Alphabet’s New Model & Novo’s Lawsuit

以下是内容的中文翻译： 《Motley Fool与Gems Investing》的最新一期节目，由特拉维斯·霍伊主持，卢·怀特曼和约翰·夸斯特共同参与，深入探讨了财报季开始之际科技、制药和能源领域的重要进展。 **Alphabet的Gemini 3.6 Flash：褒贬不一的评价** 讨论始于Alphabet最近发布的Gemini 3.6 Flash，这个AI模型受到了“AI专家圈”褒贬不一的评价。主要的失望源于备受期待的“超强”或前沿模型（Gemini 3.5 Pro）的延迟发布，高消费用户此前一直对此抱有期待。 卢·怀特曼从“普通大众”的角度出发，认为尽管尖端前沿模型主导着社交媒体的讨论，但它们价格昂贵，且只有少数（不到10%）全球AI用户使用。他认为未来大部分收入可能来自更经济实惠、更实用的AI解决方案，这使得Flash的效率（少17%的tokens，成本更低）对“实际决策者”具有潜在吸引力。然而，他承认Pro的延迟发布表明谷歌在使其按预期运行方面遇到了挑战，这标志着该公司遭遇了挫折。 约翰·夸斯特也表达了对惹恼高消费用户的担忧，他们支付了Pro订阅费，期待在六月获得新模型。他强调了Flash版本在某些情况下甚至能超越当前顶级Pro模型的讽刺之处，这表明可能存在战略失误。特拉维斯·霍伊思考这是否标志着谷歌等公司进入了一个成熟阶段，即专注于为普通用户和AI概览提供更快、更便宜、更广泛适用的Flash模型，而不仅仅是追求最前沿的技术。尽管眼前令人失望，但小组成员一致认为AI仍处于早期阶段，仅凭单一公告做出长期预测是困难的。 **GLP-1之战：诺和诺德起诉礼来** 话题随后转向制药界，具体来说是GLP-1减肥药市场日益激烈的竞争。诺和诺德（Ozempic和Wegovy的生产商）已就其营销行为起诉礼来公司（Mounjaro和Zepbound的生产商）。 约翰·夸斯特认为这起诉讼是诺和诺德“绝望”的表现。他指出，虽然诺和诺德的品牌知名度可能更高，但礼来公司占据了更大的预计市场份额（60%），并预计今年销售额将增长，而诺和诺德则预计会下降。随着价格压力挤压利润空间，销量和市场份额至关重要，而礼来公司似乎正在取得进展。 卢·怀特曼认为两家公司都有道理：礼来公司准确引用了对比试验，但省略了数据已过时且诺和诺德的新产品表现更好的事实。他将此形容为诺和诺德正在争夺一座“巨大的金矿”，并断言“最新型号就是最佳型号”的逻辑也适用于此。特拉维斯·霍伊观察到，GLP-1市场正迅速从高成本的专科治疗转变为一个更受消费者驱动、以效果和价格为首要考量的领域，这种转变是制药公司所不习惯的。这起诉讼，以及诺和诺德迅速解决与Hims & Hers的诉讼，都强调了这一竞争激烈的新阶段。 估值数据显示，诺和诺德的交易价格为销售额的3.7倍和预期市盈率的16倍，反映了市场对其增长放缓的预期。相比之下，礼来公司目前的交易价格接近销售额的15倍和预期市盈率的32倍，表明市场对其增长轨迹充满信心，这也使其能够进行战略性收购。 **GE Vernova：数据中心推动涡轮机需求** 最后，小组成员讨论了GE Vernova的财报，该公司最近从通用电气（GE）剥离出来。尽管对其公司名称有所保留，但其业绩强劲，自由现金流指引几乎翻倍。令人惊讶的驱动因素是什么？不是绿色能源（这表面上是Vernova的业务重点），而是天然气涡轮机，主要原因是数据中心的需求飙升。 卢·怀特曼指出，业绩强劲反而导致股价下跌的悖论，引发了对数据中心需求可持续性的质疑。尽管增长巨大，但投资者正在辩论其是会继续下去还是会趋于平稳。Vernova正在通过大幅增加资本支出（四年内投资110亿美元，而此前为10亿美元）来应对，以提高产能，特别是涡轮机产能。 约翰·夸斯特强调了令人印象深刻的积压订单：目前为1760亿美元，预计明年将达到2000亿美元，而年收入仅为400亿美元。这表明未来需求强劲，但也存在风险，如果需求正常化，扩大的运营基础可能会成为负担。 特拉维斯·霍伊指出该股自2024年3月以来表现出色，上涨了665%。然而，卢·怀特曼警告不要在当前市场采纳“淘金铲”投资理论。历史上，投资供应商（如为数据中心提供能源的Vernova）是廉价地接触热门趋势的一种方式。但Vernova的交易价格是预期收益的65倍（远高于Alphabet或微软等“超大规模企业”），这些“淘金铲”目前被高估了。他总结说，驱动趋势的核心公司通常比其高估值的供应商更具吸引力。

The latest episode of Motley Fool and Gems Investing, hosted by Travis Hoy with Lou Whiteman and John Quast, delved into significant developments across the tech, pharma, and energy sectors as earnings season begins. **Alphabet's Gemini 3.6 Flash: A Mixed Reception** The discussion kicked off with Alphabet's recent release of Gemini 3.6 Flash, an AI model that received lukewarm reviews from "AI pundit class." The primary disappointment stemmed from the delay of the highly anticipated "super duper" or frontier model (Gemini 3.5 Pro), which high-spending users had expected. Lou Whiteman offered a "normie town" perspective, arguing that while cutting-edge frontier models dominate social media discourse, they are expensive and used by a minority (less than 10%) of global AI users. He suggested that the bulk of future revenue would likely come from more affordable and rational AI solutions, making Flash's efficiency (17% fewer tokens, lower cost) potentially appealing to "actual decision-makers." He acknowledged, however, that the delay of Pro indicates Google faced challenges getting it to work as intended, marking a setback for the company. John Quast echoed concerns about angering high-spend users who paid for Pro subscriptions expecting the new model in June. He highlighted the irony that the Flash version can, in some instances, outperform the current top-tier Pro model, indicating a potential strategic misstep. Travis Hoy pondered if this signifies a maturation phase for companies like Google, focusing on faster, cheaper, and broadly applicable models like Flash for general users and AI overviews, rather than just the bleeding edge. Despite the immediate disappointment, the panel agreed it's still early days in AI, making long-term predictions from single announcements difficult. **GLP-1 Battle: Novo Nordisk Sues Eli Lilly** The conversation then shifted to the pharmaceutical world, specifically the growing rivalry in the GLP-1 weight-loss drug market. Novo Nordisk, makers of Ozempic and Wegovy, has filed a lawsuit against Eli Lilly, producers of Mounjaro and Zepbound, over their marketing practices. John Quast viewed the lawsuit as a sign of Novo Nordisk's "desperation." He pointed out that while Novo Nordisk might have better brand recognition, Eli Lilly holds a larger estimated market share (60%) and projects increased sales this year, while Novo Nordisk anticipates a decline. With pricing pressures squeezing profit margins, sales volume and market share are crucial, and Eli Lilly appears to be gaining ground. Lou Whiteman suggested both companies have valid points: Lilly accurately cites comparison trials but omits that the data is dated and newer Novo Nordisk products perform better. He characterized it as Novo Nordisk fighting for a "huge gold mine," asserting that the "latest model is the best model" dynamic applies here too. Travis Hoy observed that the GLP-1 market is rapidly transforming from high-cost specialty treatments to a more consumer-driven space where results and price are paramount, a shift that pharma companies are unaccustomed to. This lawsuit, following Novo Nordisk's quick resolution of a suit against Hims & Hers, underscores this new, highly competitive phase. A look at valuations showed Novo Nordisk trading at 3.7 times sales and 16 times forward earnings, reflecting market expectations of slower growth. Eli Lilly, conversely, trades at nearly 15 times sales and 32 times forward estimates, indicating market confidence in its growth trajectory, which has allowed it to pursue strategic acquisitions. **GE Vernova: Data Centers Fueling Turbine Demand** Finally, the panel discussed GE Vernova's earnings report, a company recently spun off from GE. Despite reservations about the company's name, its results were strong, with free cash flow guidance almost doubling. The surprising driver? Not green energy (which Vernova is ostensibly about), but natural gas turbines, primarily due to soaring demand from data centers. Lou Whiteman noted the paradox of a strong quarter leading to a stock dip, raising questions about the sustainability of data center demand. While growth has been immense, investors are debating if it will continue or plateau. Vernova is responding by dramatically increasing its capital expenditure (an $11 billion investment over four years, compared to $1 billion previously) to boost capacity, particularly for turbines. John Quast highlighted the impressive backlog: $176 billion currently, projected to reach $200 billion next year, against annual revenues of only $40 billion. This indicates strong future demand, but also a risk if demand normalizes, as the increased operating base could become a liability. Travis Hoy noted the stock's remarkable performance, up 665% since March 2024. Lou Whiteman, however, cautioned against the "picks and shovels" investment thesis in the current market. Historically, investing in suppliers (like Vernova for data center energy) was a way to gain exposure to a hot trend cheaply. But with Vernova trading at 65 times expected earnings (significantly higher than "hyperscalers" like Alphabet or Microsoft), these "picks and shovels" are currently overpriced. He concluded that the core companies driving the trends often present a more appealing value than their highly-valued suppliers.