AI News AI资讯 5h ago Updated 1h ago 更新于 1小时前 53

Airbnb’s Brian Chesky plans to launch a new AI lab Airbnb的布莱恩·切斯基计划推出新的AI实验室

Brian Chesky is done advising. The Airbnb co-founder, who quietly became Silicon Valley’s most influential AI whisperer—brokering Sam Altman’s return, offering counsel on hypergrowth, and reportedly being considered for OpenAI’s own board—is now putting his own money and reputation on the line. He’s launching an AI lab. This isn’t just another billionaire’s vanity project; it’s a direct challenge to the very models he helped stabilize. The move signals a profound shift in the AI power dynamic: f 布莱恩·切斯基终于不耐烦了。这位Airbnb的联合创始人兼CEO,过去一年半一直扮演着科技圈里最讨巧的角色——站在聚光灯旁,既是OpenAI的亲密盟友和非官方顾问,又是最早一批将AI融入产品实践的“应用派”领袖。现在,他显然厌倦了仅仅做个“AI王者制造者”,他要亲自下场,建立自己的AI实验室。消息一出,硅谷权力版图上又添了一道耐人寻味的裂痕。

80
Hot 热度
70
Quality 质量
75
Impact 影响力

Analysis 深度分析

Brian Chesky is done advising. The Airbnb co-founder, who quietly became Silicon Valley’s most influential AI whisperer—brokering Sam Altman’s return, offering counsel on hypergrowth, and reportedly being considered for OpenAI’s own board—is now putting his own money and reputation on the line. He’s launching an AI lab. This isn’t just another billionaire’s vanity project; it’s a direct challenge to the very models he helped stabilize. The move signals a profound shift in the AI power dynamic: from the era of the central, godlike frontier lab to an age of application-native intelligence, built by people who understand what users actually do with a product, not just what a benchmark says they can do.

The facts are straightforward. Chesky has grown disillusioned with the off-the-shelf offerings from OpenAI, Anthropic, and Google. Airbnb has used AI coding tools, but Chesky publicly stated last year that the big LLM partnerships weren’t “quite ready.” That’s a polite way of saying the general-purpose chatbots aren’t deeply useful for the complex, trust-based marketplace that is Airbnb. They can write a listing description, sure, but can they negotiate a cancellation, predict local regulatory shifts, or create a personalized travel narrative that feels human and not algorithmic? The answer is a resounding no. The current models are brilliant parrots; Chesky needs a seasoned concierge.

His lab’s rumored focus—user interaction and design—is the tell. This isn’t about chasing the next incremental GPT benchmark or training a model to pass the bar exam. It’s about the interface, the feel, the utility. This is where Chesky’s genius lies. He turned a website for air mattresses into a global hospitality icon by obsessing over trust, photography, and seamless user journeys. Now, he’s applying that same ethos to AI. He’s betting the next giant leap isn’t a smarter brain, but a smarter way for that brain to talk to us. He’s looking at the clunky, text-box-and-response paradigm dominating AI assistants and seeing a 1990s web portal: functional, but waiting for its iPhone moment.

This puts him in direct competition with the labs he advised. It’s a classic Valley narrative—the apprentice outgrows the mentor’s shadow—but with high stakes. Altman’s OpenAI is building the foundational infrastructure, the “operating system” for AI. Chesky is building what runs on top of it, or perhaps, a wholly different stack optimized for a specific experience. Think of it this way: OpenAI is making the most powerful microchips in the world. Chesky isn’t trying to make a faster chip; he’s designing the sleek, intuitive smartphone that makes the chip’s power actually matter to a human being. The battlefield is shifting from the raw capability of the model to the elegance and efficacy of its deployment.

The comparison to Brett Adcock’s Hark lab is inevitable but also instructive. Adcock, the founder of Figure AI, is chasing embodied AI and novel interfaces with a hardware component. Chesky is staying in his lane—the digital experience—but applying the same ambition. Both recognize that the current AI interaction model is a dead end for mainstream adoption. The future isn’t a chatbot you query; it’s an intelligence woven into the fabric of an app, anticipating needs and acting with context. Chesky’s advantage is he has a billion-dollar platform—Airbnb—as a potential test bed and distribution channel from day one.

Critics will say this is just another case of a tech CEO diversifying into the hotness of the moment, a hedge against AI upending his core business. That misses the point. This is a creative’s counterattack. Chesky is a designer by training. The current AI paradigm is, at its heart, a design failure. It’s a text interface forced onto a multi-modal world. His lab is a declaration that the most important problem in AI isn’t one of scale, but of translation—translating immense computational power into intuitive human utility. He’s betting that the companies that win the next decade won’t be the ones with the largest training runs, but the ones with the most intuitive, indispensable AI layers.

The risks are monumental. Building frontier AI is a capital-intensive, research-heavy endeavor that makes building a two-sided marketplace look like a lemonade stand. Chesky is entering a gladiatorial arena with trillion-dollar opponents. But he isn’t trying to out-Google Google. He’s trying to out-Apple Apple in the AI interface space. He’s leveraging his unique perspective—the kingmaker who saw the emperor’s new clothes up close—to build something different. The most telling detail isn’t the lab itself, but the narrative of its founding: the advisor, seeing the limits of the advice, rolling up his sleeves. In the gold rush of AI, the most profound insight might be that the real value isn’t in digging up the gold, but in designing the better, more intuitive wallet. Brian Chesky is betting his legacy on that idea, and frankly, the AI establishment should be nervous. The next great leap might not come from a lab in San Francisco chasing AGI, but from a design studio building an AI that finally feels like it was made for us.

布莱恩·切斯基终于不耐烦了。这位Airbnb的联合创始人兼CEO,过去一年半一直扮演着科技圈里最讨巧的角色——站在聚光灯旁,既是OpenAI的亲密盟友和非官方顾问,又是最早一批将AI融入产品实践的“应用派”领袖。现在,他显然厌倦了仅仅做个“AI王者制造者”,他要亲自下场,建立自己的AI实验室。消息一出,硅谷权力版图上又添了一道耐人寻味的裂痕。

这绝非一时兴起。切斯基对当前前沿实验室的“出品”,显然憋着一肚子意见。他去年就公开表态,Airbnb没有与任何大模型厂商达成深度合作,理由很直白:现有产品“还不够成熟”。这句话从一家全球顶级平台的CEO嘴里说出来,分量不轻。它不仅是一个商业决策的解释,更是一记敲打:你们这些估值千亿、烧掉数十亿美金的实验室,交出的答卷,在我这个挑剔的、以体验为生命线的产品经理眼里,还差点意思。那种“还不够好”的不满,如今转化成了“不如我自己来”的行动。

切斯基的底气,一半来自他多年练就的用户洞察力,另一半,则来自他那张硅谷顶级的社交网络门票。他和山姆·奥特曼的交情始于2006年的Y Combinator,那是创业公司神话的起点。当OpenAI崛起,切斯基成了奥特曼管理这家“超高速增长”公司的非正式导师。最戏剧性的一幕发生在去年奥特曼被董事会闪电解雇的风波中,切斯基充当了关键的斡旋者和“救火队长”,利用他在圈内的人脉和公关能力,为奥特曼的回归铺平了道路。他甚至一度是OpenAI董事会的潜在人选。从某种意义上说,他是OpenAI这个“巨兽”成长的关键见证者与助推者之一。

所以,这场“恩主”变“对手”的戏码才显得格外微妙和精彩。切斯基并非从零开始的愣头青,他带着对AI应用层深刻的失望、对用户体验的偏执,以及最了解竞争对手(至少是盟友)内部运作的优势入场。他的新实验室,方向已露出端倪:用户交互与设计。这恰恰是切斯基在Airbnb打造帝国的看家本领,也是当前大模型产品最普遍的短板。当所有实验室都在比拼参数规模、推理能力和基准测试分数时,切斯基很可能从另一个维度发起攻击:不追求“最聪明”,而追求“最好用、最贴心、最自然”。这就像当所有厨子都在争论哪种钢材的刀最锋利时,有个人站出来说,我们来做一把刀柄握感最舒服、最符合人体工学的刀。

这想法并不孤例,类似前Figure AI创始人Brett Adcock的Hark实验室,也在探索AI助手的新型交互范式。但切斯基的不同在于,他手握Airbnb这个庞大的真实商业场景和数亿用户的使用数据。他的实验室很可能不是纯粹的理论研究机构,而是一个快速原型、快速验证的“产品兵工厂”。它可能不会去训练一个从头开始的、与GPT或Claude正面抗衡的基础大模型,而是致力于打造一系列为特定场景高度优化的、具有独特“交互智能”的AI模块或系统。这是一种“降维打击”:用产品哲学和场景理解的深度,来弥补基础研究资源的相对不足。

当然,我们也要以一丝怀疑的眼光看待这场新战役。硅谷从不缺由成功创业者发起的、雄心勃勃的“复仇者联盟”项目,但成功率往往不高。切斯基的核心能力无疑是打造伟大的消费级产品,但AI实验室是一场需要巨额资本、顶尖研究人才和长期技术耐心的“重工业”竞赛。他的优势在于应用和设计,但基础模型的底层突破,依然是另一个战场的游戏。他能否吸引到世界顶尖的AI研究科学家?他愿意为长期的、可能没有即时产品回报的研究投入多少真金白银?Airbnb的股东是否会为他这个可能分散精力的“新玩具”持续买单?

更有趣的是一种心态的转变。切斯基曾是那个告诉别人“如何管理一家狂飙突进的科技公司”的导师,如今,他选择把自己抛入AI这个最不确定、竞争最惨烈的新战场,成为与自己学生直接竞争的选手。这其中既有科技领袖永恒的进取心和不安全感,也暗含着对现有AI巨头们“傲慢”与“迟钝”的一种挑衅。当奥特曼们在白宫和世界各地斡旋AI监管、畅谈通用人工智能的宏大未来时,切斯基选择回到“用户体验”这个具体而微的起点,试图证明另一种通往AI时代的路径。

无论成功与否,切斯基的入局都为当前由少数巨头主导的AI竞赛增添了关键变量。它宣告了一个更务实、更产品导向的AI发展路径的潜在回归。硅谷的权力游戏从未停止,只是棋盘换了。以前是操作系统和社交网络,现在是大模型和交互界面。而布莱恩·切斯基,正试图从一个精明的“辅政大臣”,转型为自立门户的“新王”。这场戏,值得持续看下去。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

产品发布 产品发布 大模型 大模型
Share: 分享到: