首页  >>  来自播客: Solving The Money Problem 更新   反馈  

Solving The Money Problem - World Reacts To SpaceX Earnings & $1 Trillion (2030) Revenue Goal

发布时间:   原节目
以下是内容的中文翻译: 该视频讨论了 SpaceX 作为一家上市公司首次发布的季度财报,重点关注其成为全球首个10万亿美元公司的潜力。 **当前估值背景:** * 苹果、Alphabet 和英伟达今天的市值约为5万亿美元。 * SpaceX 不足1.5万亿美元,这意味着它需要将其估值提高约七倍才能达到10万亿美元,而其他公司只需翻倍。 **彼得·戴曼迪斯(SpaceX、特斯拉、谷歌的早期投资者;X Prize 创始人)对 SpaceX 潜力的看法:** * 相信埃隆·马斯克能让 SpaceX 成为首个10万亿美元公司,因为它正在建设“文明级别的基础设施”。 * 他认为市场低估了 SpaceX,因为它是由四个垂直整合的独立业务组成。 * **业务1:全球通信(星链 Starlink):** * 提供全球千兆连接。 * 根据格温·肖特韦尔在财报电话会议上的说法,卫星直连手机服务拥有6000亿美元的营收潜力。SpaceX 预计将从中占据很大一部分。 * **业务2:超大规模计算(AI计算租赁):** * 该业务已签署了200亿美元的年度经常性收入(ARR)。 * 建设计算基础设施的速度比竞争对手更快。 * 宣布计划在今年底前实现2吉瓦(GW)的计算能力,目标是明年达到10吉瓦。 * 预计将成为最大的收入来源。 * 在地球之外,他们还在建设轨道计算能力。 * 埃隆·马斯克表示,他们将在2027年与英伟达合作发射首批AI卫星,这比预期快得多。SpaceX 已申请发射100万颗此类卫星。 **“解决金钱问题”节目主持人对AI计算的评论:** * 强调AI计算是 SpaceX 最关键的机会(“AI计算,TMD!”)。 * 未来几十年,对智能和AI计算的需求“几乎是无限的”,无论是在训练还是推理方面。 * SpaceX 预计到今年12月,其年度经常性收入(ARR)将达到1000亿美元以上,主要来自新的AI计算交易。 * SpaceX 的AI计算投资回收期**不到一年**,这使其在本质上更像是销货成本,而非传统的资本支出(capex)。 * 据报道,谷歌和 Anthropic 等公司“每月共计向 SpaceX 支付数十亿美元”以获取计算资源,因为 SpaceX 部署这些资源的速度更快、成本更低。 * 他对特斯拉和 SpaceX 的20年估值模型显示,从长远来看,AI计算租赁将贡献大部分收入和利润,远远超过星链、Grok(AI模型)或发射业务。 **丹·艾夫斯(Wedbush 分析师)关于资本支出和投资者耐心的看法:** * 表示 SpaceX 建设所需的资本支出是“必要的”,但将“需要投资者保持耐心”。 * 将这种情况描述为一场“军备竞赛”,其中大规模资本支出不可避免。 * 同意“彩虹尽头的宝藏是数据AI”,并且这一愿景延伸到 XAI 和可能到特斯拉。 * 承认芯片和内存成本正在上涨,SpaceX 是众多争夺这些资源的公司之一。 * 当被问及投资者为何应选择 SpaceX 而非已成熟的AI公司(Anthropic/OpenAI)或云服务提供商(微软/亚马逊)时,艾夫斯认为 SpaceX 目前“正在迎头赶上”,投资者购买的是马斯克的愿景。 * 他强调“资金成本”正在上升,尽管公司知道减少资本支出意味着落后,但投资者最终会要求看到变现。 * 他将 SpaceX 与微软和 Alphabet 等公司进行了对比,这些公司已经通过投资实现了变现。 **“解决金钱问题”节目主持人的结论:** * 反驳了丹·艾夫斯关于资本支出导致“困难阶段”的担忧,重申 SpaceX 的AI计算投资回收期**不到一年**。这使得他们可以“无限投入资金”于计算能力并迅速扩大规模。 * 再次强调,SpaceX 设定的到2030年实现1万亿美元营收的目标和预期,主要由AI计算租赁业务支撑。 * 他的模型表明,在未来二十年里,AI模型(Grok)对总收入和利润的贡献将远低于AI计算租赁。 * 预测一旦投资者充分理解对AI智能和计算的“无限需求”,他们就会认识到AI计算是 SpaceX 的主要驱动力。

The video discusses SpaceX's first quarterly earnings as a publicly traded company, focusing on its potential to become the world's first $10 trillion company. **Current Valuation Context:** * Apple, Alphabet, and Nvidia are around $5 trillion today. * SpaceX is under $1.5 trillion, meaning it needs to increase its valuation roughly seven times to reach $10 trillion, while the others only need to double. **Peter Diamandis (Early Investor in SpaceX, Tesla, Google; Founder of X Prize) on SpaceX's Potential:** * Believes Elon Musk can make SpaceX the first $10 trillion company, building "civilizational-level infrastructure." * Argues the market undervalues SpaceX because it's four distinct businesses vertically integrated. * **Business 1: Global Communications (Starlink):** * Delivering global gigabit connectivity. * Direct satellite to cell phone service has a $600 billion revenue potential, according to Gwynne Shotwell on the earnings call. SpaceX expects to capture a significant chunk of this. * **Business 2: Hyperscaler (AI Compute Leasing):** * Already signed $20 billion in annualized recurring revenue (ARR) for this business. * Building compute infrastructure faster than competitors. * Announced plans for 2 gigawatts (GW) of compute by the end of this year, aiming for 10 GW by next year. * Expected to be the greatest source of revenue. * Beyond earth, they are building orbital compute. * Elon Musk stated they will launch their first AI satellites in partnership with Nvidia in 2027, much faster than anticipated. SpaceX has sought permission to launch 1 million such satellites. **"Solving the Money Problem" Host's Commentary on AI Compute:** * Emphasizes AI compute as the most critical opportunity for SpaceX ("AI compute motherfuckers"). * Demand for intelligence and AI compute is "near infinite" for decades to come, for both training and inference. * SpaceX expects to be at an ARR of $100 billion-plus by December this year, primarily from new AI compute deals. * The payback period for SpaceX's AI compute investments is *under one year*, making it effectively a cost of goods sold rather than traditional capital expenditure (capex). * Companies like Google and Anthropic are reportedly paying SpaceX "billions of dollars a month collectively" for compute access because SpaceX can deploy it faster and cheaper. * His 20-year valuation model for Tesla and SpaceX shows AI compute leasing contributing the majority of revenue and profits over the long term, far exceeding Starlink, Grok (AI models), or launch. **Dan Ives (Wedbush) on Capex and Investor Patience:** * States that the capex required for SpaceX's build-out is "necessary" but will "require patience" from investors. * Describes the situation as an "arms race" where heavy capex is unavoidable. * Agrees that "the gold at the end of the rainbow is around data AI" and that this vision extends to XAI and potentially Tesla. * Acknowledges that costs for chips and memory are rising, and SpaceX is one of many companies vying for these resources. * When questioned why investors should choose SpaceX over established AI players (Anthropic/OpenAI) or cloud providers (Microsoft/Amazon), Ives suggests SpaceX is currently "playing catch up" and investors are buying into Musk's vision. * He highlights that the "price of capital" is increasing, and while companies know lightening up on capex means falling behind, investors will eventually demand to see monetization. * He contrasts SpaceX with companies like Microsoft and Alphabet, which are already showing monetization from their investments. **"Solving the Money Problem" Host's Conclusion:** * Rebuts Dan Ives's concern about a "difficult phase" due to capex by reiterating SpaceX's *less-than-one-year payback period* on AI compute. This allows them to "spend infinite money" on compute and rapidly scale. * Reaffirms that SpaceX's goal and expectation of achieving $1 trillion in revenue by 2030 is heavily underpinned by AI compute leasing. * Illustrates with his model that AI models (Grok) will contribute far less to total revenue and profits than AI compute leasing over the next two decades. * Predicts that once investors fully grasp the "infinite demand" for AI intelligence and compute, they will recognize AI compute as SpaceX's primary driver.