以下是内容的中文翻译： 播客节目《Motley Fool Hidden Gems Investing》开场时，主持人特拉维斯·霍利（Travis Holy）、卢·怀特曼（Lou Whiteman）和约翰·夸斯特（John Quast）对特斯拉最近的“机器人出租车”（robotaxi）活动表达了困惑和怀疑。尽管此前进行了数周的炒作和比赛，据报道该活动没有直播，没有外部邀请，值得注意的是，埃隆·马斯克（Elon Musk）本人并未出席，仅在社交媒体上进行了少量互动。卢质疑了如此低调的宣传活动的目的，认为这本可以只是一封电子邮件或一份新闻稿。 约翰·夸斯特提出了一个更乐观的看法，认为考虑到这只是一个缓慢的部署过程而非全面发布，低调推出是合理的。他认为，避免过度炒作有助于管理预期并防止失望。然而，特拉维斯反驳说，埃隆·马斯克有过度炒作的历史，指出机器人出租车的讨论始于2018-2019年，而最初预测的2024年推出是在两年前。他提到，虽然特斯拉预计在自动驾驶汽车领域拥有显著的先行者和成本优势，但Waymo等竞争对手正在扩大车队规模并大幅削减硬件成本，可能削弱特斯拉的优势。卢补充说，PowerPoint演示文稿已经无法再打动投资者；真正的大规模部署数千辆汽车才能做到。 讨论随后转向了OpenAI宣布推出GPT-6，该产品已开始向企业客户推出。主持人，特别是特拉维斯，表示失望的是，OpenAI领导层所宣扬的“AGI时代”（通用人工智能）似乎归结为一种能够“填写在线表格、更新客户记录……以及安排日程”的机器人。约翰持怀疑态度，认为如果这真的是AGI，“我们看到了就会知道”，并暗示这可能是一种IPO前炒作策略。两人都认为，OpenAI、谷歌（通过Gemini）和Meta对企业客户的关注表明了真正的利润所在，但质疑如果所有公司都瞄准相同的客户，能有多少收入。卢指出，许多企业正在选择“足够好”的更便宜模型，而不是尖端的前沿模型，以优化成本，这一趋势在Duolingo和Uber等公司中已有所体现。 此外，还讨论了Adobe最近的CEO过渡，市场反应负面，导致股价下跌。卢解释说，即将离任的CEO在任职20年后退休，这是一次常规的内部晋升。然而，约翰指出，负责Adobe最创收产品（Photoshop、Premiere、Acrobat、Firefly）的高管大卫·瓦德瓦尼（David Wadwani）被越过，现在正离开公司，寻求一个“全新开始”，在一个“技术改变……公司构建方式”的领域。约翰怀疑瓦德瓦尼可能正在前往一家AI公司与Adobe竞争，这引发了人们对在潜在颠覆性环境中出现“看守CEO”的担忧。 在零售新闻方面，Lululemon报告了糟糕的财报，同店销售额下降5%，导致其股价大幅下跌。约翰认为这是“Lululemon自身的问题”，而非广泛的零售业低迷，他引用了行业内其他品牌如American Eagle、Abercrombie和Yeti强劲的服装和体育用品销售数据。卢对此表示认同，指出“热门零售是危险的”，当潮流零售商的热度消退时，很难重新提振。 主持人随后对未来十年进行了预测： * **自动驾驶里程（2036）：** 卢预测，鉴于汽车行业商品化的历史趋势，商品化技术提供商（如Mobileye或英伟达NVIDIA）或Waymo将领先，而非特斯拉。然而，约翰认为，如果监管障碍得以清除，特斯拉将拥有最多的里程数，这得益于其制造规模以及潜在的车辆补贴能力。他们还讨论了个人汽车所有权的未来。 * **最佳AI模型构建者：** 卢预测，谷歌和Meta等老牌多元化公司将成为主导的AI供应商，因为他怀疑Anthropic和OpenAI等独立的、尖端公司在没有其他收入来源的情况下能否在财务上维持自身。约翰认为，由于持续的迭代，“最佳AI模型”这个问题本身可能会变得无关紧要，但他承认，那些使其模型更难被“提取”的公司可能会扩大其领先优势。 * **购物的未来：** 两位主持人都对AI演示中展示的“机器人购物”愿景表示怀疑，提及苹果公司在2012年类似的未兑现承诺。卢认为，购物将经历“渐进式变化”，AI将充当搜索向导，而非完整的采购代理。约翰承认“代理式商务”（agentic commerce）是未来，但他强调了支付和金融科技公司的作用，认为“网站不再是任何竞争优势”。 最后，在“我们关注的股票”环节： * **约翰推荐了Reddit (RDDT)**，理由是其用户增长超过60%，通过Reddit Max增加变现能力，财务状况改善（运营现金流翻倍，净利润率30%），无债务，以及20倍远期市盈率的吸引力估值。 * **卢推荐了CECO Environmental (CECO)**，一家提供空气质量、废水管理和能源转型产品的公司。他强调该公司涉足电动汽车电池和半导体制造等高增长市场，拥有18亿美元的订单积压，并有望实现30亿美元的年订单额，尽管其市盈率接近30倍。 丹·博伊德（Dan Boyd）将CECO Environmental列入了他的观察名单。

The podcast "Motley Fool Hidden Gems Investing" opened with hosts Travis Holy, Lou Whiteman, and John Quast expressing confusion and skepticism about Tesla's recent robotaxi event. Despite weeks of hype and contests, the event reportedly had no live stream, no outside invitations, and notably, Elon Musk did not attend in person, only offering minimal social media engagement. Lou questioned the purpose of such a low-profile publicity event, suggesting it could have been an email or press release. John Quast offered a more bullish perspective, arguing that the quiet rollout made sense given it's a slow deployment rather than a full launch. He posited that avoiding over-hyping would manage expectations and prevent disappointment. However, Travis countered that Elon Musk has a history of over-hyping, pointing out that robotaxi discussions began in 2018-2019 and the original 2024 launch prediction was two years ago. He noted that while Tesla was expected to have a significant first-mover and cost advantage in autonomous vehicles, competitors like Waymo are scaling their fleets and cutting hardware costs dramatically, potentially eroding Tesla's edge. Lou added that a PowerPoint presentation won't impress investors anymore; actual deployment of thousands of vehicles in force will. The discussion then shifted to OpenAI's announcement of GPT-6, which is starting to roll out to enterprise customers. The hosts, particularly Travis, expressed disappointment that the "AGI era" (Artificial General Intelligence) being touted by OpenAI's leadership seemed to boil down to a bot capable of "filling out online forms, updating customer records... and organizing your calendar." John was cynical, suggesting that if it were truly AGI, "we're going to know it when we see it" and implying it might be a pre-IPO hype strategy. Both agreed that the focus on enterprise customers by OpenAI, Google (with Gemini), and Meta indicates where the real money lies, but questioned how much revenue there is if all companies target the same customers. Lou noted that many enterprises are opting for "good enough" cheaper models over bleeding-edge frontier models to optimize costs, a trend seen with companies like Duolingo and Uber. Adobe's recent CEO transition was also discussed, with the market reacting negatively, sending the stock down. Lou explained that the outgoing CEO retired after 20 years, making it a routine internal promotion. However, John pointed out that David Wadwani, the executive in charge of Adobe's most revenue-generating products (Photoshop, Premiere, Acrobat, Firefly), was passed over and is now leaving the company to pursue a "fresh start" where "technology changes... how companies are built." John suspected Wadwani might be headed to an AI company to compete with Adobe, raising concerns about a "caretaker CEO" in a potentially disruptive environment. In retail news, Lululemon reported poor earnings with a 5% drop in same-store sales, causing its stock to fall significantly. John argued it was a "Lululemon problem" rather than a broad retail downturn, citing strong clothing and sporting goods sales across the industry from other brands like American Eagle, Abercrombie, and Yeti. Lou echoed this, stating that "hot retail is dangerous" and that when the air comes out of the balloon for trendy retailers, it's hard to reinflate. The hosts then engaged in 10-year predictions: * **Autonomous Drive Miles (2036):** Lou predicted that a commoditized technology provider (like Mobileye or NVIDIA) or Waymo would lead, not Tesla, due to the historical trend of commoditization in the auto industry. John, however, believes Tesla would have the most miles if regulatory hurdles are cleared, thanks to its manufacturing scale and ability to potentially subsidize vehicles. They also debated the future of individual car ownership. * **Best AI Model Builder:** Lou predicted that established, diversified companies like Google and Meta would be the dominant AI vendors, as he doubted independent, bleeding-edge companies like Anthropic and OpenAI could sustain themselves financially without other revenue streams. John suggested the "best AI model" question itself might become irrelevant due to constant iteration, but conceded that companies making models harder to distill could widen their lead. * **Future of Shopping:** Both hosts expressed skepticism about the "bot making purchases" vision shown in AI demos, citing Apple's similar unfulfilled promises from 2012. Lou believes shopping will see "incremental change" with AI as a search guide, not a full purchase agent. John acknowledged "agentic commerce" as the future but emphasized the role of payment and financial technology companies, suggesting that the "website is not anything of a competitive distinctive anymore." Finally, for the "Stock on Our Radar" segment: * **John recommended Reddit (RDDT)**, citing its 60%+ user growth, increasing monetization through Reddit Max, improving financials (doubling operating cash flow, 30% net margin), no debt, and an attractive 20x forward earnings valuation. * **Lou recommended CECO Environmental (CECO)**, a company providing air quality, wastewater management, and energy transition products. He highlighted its involvement in high-growth markets like EV batteries and semiconductor manufacturing, a $1.8 billion backlog, and a path to $3 billion in annual orders, despite trading at almost 30 times earnings. Dan Boyd chose CECO Environmental for his watch list.