The Motley Fool Hidden Gems Investing播客，由Travis Hoy、Lou Whiteman和Tyler Crow主持，最近讨论了Muse，该应用被描述为自ChatGPT以来最大的新应用发布。Muse在应用商店榜首已近一周，这促使Meta的股价“本周暴涨”，市值增加了数千亿美元。然而，并非所有大型科技公司都欢迎Muse；例如，亚马逊已阻止其抓取其网站内容，这表明公司之间正从过去的合作转向“追逐相同的东西”。 Lou Whiteman指出，现在断言Muse的影响还为时过早，但他观察到亚马逊的抵制表明Muse正试图获取他们的一些“秘诀”。Tyler Crow提出，Meta推出Muse（提供慷慨的免费令牌额度，在收费层级之前有1亿个）的主要目标是“获取更多信息”，特别是购物数据，以补充其社交媒体帝国。他担忧亚马逊和Google/Alphabet等“围墙花园”保护其广告数据，可能导致用户体验碎片化，届时个人需要多个AI代理，类似于当前的“流媒体服务重蹈覆辙”的困境。 讨论触及Muse是否可能颠覆以广告为中心的商业模式，例如亚马逊（其超过100%的利润来自广告）或Meta的社交网络广告。Travis Hoy质疑Muse是代表着根本性颠覆还是“缓慢燃烧”。Lou Whiteman倾向于“缓慢燃烧”的观点，他引用了AI预测的历史，并强调Muse“之所以是新的热门事物，是因为它今天表现更好”，但未来的竞争者很可能会出现。Tyler Crow补充说，技术进步的快速和成本的下降意味着“改变游戏规则的时刻”正变得商品化，这使得任何单一创新都难以作为关键突破而脱颖而出。 尽管亚马逊采取敌对立场，Shopify、PayPal和Expedia等公司已与Muse合作，将其视为低风险机会。Tyler Crow认为这些合作有益，但对于这些公司的投资论点而言，并非“改变游戏规则”。Lou Whiteman警告不要高估这些早期合作的重要性，将它们比作可退还的押金，这意味着合作伙伴“几乎没有不利影响”，但长期保障有限。 对话随后探讨了中间商可能受到的颠覆。Travis Hoy思考Expedia（一个合作伙伴）是否容易受到攻击，因为智能AI可以直接访问酒店/航空公司网站，而Uber（一个潜在目标）可能因需要人类司机而受到更多保护。然而，Lou Whiteman反驳说，AI代理的承诺正在“摧毁中间商”。他提出，从长远来看，Uber司机也可能被激励使用新系统来规避Uber的费用，因此“在假设一个中间商成功而另一个不成功时要非常小心”。 关于“隐藏机会”，Tyler Crow对代理式AI立即影响个人消费者“可支配支出”表示怀疑，他指出人们常凭一时冲动购物或重视亲身体验。他认为代理式商务的**B2B方面**有更大的潜力，因为它具有定期的采购计划和数字化的库存。他引用了**AAR Corp.**作为例子——该公司正在将其ERP系统整合代理式AI，用于商业和军用飞机零部件分销，可以自动化航空公司的库存规划。他认为，与消费领域的“激烈竞争”相比，这种“小众”B2B应用将产生“巨大价值”。Lou Whiteman表示同意，他引用智能冰箱用于购物清单的缓慢普及作为消费者不愿完全自动化的证据。 最后，Travis Hoy质疑Google凭借其Android操作系统的Sdominance，是否能“快速跟进”Muse。Lou Whiteman回应称，虽然Google *可以*凭借现有联系进行复制并获胜，“但几乎任何人都可以复制这个”，而且“Alphabet或Meta之外的某个公司会做得更好。也许是Apple。”他总结说，五年后谁是最终赢家是无法预测的，投资者“不应过度解读今天，也不应妄下任何断言。”

The Motley Fool Hidden Gems Investing podcast, featuring Travis Hoy, Lou Whiteman, and Tyler Crow, recently discussed Muse, described as the biggest new app introduction since ChatGPT. Muse has been number one in app stores for nearly a week, contributing to Meta's stock "exploding this week" and adding hundreds of billions in value. However, not all big tech companies are welcoming Muse; Amazon, for instance, has blocked it from crawling its website, signaling a shift from past collaborations as companies now "chase the same thing." Lou Whiteman noted that it's too early to definitively assess Muse's impact but observed that Amazon's resistance indicates Muse is trying to take some of their "secret sauce." Tyler Crow suggested Meta's primary goal with Muse, offered with generous free token limits (100 million before pay tiers), is to "capture more information," particularly shopping data, to supplement its social media empire. He raised concerns about "walled gardens" like Amazon and Google/Alphabet protecting their advertising data, potentially leading to a fragmented user experience where individuals need multiple AI agents, akin to the current "streaming services all over again" dilemma. The discussion touched on whether Muse could disrupt ad-centric business models like Amazon's (which derives over 100% of its profit from advertising) or Meta's social network ads. Travis Hoy questioned if Muse represents a fundamental disruption or a "slow burn." Lou Whiteman favored the "slow burn" view, citing the history of AI predictions and emphasizing that Muse is "the new big thing because it is better today," but future contenders are likely. Tyler Crow added that the rapid pace of technological advancements and falling costs mean that "game changing moments" are becoming commoditized, making it harder for any single innovation to stand out as pivotal. While Amazon takes an adversarial stance, companies like Shopify, PayPal, and Expedia have partnered with Muse, viewing it as a low-risk opportunity. Tyler Crow deemed these partnerships beneficial but not "game changing" for the companies' investment theses. Lou Whiteman cautioned against overestimating the significance of these early partnerships, comparing them to refundable deposits, implying "very little downside" for partners, but limited long-term guarantees. The conversation then explored the potential for disruption to middlemen. Travis Hoy pondered if Expedia (a partner) is vulnerable because smart AI could go directly to hotel/airline websites, whereas Uber (a potential target) might be more protected due to the need for human drivers. Lou Whiteman, however, pushed back, arguing that the promise of AI agents is "destroying the middlemen." He suggested that, long term, Uber drivers could also be incentivized to use new systems to bypass Uber's fees, so "be very careful assuming one middleman makes it and another doesn't." Regarding "hidden opportunities," Tyler Crow expressed doubts about agentic AI's immediate impact on individual consumer "discretionary spending," noting that people often buy on whim or value in-person experiences. He sees greater potential in the **B2B side** of agentic commerce due to regular purchasing schedules and digitized inventories. He cited **AAR Corp.** as an example—a company building agentic AI into its ERP systems for commercial and military aircraft parts distribution, which can automate inventory planning for airlines. This "niche" B2B application, he argued, will generate "immense value" compared to the "knife fight" in the consumer space. Lou Whiteman agreed, referencing the slow adoption of smart refrigerators for grocery lists as evidence of consumer reluctance to fully automate. Finally, Travis Hoy questioned if Google, with its Android operating system dominance, could "fast follow" Muse. Lou Whiteman responded that while Google *could* copy and win due to existing connections, "almost anyone can copy this," and "someone other than Alphabet or Meta will simply do it better. Maybe Apple." He concluded that it's impossible to know the ultimate winner in five years, and investors "shouldn't read too much into today and we shouldn't assume anything."