在汉堡排或牛排表面，就能隔空精准测出肉块内部的核心温度和熟度。主持人指出这在工程上难度极大，因为在高温煎烤（Searing）或反向慢烤（Reverse Sear）时，肉排表面与内部的核心温差极为悬殊。想要不刺穿肉质测温，除非锅铲里装了一台高能 X 射线机或像医疗脑部扫描仪、核磁共振（MRI）一样的庞大设备，现场脑补一把连着全身扫描仪的大号锅铲让人捧腹。 与锅铲的小切口相比，华盛顿州莱西（Lacey, Washington）的雷·卢尔（Ray Lure）则描绘了一个宏大的全自动主厨梦想。他希望未来的机器人能够成为自家厨房的“行政主厨”： - 能够自主致电生鲜超市并订购高端食材； - 协调配送上门，主动拆箱并分门别类码进冰箱； - 自主操作家中的现有厨电烹饪完整正餐。 主持人们认为这个设想处于极其精准的科幻与现实交界点——它绝非周末用一块 Arduino 开发板就能拼凑出的小玩具，但在未来20年内完全有可能实现。与其说是天马行空，不如说这是给硬件与具身智能创业者提供了一份清晰的“创业需求征集令”（Request for Startups）。 --- ### [23:34](https://www.youtube.com/watch?v=uQawEBikW8U&t=1414s) 婴儿房隔音痛点与户外主动降噪的奇思妙想 宾夕法尼亚州西格罗夫（West Grove, Pennsylvania）的雪莉·富兰克林（Cheryl Franklin）则聚焦于噪音污染对健康的侵害，指出噪音会导致高血压、心脏压力骤增并拉低房产价值，因此期待能自带隔音特性的新型石膏板或便携轻量隔音贴膜。这段诉求引发了主持人对“铅皮与重度质量隔音”的无厘头插科打诨（戏称“把含铅涂料带回生活”），但也引出了一个极具商业潜力的真实创业想法： 在照顾新生儿时，换尿布过程往往伴随着婴儿剧烈的啼哭与回音轰鸣。有朋友为了减震降噪，直接在婴儿房整面墙安装了播客录音棚同款的木质吸音格栅条板（Slat Wall），实现了“双重用途技术”。但绝大多数家庭并不希望婴儿房看起来像个专业播客录音棚，他们需要的是小动物、田园风光和温暖柔和的浅色调。因此，一个极佳的商业点子是：将吸音棉板制成平整画布，表面手绘温馨的儿童插画，既完美融入婴儿房装饰风格，又具备专业的吸音降噪功能，主持人们一致认定这绝对是一门能大卖的成熟生意。 随后，詹妮弗·史密斯（Jennifer Smith）提出了更宏大的愿望：她希望拥有“控制室外噪音的超能力”，能在自家纱窗门廊上惬意地听鸟鸣，同时完全屏蔽附近街道震耳欲聋、宛如直线加速赛车（Drag Racing）的交通噪音。针对“能不能造出户外开放空间的主动降噪音响，只抵消车流噪音却保留鸟叫”这一问题，主持人们展开了趣味讨论： - **电动汽车普及**：纯电车没有内燃机轰鸣，是解决街道噪音最现实的方案（除非有人恶意改装排气直通管甚至给 Model 3 换装燃油引擎）。 - **物理防御与美式幽默**：如果无法实现开放声场主动降噪，要么把整栋房子加装超厚隔音窗、把鸟养在屋里；要么开启《兔八哥》里的“歪心狼”（Wile E. Coyote）模式——直接在马路上画一堵假砖墙隧道，让飙车党直接撞上去，或者从天上掉下个铁砧砸扁噪音车。 --- ### [27:38](https://www.youtube.com/watch?v=uQawEBikW8U&t=1658s) Rivian 试驾体验、自动驾驶大众化与科技愿望清单的温差 话题自然转到了电动车体验上。Rivian 团队最近寄送了一台测试样车供他们体验，厂方的态度非常开放且硬核，甚至直接告诉他们“想开多快就开多快，就算把它开翻车（Roll it）也无所谓，尽情去体验它的极限”。主持人们对这款车的性能赞不绝口，尤其是其辅助驾驶功能被严重低估。 围绕自动驾驶，节目给出了一个极为鲜明的核心观点： - **自动驾驶正在迅速大宗商品化（Commoditizing）**：其普及速度将与大语言模型（LLM）相当。只要车企认真对待并投入资源，自动驾驶并非不可逾越的独家技术护城河。 - **全行业普及浪潮**：就像现在连连锁快餐店 Chipotle 都能用上客服 AI 聊天机器人一样，未来几乎每一辆车都会具备自动驾驶能力，这项技术绝不会永远被一两家先跑出来的寡头所垄断。 本段最后聚焦于《华尔街日报》这份科技心愿征集呈现出的巨大反差幽默。当大家翻阅这份愿望清单时会发现，有人只是想要一台“稍微高一点的洗碗机”，有人想要一把“免穿透的测温智能锅铲”，但来自诺兰·威廉姆斯（Nolan Williams）的愿望却是：“希望看到癌症药物与抗衰老延寿药物取得重大突破，因为更长的寿命意味着杰出的人才能有更充裕的跑道去推动创新、传承知识。”这种从锅铲直接跳跃到人类攻克癌症与长生不老的强烈对比，让演播室充满了既荒谬又深受启发的戏剧性张力。 ### [29:09](https://www.youtube.com/watch?v=uQawEBikW8U&t=1749s) 德里克·汤普森论 AI 时代的“超级个体”与单人创业潮 主持人谈到了科技评论家德里克·汤普森（Derek Thompson）最新发布的通讯文章，其核心论点是：“现在是凭借单打独斗发家致富的最佳时机（There's never been a better time to get rich working alone）。”节目中主持人插话打趣道：“他这是在说‘下金蛋的鹅本身没有被充分估值’吗？真正的价值其实在于那只持续下蛋的母鹅自身。”随后他们详细拆解了汤普森对 AI 与就业关系的深刻洞察。汤普森指出，当前围绕 AI 与就业的争论往往走向两极化，且双方都缺乏扎实的数据支撑：一边是宣称“AI 将夺走所有人饭碗”的末日论者（Doomers），尽管现实中美国核心劳动年龄群体的就业率依旧稳固；另一边则是坚称“AI 完全是一场毫无价值的骗局”的虚无主义者（Deniers），这种偏见让他们忽视了 AI 正在深刻重塑工作与经济结构的事实。 汤普森给出了基于数据的务实论断：当前正处于“高营收微型初创企业（tiny startups with big revenue）”的黄金时代。相关图表与行业数据清晰显示：独立创业者（solo entrepreneurs）与微型初创公司的数量正在爆发式激增；这些小型企业更倾向于深度整合 AI 工具，或直接销售 AI 相关的软件、咨询与设计产品；而支付巨头 Stripe 的跟踪数据更表明，当今的微型初创团队跨过数百万美元营收门槛的速度与频率，超越了以往统计过的任何一代企业。不过，这种“超级个体”的崛起也伴随着隐忧：它可能加剧美国社会的个体孤独感，同时大量穿透实体（pass-through firms）的涌现相较于传统的 W-2 高薪职业（如医生）缴纳更少税款，可能会对面临财政赤字危机的政府税收造成进一步冲击。 ```mermaid flowchart TD accTitle: AI时代单人创业崛起机制 accDescr: 展示AI工具如何推动微型企业快速增长并产生社会经济效应 subgraph Catalyst["技术驱动"] AI["AI 生产力工具普及<br/>软件 / 咨询 / 设计"] end subgraph Business["商业形态演变"] Solo["单人创业者 / 微型团队爆发"] Scale["Stripe 数据：<br/>极速跃升为数百万美元营收实体"] end subgraph Impact["宏观与社会影响"] Social["社会层面：个体孤独感上升"] Fiscal["财政层面：税源结构变化<br/>穿透企业税制挑战财政收入"] end AI --> Solo Solo --> Scale Scale --> Social Scale --> Fiscal ``` - **两极化的就业争论**：末日论者（Doomers）无视目前仍处于低位的整体失业率，而虚无否认者（Deniers）则对生产力的实质改变视而不见。 - **微型企业黄金期**：小规模团队利用 AI 实现杠杆效应，提供软件、设计与咨询服务，以极低的人力成本撬动巨大收入。 - **Stripe 数据印证**：现代微型企业达到数百万美元营收规模的速度创下历史纪录。 - **潜在财政与社会冲击**：高技能人才从传统高税率薪酬岗位转向轻资产独立运营公司，可能对公共财政税收形成压力。 --- ### [31:11](https://www.youtube.com/watch?v=uQawEBikW8U&t=1871s) 共享办公空间的困境与“独居小屋”的幽默畅想 顺着“单人创业者全面爆发”的话题，主持人探讨这是否会成为共享办公空间（Co-working spaces）的重大利好。然而另一位主持人随即泼了冷水，笑称：“对我来说，没有任何事情算得上是共享办公空间的利好（Nothing is bullish for co-working spaces to me）。”两人开始吐槽为什么至今没有人能打造出一个真正令人满意的共享办公产品——核心症结在于，人们归根结底渴望拥有真正属于自己的独立私密空间，而不是退而求其次地忍受公共环境。 这种逻辑随即延伸到胶囊公寓和极简生活的调侃上。主持人笑称，这就像有人幻想开发一种“每间只有 10x10 英尺、所有生活设施全靠共享”的蜂巢式公寓。两人互相打趣道，对方其实心里最渴望的生活状态就是直接搬进臭名昭著的“大学炸弹客（Unabomber）泰德·卡钦斯基”那种避世荒野小木屋，甚至曾由衷赞叹存放在仓库里的小木屋照片“看起来太酷了、简直是生活方式的巅峰”。玩笑归玩笑，大家一致认同那种极端微型的空间要想行得通，周围的大片专属私有土地才是真正的核心溢价与“配套设施”。在轻松愉快的插科打诨中，主持人总结了全场内容，预告节目将于太平洋时间工作日上午 11 点继续直播，并呼吁听众在 Spotify 等播客平台打出五星好评、订阅官方通讯。 ## Notable quotes > [00:02](https://www.youtube.com/watch?v=uQawEBikW8U&t=2s) 那天线要是离脑门太近感觉都能被辐射干掉 > [02:59](https://www.youtube.com/watch?v=uQawEBikW8U&t=179s) 上一次我穿这件蓝夹克还是在宣布 Neutron 火箭的时候，所以我想这代表今天又有大事要宣布了。 — Peter Beck > [02:59](https://www.youtube.com/watch?v=uQawEBikW8U&t=179s) 我得去买件新夹克了 — Peter Beck > [08:14](https://www.youtube.com/watch?v=uQawEBikW8U&t=494s) 他们这根本不是出了事故，纯粹是拿爆炸当旧设施的拆迁工程用了。 > [08:14](https://www.youtube.com/watch?v=uQawEBikW8U&t=494s) 如果知道负责给我主刀做手术的医生私下里还在兼职跑去搞对冲基金，我躺在手术台上恐怕会非常担忧。 > [12:12](https://www.youtube.com/watch?v=uQawEBikW8U&t=732s) 预测市场终将取代万事万物 > [27:38](https://www.youtube.com/watch?v=uQawEBikW8U&t=1658s) 想开多快就开多快，就算把它开翻车也无所谓，尽情去体验它的极限 — Rivian 团队 > [27:38](https://www.youtube.com/watch?v=uQawEBikW8U&t=1658s) 希望看到癌症药物与抗衰老延寿药物取得重大突破，因为更长的寿命意味着杰出的人才能有更充裕的跑道去推动创新、传承知识。 — Nolan Williams > [29:09](https://www.youtube.com/watch?v=uQawEBikW8U&t=1749s) 现在是凭借单打独斗发家致富的最佳时机 — Derek Thompson > [29:09](https://www.youtube.com/watch?v=uQawEBikW8U&t=1749s) 下金蛋的鹅本身没有被充分估值 > [31:11](https://www.youtube.com/watch?v=uQawEBikW8U&t=1871s) 对我来说，没有任何事情算得上是共享办公空间的利好 > [31:11](https://www.youtube.com/watch?v=uQawEBikW8U&t=1871s) 看起来太酷了、简直是生活方式的巅峰 ## Mentioned Rocket Lab, SpaceX, Iridium Communications, Brandon Guerell, Delian Asparouhov, Jeff Bezos, 蓝色起源 (Blue Origin), Starlink, Electron, Neutron, Peter Beck, Comcast, NBCUniversal, 华尔街日报 (The Wall Street Journal), 通用电气 (GE), Sky, 梦工厂 (DreamWorks Animation), Peacock, Verizon, AOL, Yahoo, Apollo, AT&T, DirecTV, 时代华纳 (Time Warner), Warner Bros. Discovery, Dave Limp, New Glenn, ChatGPT, Kalshi, 美国商品期货交易委员会 (CFTC), Meta, Mark Zuckerberg, Instagram, Facebook, Llama, CNN, CNBC, 纽约证券交易所 (NYSE), Mary Gillespie, Avalanche Energy, Morgan Clayton, Mary Gifford, Paul Blanco, Ray Lure, Cheryl Franklin, Jennifer Smith, Nolan Williams, Rivian, Chipotle, Derek Thompson, Stripe, Ted Kaczynski ## 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=uQawEBikW8U)

[0:02] 那天线要是离脑门太近感觉都能被辐射干掉 [2:59] Peter Beck: 上一次我穿这件蓝夹克还是在宣布 Neutron 火箭的时候，所以我想这代表今天又有大事要宣布了。 [2:59] Peter Beck: 我得去买件新夹克了 [8:14] 他们这根本不是出了事故，纯粹是拿爆炸当旧设施的拆迁工程用了。 [8:14] 如果知道负责给我主刀做手术的医生私下里还在兼职跑去搞对冲基金，我躺在手术台上恐怕会非常担忧。 [12:12] 预测市场终将取代万事万物 [27:38] Rivian 团队: 想开多快就开多快，就算把它开翻车也无所谓，尽情去体验它的极限 [27:38] Nolan Williams: 希望看到癌症药物与抗衰老延寿药物取得重大突破，因为更长的寿命意味着杰出的人才能有更充裕的跑道去推动创新、传承知识。 [29:09] Derek Thompson: 现在是凭借单打独斗发家致富的最佳时机 [29:09] 下金蛋的鹅本身没有被充分估值 [31:11] 对我来说，没有任何事情算得上是共享办公空间的利好 [31:11] 看起来太酷了、简直是生活方式的巅峰

[0:02] Peter Beck: This is a deal where 1 + 1 equals three, not just two... a fully integrated self-launching space superpower. [9:14] I think you just created vertical AI agents for tax fraud. [18:23] Mary Gifford: My father's sight was so bad that he had to look at the music about 2 inches away, memorize one or two phrases, then practice what he memorized. He gave his last recital from memory at age 94. [25:45] My latest hot take is that self-driving technology is commoditizing as fast as LLM technology... It will not be the domain of a single company.

[0:02] People were clamoring for a Rocket Lab [0:03] story that never came yesterday. Uh, [0:06] Rocket Lab enters the arena with SpaceX. [0:08] Brandon Grell fired up the TBPN [0:10] newsletter, which you can subscribe to [0:11] at tbpn.com today because we were on the [0:13] horn with Prof. We did his podcast. Very [0:16] excited for that to come out. Anyway, [0:18] Rocket Lab, they are duking it out with [0:20] SpaceX. Let's go through it. Yesterday, [0:22] the launch services and space systems [0:25] provider Rocket Lab agreed to purchase [0:27] satellite manufacturer/operator [0:29] Aridium in a cash and stock deal that [0:32] values the company at 8 billion or 20% [0:36] more than the share price at Friday's [0:37] close. Aridium pioneered LEO's LEO [0:40] satellites, launching its first one some [0:42] 30 years ago. The Aridium network was [0:44] the original satphone. this huge block [0:47] of a phone. You would take out the [0:49] antenna out of the back and the antenna [0:50] was big enough to kill you instantly if [0:53] you let it get near your brain. No, just [0:54] kidding. It was very sad, but you it [0:56] would be terrifying to anyone worried [0:58] about uh microwave technology or or uh [1:01] >> any sort of radiation, but it was wildly [1:03] successful. The company grew and grew [1:05] and now it's part of Rocket Lab. Uh [1:07] today, they operate a fleet of 66 [1:09] satellites. That seems low. That's [1:11] because these are bigger, older [1:12] satellites, but this partnership is [1:14] probably going up against Starlink. So, [1:16] it connects handsets and other devices [1:18] used by ships, mining sites, US [1:20] government agencies, enterprise [1:23] connectivity play. The move represents [1:25] yet more vertical integration at Rocket [1:27] Lab. Over the years, the company has [1:28] grown from small niche satellite launch [1:30] provider to an integrated space [1:32] powerhouse. It now designs satellites [1:34] and and aerospace components, serves the [1:36] defense primes, has the second most [1:39] launched US rocket annually, the [1:42] Electron rocket, and is currently [1:44] developing a reusable medium lift launch [1:47] vehicle. But Jeff Bezos is probably [1:49] coming for that silver medal any day [1:51] now. With the Idium acquisition, Rocket [1:52] Lab is positioning itself to compete [1:54] directly with SpaceX, which already [1:56] connects which already operates a 10,000 [1:58] satellite fleet. Rocket Lab has its work [2:01] cut out for it, says Brandon Guerell. On [2:04] August 25th, 2021, Rocket Lab spacked at [2:07] 4.1 billion, the usual spack price, [2:10] probably $10 a share, four billion [2:11] valuation. It stayed at or below that [2:13] price in the public markets for three [2:15] whole years until September 2024 when it [2:18] started to grow both revenue and get [2:20] rerated. Now, it's acquiring Aridium for [2:24] 8 billion, twice its original IPO price. [2:27] Uh, and Delian says, [2:28] >> "Yeah, you could buy it for around [2:30] $4.80ish [2:32] a share on July 1st in 2024." Yeah, it [2:35] is now sitting at $11. [2:39] >> Yeah, what a run. And uh, yeah, the [2:41] whole space economy has been booming. [2:43] Very, very good. It's a $60 billion [2:45] company, so they're picking up an $8 [2:46] billion uh, connectivity provider. Uh, [2:50] and we can play a little bit of this [2:52] from the founder of Rocket Lab, the CEO, [2:55] Rocket Man himself. Uh, Rocket Lab is [2:58] acquiring Aridium. [2:59] >> Last time I was wearing this blue [3:00] jacket, [3:01] >> look at this. [3:02] >> I was announcing Neutron. [3:04] >> Yeah. [3:04] >> So, I guess that means [3:06] there's something important to announce. [3:08] >> Do you ever Wait, pause for a second. Do [3:09] you ever put on a jacket like that? [3:11] [laughter] [3:12] >> Is this really the important? [3:14] >> This seems to be one of the more [3:17] >> Look at this. No, wait. [3:18] >> Whoa. W something important to [3:21] >> I don't know. [3:22] >> I mean, I don't think I have, but now I [3:24] want to try [3:24] >> something important to announce today. [3:26] >> This thing broke, too. [3:29] >> Okay, [music] we're going to go back to [3:30] school for a little bit. I want to [3:32] introduce you to the space application [3:35] equation. [3:36] >> Mhm. [music] [3:40] >> So, you've heard me talk a lot in the [3:42] past about applications. [3:44] It's where all the value in space truly [3:46] lies. But in order to exploit that value [3:49] to the fullest, you need a few other [3:50] things. The first thing is unfeted [3:53] access to space. So you need your own [3:55] launch. [4:00] And then of course you need your own [4:01] ability to build spacecraft at scale. [4:05] We've got that too. [4:10] But there's some other really big [4:11] barriers to building large successful [4:14] constellations in orbit. And the first [4:16] one [4:16] >> romantic companions [4:18] >> spectrum [laughter] spectrum is a finite [4:21] almost impossible to get and of course [4:23] not all spectrum is created the same. [4:25] The next thing is takes a long time to [4:29] build and launch your infrastructure. A [4:32] long time to design and build your [4:33] satellites, launch them and even longer [4:35] time to get your first $1 of revenue. [4:38] And then finally, it's a long time to a [4:41] sustained cash flow model, a long time [4:43] to build your business model, a long [4:44] time to build your customer base. And it [4:46] takes just that long extended time until [4:48] you've got proper reoccurring cash flow. [4:51] So, those of you who know me will know [4:53] that I'm way too impatient for that. So, [4:55] we found a bit of a shortcut. [4:59] Rocket Lab is acquiring Aridium [5:01] Communications. [5:04] >> Good music. Very clear video. This will [5:07] be one of the most transformative deals [5:08] in the space industry. It's the ultimate [5:10] combination for growth. And if you think [5:13] back to our little whiteboard session a [5:14] moment ago, this is a deal where 1 + 1 [5:17] equals three, not just two. One being [5:20] rocket lab, we have unfeted access to [5:22] space and unfettered [5:23] >> unfettered [5:25] missions again. [5:26] >> And then you think of Aridium. They have [5:28] an already operational constellation [5:30] spectrum. Not just any spectrum, [5:32] extremely valuable spectrum. the [5:33] spectrum [5:34] >> millions of customers and they're a [5:35] profitable business. [5:36] >> SpaceX just bought some spectrum. [5:37] >> Where the three comes in is the result [5:38] of these two things is a fully [5:40] integrated self-launching space [5:42] superpower. One that will unlock more [5:44] growth from existing network and build [5:47] new constellations to unlock new [5:48] services and market opportunities. This [5:51] is officially our entrance into the [5:53] space applications market. A thing we've [5:55] been talking about for a long time now. [5:57] But to be clear, this is not the finish [5:59] line. Now, we won't just continue [6:02] network as is. We're going to build upon [6:03] it to unlock new markets and pioneer new [6:06] space-based services. Rocket Lab's [6:08] future in space applications has just [6:11] been unlocked and accelerated. [6:17] [music] [6:18] >> And with that, [6:18] >> I like this guy. [6:19] >> I'm off to buy a new jacket. [6:21] >> This [laughter] thing's too small. [6:24] >> See, it it goes full circle. The reason [6:28] he had to put the jacket on sort of [6:29] awkwardly was it's too small. and it was [6:31] a part of the story. [6:32] >> Yeah, it was good. That's a really good [6:34] video. I like that. Uh anyway, Comcast [6:37] split. In 2011, Comcast, which was [6:40] primarily a cable, TV, internet, and [6:42] phone provider to something like 30 [6:44] million Americans, bought a 51% stake in [6:47] NBC Universal from General Electric. I [6:51] love it. I love an industrial company [6:52] that owns a media company. It's happened [6:54] before, folks. It's not that crazy. Just [6:57] might work. In 2013, it purchased the [7:00] remaining 49% stake. Uh, since then, [7:03] Comcast has acquired the European media [7:05] company Sky, the movie studio DreamWorks [7:08] Animation, and through NBC Universal [7:10] spun up Peacock, among other activity in [7:12] the entertainment space. Yesterday, the [7:14] Wall Street Journal reported that [7:16] Comcast's decadel long into trying to be [7:19] a content plus distribution company is [7:21] coming to an end. It's planning to split [7:23] the business into two. The Comcast [7:25] entity will once again becoming a become [7:27] a connectivityonly business and NBC [7:30] Universal will focus entirely on [7:33] content. And there's been a number of [7:35] spinouts uh that have been happening for [7:37] a while. Um and this is the the the [7:40] final like big one I guess. Uh there's a [7:42] recent history of telecom connectivity [7:43] companies rolling up entertainment [7:45] businesses and then spinning them out or [7:47] get rid of or getting rid of them. Over [7:49] the past decade, Verizon bought AOL and [7:51] Yahoo, reorganized the two into a entity [7:54] called Verizon Media, and ultimately [7:56] sold Verizon Media to Apollo in 2021. [7:59] Over roughly the same period, AT&T [8:01] bought DirecTV and Time Warner, but [8:04] ended up fully exiting its position in [8:06] Direct TV and spun off Time Warner as [8:08] Warner Media into its own entity in [8:10] 2022. So, interesting deal to follow [8:13] along. The other good news in the space [8:15] world is that uh after that explosion at [8:19] Blue Origin launch, that new Glenn [8:22] launch that was uh so disheartening, [8:24] there were rumors that uh it might be 2 [8:28] years until they got that launchpad back [8:30] online, but there is some updates here. [8:33] There are some updates to return to [8:35] flight this year. We're not rebuilding [8:36] the same pad. We're going straight to a [8:38] horizontal vertical hybrid conops. We [8:41] already we had already been developing [8:43] for our 9x4 new Glenn launch vehicle [8:46] using existing infrastructure skipping a [8:48] new transporter erector and creating a [8:52] common conops across two paths. So, uh [8:55] Dave shares [8:56] >> they just used the explosion as sort of [8:58] like a demo. They demoed the old setup, [9:00] >> a demolition. Yeah, they did. Uh, but [9:04] uh, good news from Dave Limp, the CEO of [9:06] Blue Origin, that they're going to get [9:09] New Glenn flying again sooner than [9:12] people thought. I have a pitch for you. [9:14] I have I have a I have a business idea. [9:16] I want I want to get your feedback. So, [9:17] let's hear it. [9:18] >> What's the biggest problem with being a [9:19] doctor? You make a lot of money as a [9:20] doctor, right? What's the biggest [9:21] problem? [9:23] >> Uh, the odd hours. [9:25] >> No, taxes. You make so much money, you [9:27] get taxed on on normal income, ordinary [9:30] income. Even if you're a doctor, you're [9:32] saving people's lives. You're doing the [9:34] most important job in our economy. [9:35] >> You have a doctor buddy that also runs a [9:38] hedge fund, though. [9:39] >> And and when he runs the hedge fund, [9:40] he's making capital gains. But when he [9:42] when he does an operation, he makes [9:45] ordinary [9:45] >> to me was personally very concerning. I [9:48] don't know if I'd want to go to see a [9:49] doctor that does surgery that that also [9:52] >> to be fair, he's building a new clinic [9:54] and he has time off while while he's not [9:56] up and running. But but when I heard [9:57] that, you know, a surgeon was also [9:59] running a hedge fund, I I got a little [10:01] bit worried just, you know, if I was [10:03] under under the knife. [10:04] >> Yes. [10:05] >> And uh [10:06] >> but [10:06] >> there were some and the market was open, [10:08] I'd be I'd be a little nervous. [10:10] >> But the the doctor makes ton of money, [10:13] saves people's lives, does all sorts of [10:15] good things, gets taxed at ordinary [10:16] income, not capital gains. So you're [10:18] taxing doctors who are saving lives at a [10:20] higher rate. So my business idea is with [10:24] the power of Agentic AI, we create [10:27] thousands and thousands of Delaware SE [10:30] corps for every doctor. So every time [10:33] that there's a doctor's appointment and [10:35] you have to pay the doctor, you acquire [10:39] one of their Delaware C corps and within [10:43] that CC Corp there is uh on the balance [10:46] sheet of that company basically an IOU [10:49] for the services the doctor must render [10:51] you [10:52] >> sitting on the balance sheet. [10:53] >> Sitting on the balance sheet. [10:53] >> Yeah. It's an asset. [10:54] >> It's an asset. And so you can you can [10:57] then depreciate that asset since which [10:59] since you've acquired it. The doctor [11:01] reaps capital gains and then you cash in [11:05] your [11:06] >> Yeah. And they can have a sort of a farm [11:07] of C corps. So there's they're not, you [11:09] know, they've had them the shares for [11:11] over a year. [11:13] >> Exactly. Long-term capital gains. And [11:15] then and then the the patient on their [11:17] way out, they sign one form. Agentic AI [11:20] does its thing, winds down the CC corp [11:22] and and cashes in the asset and then uh [11:26] the doctor can reap capital gains. What [11:28] do you think? You think we got a [11:29] business there? [11:30] >> I think you just created vertical AI [11:32] agents for tax fraud. [11:34] >> I asked JGBT and it said, uh, this is [11:37] almost certainly dead on arrival. Not [11:40] clever tax optimization, more like [11:42] abusive tax shelter with healthcare law [11:44] side quests. [11:46] >> [laughter] [11:46] >> The core problem is the economic [11:48] substance is obvious. The patient is [11:49] paying for medical care. Doctors being [11:51] paid for performing medical services. [11:53] [laughter] [11:54] IRC61 [11:56] includes compensation for services and [11:58] fees and gross income. And the IR and [12:00] the IRS says you generally include [12:02] everything received in payment for [12:03] personal services. Wrapping the bill in [12:05] a newly minted CC Corp does not change [12:08] the income character. Well, I think [12:10] we'll have to take that one to the [12:11] Supreme Court. Take it away. [12:12] >> I think if we take it to the Supreme [12:13] Court, there's a chance. I I think you [12:15] can wrap it in a prediction market, [12:17] right? So every single [laughter] time [12:18] you go to the doctor, you say, "Okay, [12:20] there's like 0% chance the doctor's [12:22] going to do well." And then he bets yes. [12:24] >> He bets yes. [12:25] >> And then we'll see what happens. But you [12:26] know, if it's a good doctor, then he'll [12:28] be paid out in full. I'm pretty sure [12:29] >> it adds insurance to it as well because [12:31] you don't pay if the doctor doesn't do a [12:32] good job. [12:33] >> Yeah. It's like insurance is built in. [12:34] >> Built in. [12:35] >> So I think this is actually [12:37] >> Are winnings on on prediction markets [12:39] capital gains? [12:40] >> I think it's still mostly unclear. So [12:42] for the next few years, I think this [12:43] would definitely be a good strategy. [12:45] >> Okay. I like it. I like it. I You have [12:47] to imagine because they say they don't [12:48] use the word bet, they use the word [12:51] trade, [12:51] >> trade, predict. So capital gains, [12:54] >> trading, capital gains treatment maybe. [12:56] >> I think that the post uh last year where [12:59] someone was saying like prediction [13:01] markets will replace everything. If I [13:03] want blueberries delivered to my house, [13:05] I go to the market for blueberries being [13:07] delivered to my house and I bet $15 and [13:10] I bet no. someone else. They bet yes. [13:13] They deliver me blueberries. [snorts] [13:14] They get the $15. [13:16] >> Works [laughter] every time. [13:19] Works every every single time. We we we [13:22] keep debating this uh this meta question [13:25] of will they launch a financially [13:28] incentivized prediction market or will [13:30] it purely be for social cloud? You're [13:32] still in the camp of there's going to be [13:34] real dollars. It just feels like the [13:35] first time meta. Only reason the reason [13:38] I I I believe that is because I believe [13:42] they had Meta [13:43] >> Yeah. [13:44] >> like was reaching out to the various [13:46] prediction market providers to try to be [13:49] potential like partners andor vendors. [13:52] >> But I mean partnership could mean so [13:54] many different things, right? [13:55] Partnership with with a prediction [13:57] market can mean we'll give you ad reads [14:01] or discounted ads or we'll vend your [14:04] data so that if someone is on Instagram [14:07] and hits the search box acting for [14:09] something and Llama 5 or Muse Spark [14:12] wants to answer it and surface data from [14:15] a prediction market, we have the rights [14:17] to do that without infringing on your [14:20] copyright or something like that. Like [14:21] it could be a data licensing [14:23] partnership. It could be an acquisition [14:25] funnel. It could be something where [14:27] Meta's funneling people to the [14:30] prediction markets or back like that. [14:32] That whole I I I read the same article [14:34] that you read and it doesn't make any [14:35] sense to say, "Oh yeah, I want to create [14:37] a competitor, so I'm going to reach out [14:39] to my direct competitors and see if I if [14:41] they want to give me their customers." [14:43] Like, why would they ever do that? That [14:44] makes no sense. You see what I'm saying? [14:45] The only thing is like the prediction [14:47] markets have a history of partnering [14:49] with with other applications and [14:51] basically being the infrastructure [14:53] providers. [14:54] >> Yeah. to actually place real [14:56] >> but if you go to any any partner whether [15:00] it's you know there was that whole suite [15:02] of like [15:03] >> I I agree with you like [15:05] >> I I I generally agree that there is a [15:07] possibility that there's no dollars and [15:09] it's [15:09] >> CNN CBC [15:11] >> but I just don't I I [15:12] >> New York Stock Exchange they all have [15:14] partnerships with uh prediction markets [15:17] none of them have like oh go in the CNBC [15:20] app and and upload your credit card [15:22] information like Meta's been on this [15:25] path of like maybe we'd like a financial [15:27] relationship with our customers for [15:29] decades and like it's never really [15:30] matured. There was always this idea that [15:32] you would be able to like save your [15:34] payment infrastructure in information [15:35] and then when you see a t-shirt you [15:37] could just click and oneclick checkout [15:39] in the Instagram app or in the Facebook [15:41] app and that never really took off for a [15:43] variety of reasons. [15:44] >> So John in this case it would be like [15:45] you see an ad and then it's like I place [15:47] a bet that there won't be a t-shirt [15:50] delivered to my house and then it'll get [15:52] fulfilled. This is the final boss. [15:54] >> Hopefully [15:54] >> hopefully the answer to your question on [15:57] can doctors reap capital gains by uh [16:00] using prediction markets instead. [16:02] Prediction market profits are taxable [16:04] but the characters unsettled at the [16:06] moment. CFTC regulated event contracts [16:09] like kshi style contracts. There is a [16:11] decent argument that they are financial [16:13] derivatives uh rather than gambling. The [16:15] CFTC describes event contracts as [16:17] derivative contracts whose payoff [16:19] depends on the specified event. uh it [16:21] also granted kshi status as a designated [16:25] contract market. So best case some [16:29] contracts might be treated as 1256 [16:32] contracts that means marked to market [16:34] annually and gains and losses split [16:37] across 60% long-term capital 40% [16:40] short-term capital regardless of holding [16:42] period. So there's a chance that that's [16:44] the that that's the solution. And I [16:46] think we might we might have solved it, [16:47] but we'll have to use every tool in the [16:49] modern financialization and technology [16:51] box. The Wall Street Journal asked their [16:54] readers what innovations they want to [16:57] see in the next 20 years, and the [16:59] answers are sort of all over the place. [17:01] And [17:01] >> number five will shock you. [17:03] >> Number five might shock you. Um, what [17:06] innovation do you want to see in the [17:07] next 20 years? Tyler, think of [17:09] something. Jordy, think of something. [17:10] I'll read some of what the Wall Street [17:13] Journal readers wrote in. Mark uh Mary [17:15] Gillespie from Irving, Texas said, [17:17] "Solar on wheels." She wants She says, [17:20] "Why can't cars be manufactured with [17:22] solar panels integrated into the top of [17:24] the vehicle? Then it could recharge the [17:26] car's battery as you travel and when you [17:29] park." So, this is not actually a new [17:31] idea. There are cars that have solar [17:33] panels on the roof, but solar panels are [17:36] extremely weak and so you'd have to [17:38] leave the car outside for like 5 months [17:39] to charge a full electric car. Now, [17:41] there have been DARPA Grand Challenge [17:44] cars and specific uh almost science [17:47] experiments vehicles where they have a [17:49] huge surface area and it's this massive [17:52] wing of solar panels and uh I think [17:55] there's a plane that's also potentially [17:57] be able to fly indefinitely by having [18:00] solar panels that recharge a motor and [18:02] it spins. Um but for a normal car, a [18:06] single solar panel on the roof doesn't [18:08] get you much in the way of charging. [18:10] She's basically asking for stronger [18:12] solar density, stronger energy density. [18:14] And I'm in. I'm in. In 20 years, totally [18:17] feasible. Mary from Irving, Texas. I [18:20] think it's gonna happen. I think it's [18:21] possible in 20 years. Anyway, traffic [18:23] jam plan. I would love to see an [18:25] autopilot feature. I would I would love [18:28] to see autopilot features on cars talk [18:30] to each other. You're saying so cars can [18:32] nuclear cars. Just get a nuclear car. [18:34] >> That's very uh very Fallout. I I I like [18:37] the What is that? Nuclear punk or [18:39] something. What was that? was that [18:40] there's like this alternate history of [18:42] like the 1960s if it was powered by [18:44] nuclear. Uh not solar punk, atom punk. [18:47] Atom punk is the term. Uh and it has all [18:49] these really cool things that could be [18:51] powered by nuclear. Maybe we'll get [18:53] there. Fusion uh is sort of promising [18:55] that. We've talked to a couple companies [18:56] that do smaller fusion reactors. Um and [19:01] they uh tend avalanche is one up in uh [19:04] Seattle or maybe Washington where uh [19:07] it's basically the nuclear nuclear [19:10] reactor but it's uh it's fusion not [19:12] fision and it's about the size of you [19:15] know a large battery and so you could [19:17] potentially put it in a car, put it in a [19:18] drone, put it in a space vehicle, uh [19:22] keep it up there for a long time. Uh [19:23] accordion traffic jams would be a thing [19:25] of the past. What is an accordion [19:26] traffic jam? I think that's when the [19:28] traffic's like [19:29] >> oh going in and out in and out. It's [19:32] like people are stopping and going and [19:33] there's that slight delay when [19:35] theoretically if everyone was like [19:36] moving at the perfect [19:38] >> Yeah. [19:38] >> you know. [19:38] >> So Morgan Clayton from Birmingham, [19:40] Alabama wants autopilot features on all [19:43] the cars to talk to each other so they [19:45] never do that. So they know, okay, I got [19:47] to speed up, I got to slow down, and [19:49] then they're all perfectly in sync. [19:52] Probably doable. You need some standard [19:54] or something. What are you thinking? [19:56] Musical vision. This is a good one. [19:57] These people are thinking outside the [19:59] box. This is not like, oh, I want [20:01] inference at onetenth the cost. [20:03] [laughter] [20:04] Silicon Valley is unimagin unimaginative [20:07] by comparison. But Mary Gford from [20:11] Alpine, Utah says she wants musical [20:13] vision. I'd like to see the invention of [20:15] a device to help musicians with eyesight [20:18] limitations read music. There are many [20:21] ways to help the sight impaired read the [20:23] printed word, but pianists and others [20:26] need to read music at the piano. My [20:28] father's sight was so bad that he had to [20:30] look at the music about 2 in away, [20:33] memorize one or two phrases, then [20:35] practice what he memorized. He gave his [20:37] last recital from memory at age 94. What [20:40] a legend. Let's give it up for Mary [20:42] Schikner Gford's [20:44] father, grandfather. Um, powerful crazy [20:48] idea. Can't you just print it bigger if [20:52] you're [20:52] >> Whoa. [20:52] >> having trouble seeing it. [20:54] >> You just invented the semi-truck size [20:57] printer. [20:58] >> Yeah. I mean, you could just print like [20:59] a poster board and then you could see it [21:01] at the piano. Um, also glasses. [21:04] >> Yeah. I wonder I mean it might be a [21:07] small might be a a small market of [21:09] people whose vision is so impaired that [21:11] even with glasses it needs to be [21:13] >> up close. [21:13] >> Up close. I'm just confused by why this [21:15] the invention of a device to help [21:17] musicians with eyesight limitations read [21:19] music. Why are we focused on reading [21:23] music? Wouldn't this be wouldn't any [21:25] technology that allows you to read music [21:28] if you're hard of sight [21:30] allow you to read anything? [21:33] >> Are we talking general technology? [21:35] >> I mean, couldn't you have uh couldn't [21:37] you learn individual notes and then [21:41] learn the order of them? have it spoken [21:43] to you. [21:44] >> Yeah. And [21:44] >> I do think maybe iPad app, iPads often [21:48] sit at the at the piano and you could [21:50] have an iPad app that shows like sort of [21:52] like one note or two notes. [21:54] >> What about a humanoid robot that grabs [21:56] your hand and bends [21:58] >> puppeteers you like a marionette [22:00] >> and you know slams your hands. [22:01] >> Yeah, totally possible. [22:03] >> Paul Blanco wants something else. He [22:05] wants a smart spatula. [22:06] >> This is so much more attainable. [22:08] >> I'd love a cooking spatula. human [22:09] teleportation [22:10] >> that could tell without penetration what [22:12] temperature or how thoroughly cooked a [22:14] hamburger or steak is which is rare. Do [22:16] we not have uh we we we literally have [22:19] things you just [22:20] >> No no wants without penetration. It has [22:23] to you just rest the spatula on top of [22:25] the steak and it tells you inside the [22:28] steak what temperature it is which is [22:30] very difficult because if you're searing [22:31] or doing a reverse sear or something the [22:33] outside could be a wildly different [22:35] temperature than what's going on inside. [22:36] So, how do you detect the temperature [22:39] without penetrating the steak with a [22:41] thermometer? [22:42] >> A powerful X-ray. [22:43] >> Imagine it's as big as that brain [22:45] scanner thing. [laughter] It's just [22:47] like, yeah, here's your smart spatula. [22:50] Load it up. It's like MRIing the entire [22:53] steak. That'd be a good time. Robot [22:55] chefs. Oh, this is related. I want a [22:57] robot that can work as an executive chef [22:59] in my own kitchen, says Ray Lure from [23:02] Lacy, Washington. It could call a [23:04] grosser, order ingredients, insist on [23:07] insisting on it all being upscale and [23:10] have those ingredients delivered, [23:11] unpack, load the fridge, and then use my [23:13] appliances to make dinner. 20 years, I [23:16] think it's possible. I like that. That [23:17] that's right on the right like, you [23:19] know, level of sci-fi. Seems doable. Uh, [23:22] not [23:22] >> for what it's worth, but also not [23:24] something we can whip up in a weekend [23:25] with an Arduino like the smart. [23:26] >> Some of these ideas are silly, but I'm [23:28] viewing it as a very literal request for [23:30] startups. [23:31] >> Yes. Yeah. No, I like it. Uh, quiet [23:34] inside. Cheryl Franklin from West Grove, [23:36] Pennsylvania says, "Given the negative [23:39] effects of of excessive noise such as [23:41] elevated blood pressure, cardiac stress, [23:44] and lower real estate values, I'd love [23:46] to see improvements in building [23:47] materials. For example, drywall with [23:49] built-in soundproofing capabilities. [23:52] Love that. Perhaps a lightweight sound [23:55] >> called lead paint. Cheryl, [laughter] [23:56] bring it back. [23:57] >> Lead paint. That doesn't do anything to [23:58] sound." Uh perhaps a lightweight sound [24:01] blocking film could also be developed [24:03] for use in existing buildings at home. I [24:05] had a startup idea a while back. So uh [24:07] you know when you have a young baby, a [24:09] child, and they every time you go to [24:11] change the diaper, very noisy. One of my [24:13] friends put up sound panels like a [24:15] podcast studio, full slat wall in the [24:19] slat wall [24:19] >> in the nursery. [24:20] >> Dual use technology. [24:21] >> Dual use technology for sure. And it and [24:25] it dampens the cry so you can so you're [24:27] not so echoey because sometimes if a [24:29] kid's crying in a very echoey room it [24:31] really reverberates and it can be kind [24:33] of crazy. [24:34] >> And lead paint is dense. So in theory [24:36] any added mass can reduce sound [24:38] transmission. So maybe add a little bit [24:40] of lead paint [24:42] >> to your nursery. [24:43] >> Lead plates. Whole lead plates. Maybe [24:45] whole Faraday cage. Um but here was the [24:48] idea. You don't. No one wants a slat [24:49] wall in their nursery in their baby's [24:53] room. [24:53] >> I don't know. [24:54] >> Maybe if you're trying to breed the [24:56] super podcaster, but most people don't [24:58] want that. They want animals. They want [25:00] a pastoral nature vibe, some warm [25:02] pastels, some nice warm tones, something [25:05] welcoming for the new child who's just [25:07] been brought into this world. And so the [25:09] idea was a typical sound panel, square, [25:13] flat. You know, you might see the egg [25:15] crate, the black material, but you you [25:18] paint on it an animal. You paint it on [25:20] it childlike, you know, paintings that [25:23] you would hang on the wall. It looks [25:24] appropriate. It fits the decor of the [25:26] room, but it also has the added effect [25:29] of sound treating the room so that it's [25:31] less noisy when everyone's yelling. Good [25:33] business. [25:34] >> Banger. [25:34] >> Banger. There we go. [25:35] >> Banger. [25:36] >> Got a banger. better than C corps on [25:39] demand for doctors for doctors. [25:44] >> Um [25:44] >> maybe certainly simpler. [25:45] >> Jennifer Smith, we have made so many [25:47] advancements in so many areas, but what [25:49] I want most is a way to control outside [25:52] noise. This is the superpower she wants. [25:54] It would be so nice to be able to sit on [25:55] my screened in porch and enjoy listening [25:57] to the birds without having to hear [25:59] nearby traffic, which sounds like drag [26:02] racing these days. Well, that's because [26:05] it is drag racing, Jennifer. We're drag [26:07] racing in your neighborhood. If sounds [26:11] are waves, can't we block some? How have [26:13] we not invented an open air type of [26:16] noise cancelling headphones to shield [26:17] noise waves and give us some quiet [26:19] spaces outdoors? [26:21] >> Interesting. I mean, the the simple [26:24] solution here is just more electric [26:25] vehicles. They're quieter. You don't [26:27] sound like no uh drag racing. Obviously, [26:30] there's a whole bunch of emission [26:31] standards that also reduce reduce noise [26:34] from internal combustion engines, at [26:36] least until you chop the exhaust and [26:39] straight pipe that thing. But if you're [26:41] not doing that, [26:42] >> or you engine swap Yeah. [26:43] >> your Model 3. [26:45] >> Yeah. Yeah. I don't know. Is this [26:46] possible? Could you just put out a big [26:49] speaker that selectively noiseancells [26:52] car noise but not bird noise? That seems [26:56] possible. seems [26:58] >> I personally I would, you know, take my [27:00] home. I' I'd get really thick windows, [27:03] ideally lead paint in the whole home, [27:05] you know, [laughter] [27:06] >> the lead paint. Just love the lead [27:08] paint. [27:09] >> And I'd bring the birds inside. [27:11] >> I got an alternative theory. If you got [27:13] drag racers racing down your your your [27:16] quiet uh birdinfested [27:19] street, uh you got to go wy coyote mode. [27:22] You got to put out a fake tunnel that [27:25] out of bricks. Paint it with some lead [27:27] paint. Maybe uh the fake tunnel the drag [27:30] racer smashes into the bricks. They'll [27:32] never be racing down your street again. [27:33] >> Drop drop an anvil on their [27:35] >> Yeah. Dropping an anvil on the car. That [27:37] works too. [27:38] >> We have to talk about the Rivian team [27:39] cuz they sent you this this demo unit to [27:41] try and it's a demo unit so drive it as [27:45] fast as you want. You can roll it. You [27:47] can roll it if you want. Like it's no [27:49] big deal. Just like really enjoy the car [27:51] to the fullest. I was pretty surprised [27:53] that they said that. [27:54] >> That's great. It's actually an amazing [27:56] car. I like it. Uh the self-driving [27:58] feature, deeply underrated. My latest [28:00] hot take is that self-driving technology [28:02] is commoditizing as fast as LLM as LLM [28:06] technology and that uh companies that [28:08] take it seriously and actually [28:09] prioritize it. Uh it is not some insane [28:13] uh insane tech breakthrough that cannot [28:15] be uh implemented in other vehicles. It [28:18] in fact can be implemented in other [28:19] vehicles. Just like we've seen Chipotle [28:21] get a chatbot, we will see every car be [28:24] autonomous. It will not be the domain of [28:26] a single company. Um, funniest thing [28:28] about this, you got someone who's asking [28:30] for just a dishwasher that's a little [28:32] bit taller. You got somebody who's [28:34] asking for a smart spatula. And then you [28:38] have Nolan Williams who says, "I would [28:40] like to see significant developments in [28:42] cancer and longevity drugs." You're [28:45] sitting there, you're going around with [28:46] the Wall Street Journal and uh and [28:48] you're like, "Oh, like what did you ask [28:50] for?" Like like [28:51] >> smart spatula. I was like, "What did you [28:53] ask for?" Like [28:55] >> longer curing cancer. You're like, [28:58] "What? I didn't know. I didn't know it [29:00] was that type of question." [29:01] >> Uh he says, "Longer lives mean longer [29:03] runway for brilliant people to bring [29:05] forth their own innovations and pass on [29:06] their knowledge." Wow. Nolan Williams, [29:09] really, really great answer. Um Derek [29:11] Thompson has a new newsletter out. He's [29:13] breaking it down. He says, "There's [29:15] never been a better time to get rich [29:17] working alone." Block in. [29:19] >> Is he saying the goose is not valued? [29:21] >> I think he's saying the goose is not [29:23] valued. [29:24] >> He's saying the goose is not valued. [29:26] >> He says the debate about AI and jobs [29:28] often breaks down into two extreme [29:30] groups. Both of which have an evidence [29:32] problem. On the one hand are the doomers [29:34] who say AI will take everyone's job even [29:36] though unemployment remains low and the [29:38] unemployment rate for prime age [29:39] Americans is still very high. on the [29:41] other hand are the are the deniers whose [29:44] exist whose insistent that whose [29:46] insistence that AI is a worthless scam [29:49] prevents them from seeing the many ways [29:52] it's changing work and the economy [29:54] broadly. If you want a strong and [29:56] evidence-based take about AI and jobs, I [29:59] have one for you. Quote from Derek [30:01] Thompson. There's never been a better [30:04] time for workers to get rich by going [30:06] independent. This is the golden age for [30:08] tiny startups with big revenue. The [30:11] evidence charts one and two, the number [30:12] of solo entrepreneurs and tiny startups [30:15] is taking off. Great news. Chart three, [30:18] there is rising evidence that these [30:19] small firms are more likely to use AI [30:21] and sell AI related products, whether [30:23] it's software, consulting or design. [30:25] Chart four, [30:26] >> you're saying that the matters is not [30:28] the eggs, it's the goose itself. Is the [30:30] true value is the power to keep laying [30:31] eggs. [30:32] >> Yes. Yes. [30:33] >> Okay, continue. I just want to make sure [30:35] I understood. Today's micro startups are [30:38] becoming multi-million dollar firms [30:40] faster and more frequently than any [30:42] generation of firms that stripe has [30:43] measured. Today's article is about the [30:45] rise of the solo act in the US labor [30:47] force and its implications for the [30:48] future of business social life. More [30:50] solo entrepreneurs even more aloneeness [30:52] and the fisc in America's fiscal crisis. [30:54] More pass through firms even less tax [30:56] revenue for a revenue starve government. [30:58] Interesting. So if if you have a bunch [31:00] of talented people that would be earning [31:02] money as doctors but then they go and [31:04] and they wind up uh starting a business [31:07] then yeah maybe maybe that's lower tax [31:09] that feels like that's further off. Is [31:11] this a bull case for co-working spaces? [31:14] The rise of solo entrepreneurship. [31:16] >> Nothing is bullish [laughter] [31:18] for co-working spaces to me. [31:21] >> It's rough. Why is it bad? Why has no [31:24] one been able to make a good co-working [31:26] space? because I think it's just nice to [31:28] have your own space. [31:31] It's the same I mean the same logic [31:33] would be like why don't they make an you [31:34] know why don't they make an apartment [31:36] building where you get a tiny little I [31:38] mean they do do this [laughter] [31:41] >> like why don't they have a pod [31:43] >> yeah basically I mean it's basically [31:45] like imagine if you rented your [31:46] apartment and it was actually like as [31:48] big as like one [31:49] >> why don't they make like a smaller [31:51] version of a house that you can rent [31:53] that's like in a building with other [31:55] other like of these mini houses [31:57] >> I'm just saying like a a 20 a 10t by 10t [32:00] room that you rent and then all the [32:02] amenities. [32:02] >> I was telling you that you would happily [32:03] live in the uni bomber shed. [32:07] >> You were saying that it's the pinnacle [32:08] of lifestyle. [32:10] >> I I did think the picture of it in [32:12] storage looked cool, [32:14] >> but it needed a slight [32:16] >> The difference the difference there is [32:18] like that shed on the right property [32:20] would be very cool. [32:21] >> That's true. That's true. [32:22] >> But but the property is is is sort of an [32:25] amenity. [32:26] >> I don't know. We'll figure it out. Thank [32:28] you for watching TVPN. We'll be back [32:29] tomorrow at 11:00 a.m. Pacific. [32:31] >> Oh, hang out with us. We'll close it [32:32] out. [32:33] >> Five stars on our podcast and Spotify. [32:35] Sign up for our newsletter at tbpn.com. [32:43] >> [music]