ChiNext Index rises over 2% 创业板指涨超2%

Perplexity CEO Aravind Srinivas stated on camera that regardless of when giants like Anthropic and OpenAI decide to go public, their own IPO timeline is firmly set for 2028. The remark sounds calm, assertive, and even carries a touch of "lone bravery"—focusing on their own path while others compete. Yet, beneath this calm, I detect a pervasive and nearly irrepressible industry-wide anxiety—a deep-seated yearning for valuations, market entry, and the realization of gains in this ultimate high-stakes gamble.

2028. This timeline itself is a strategically calculated statement. It’s not too distant, providing investors and employees with a clear "beacon of hope," while also leaving ample room to navigate any potential black swans or gray rhinos in the AI field. This isn’t a casual plan—it’s a meticulously crafted exercise in "expectation management." Srinivas’s underlying message is: "Look, we haven’t been rattled by the giants’ pace; we have our own strategic resolve." But this very resolve underscores an incredibly stark reality at the heart of the AI race: In this winner-takes-all, cash-burning domain, going public has almost become the most vital "safe haven" and "refueling station" for startups.

Why the urgency? The answer lies in every massive funding report and every heated debate over GPU computing power. Training a frontier large language model costs enough to wipe out a mid-sized corporation’s annual profits. Perplexity operates in the AI search arena, directly facing the squeeze from trillion-dollar behemoth Google and the Microsoft/OpenAI alliance. They need continuous funding to purchase exorbitantly priced NVIDIA chips and must lure top AI scientists worldwide with sky-high salaries. All these efforts require a stage more imaginative and liquid than private markets. An IPO is that ultimate stage. It signifies not just capital, but also industry validation and a stronger voice within the broader ecosystem.

Shift your gaze from Perplexity’s ambitions and look around—you’ll find this "sense of urgency" everywhere. The news headlines reveal more clues: ChatGPT’s push to become a "super app" is clearly an attempt to cement its position in the OS-level ecosystem before user habits solidify. Apple’s launch of the new Siri reflects a desperate counterattack after being cornered by Google Gemini and Copilot. Even headlines like "Companies stop hiring after adopting AI" brutally expose the cold and immediate reshaping power of technological iteration on the labor market. Every player is accelerating, sprinting as if the starting gun has long sounded and only one golden throne awaits at the finish line.

Thus, Perplexity’s "2028 manifesto" is less a calm blueprint and more a pledge of allegiance to capital and the market. Its translation is: "We are serious, we embrace long-termism (even if pressured), and we ask for your trust and valuation." In an industry like AI, which heavily depends on future narratives, confidence is more precious than gold—and a clear IPO expectation is the most effective confidence booster.

However, this path is crowded and brutal. OpenAI and Anthropic are like two racing cars already far ahead, while Perplexity, Mistral, and numerous Chinese LLM companies are slamming on the gas in pursuit. Will 2028 be a season of harvest or a new red ocean? Only time will tell. What’s certain is that when only a few players eventually cross the IPO finish line, draped in valuations and glory, all present-day rhetoric will be recalibrated by the harsh market reality of that time. The current composure is merely the posture a helmsman must maintain before the storm arrives.

Perplexity CEO 阿拉温德·斯里尼瓦斯对着镜头说，不管Anthropic和OpenAI这两家巨头何时敲钟，我们自己的IPO时间表是钉死在2028年了。这话听起来冷静、有主见，甚至有点“你们打你们的，我走我的路”的孤勇。但在这份冷静背后，我嗅到的是一种行业普遍性的、几乎按捺不住的焦虑——对估值、对门票、对在这场终极豪赌中离场变现的深层渴望。

2028年。这个时间点本身就是一个充满心机的声明。它既不算太远，足以让投资人和员工看见一个明确的“盼头”；又留有足够的余地，可以应对AI领域任何可能的黑天鹅或灰犀牛。这不是一句随口的规划，这是一次精心计算的“预期管理”。斯里尼瓦斯真正想说的是：“看，我们没有被巨头的节奏打乱，我们有自己的战略定力。” 但这种定力，恰恰反衬出当前AI赛道一个无比现实的底色：在这个赢家通吃、烧钱如流水的领域，上市，几乎成了初创公司最重要的“安全屋”和“加油站”。

为什么这么急？答案写在每一份巨额的融资报告和每一次关于GPU算力的争吵里。训练一个前沿大模型的费用，足以让一家中型上市公司一年的利润蒸发。Perplexity所在的AI搜索赛道，直面的是Google这个万亿美金巨兽和微软/OpenAI联盟的挤压。他们需要持续融资来购买天价的英伟达芯片，需要高薪挖角那些全世界都在抢的AI科学家。而所有这些动作，都需要一个比私募市场更具想象力、更具流动性的舞台来支撑。IPO，就是那个终极舞台。它不只意味着钱，更意味着行业地位的“认证”和更广泛生态的“话语权”。

把视线从Perplexity的野心移开，扫一眼周围，你会发现类似的“急迫感”无处不在。那些简讯标题里，藏着更多线索：ChatGPT要搞“超级应用”变身，这分明是想在用户习惯固化前，把自己焊死在操作系统级别的生态位上；苹果推出全新Siri，是被Google Gemini和Copilot逼到墙角后的绝地反击；甚至“用上AI后，公司停止招人了”这种新闻，都赤裸裸地揭示了技术迭代对劳动力市场那冷酷而即时的重塑力。所有玩家都在加速，都在抢跑，仿佛起跑枪早已打响，而终点线旁只有一个黄金宝座。

所以，Perplexity的“2028年宣言”，与其说是一份从容的计划书，不如说是一张写给资本和市场的投名状。它翻译过来就是：“我们很认真，我们有长期主义（虽然可能被逼），请给我们信任和估值。” 在AI这个极度依赖未来故事的行业里，信心比金子更贵，而一个清晰的上市预期，就是最有效的信心强心剂。

只是，这条路拥挤且残酷。OpenAI和Anthropic就像两辆已经开出很远的赛车，而Perplexity、Mistral、甚至更多的中国大模型公司，都在拼命踩油门。2028年，会是一个收获的季节，还是一片新的红海？时间会给答案。但可以肯定的是，当最终只有少数玩家能带着满身估值和光环冲过IPO终点线时，现在所有的豪言壮语，都会被那时残酷的市场现实重新称量。此刻的从容，不过是暴风雨来临前，掌舵者必须保持的姿态罢了。