DJI (大疆) and Yingshi (影石)'s stunt — us onlookers really can't benefit

Market shock: a cash-rich company punished anyway

Kuaishou (快手) plunged about 14% in Hong Kong trading on March 26, knocking its market value below HK$200 billion, even after the company posted record annual revenue and profit. The disconnect is stark: solid last-year numbers, a video-generation model with reportedly more than 60 million users and an annualized revenue run‑rate above $300 million in January, and yet the market stamped a sell signal. Why did investors turn away so quickly? Perhaps because headline AI projects — think of the spectacle around DJI (大疆) or marketing plays by smaller camera vendors such as Yingshi (影石) — no longer cut it.

Fundamentals vs. theatre

Kuaishou told investors it will spend RMB 26 billion (260亿元) on AI-related capex this year to build out datacenter and model capacity. But much of the company remains tied to legacy video advertising and e‑commerce economics; AI is framed as an accelerator, not a new business core. It has been reported that its flagship model “KeLing” (可灵) made an early splash, but competitors such as ByteDance’s Seedance 2.0 have taken the conversation elsewhere. When AI is used mainly to optimize old workflows rather than to rewrite the company’s business model, analysts and funds are less willing to re‑rate growth trajectories.

A broader lesson for China’s tech players

This episode underscores a wider trend in China: the AI era accelerates winners and losers. Firms that build independent AI stacks and treat model development as a strategic frontier — akin to the big platform groups that can afford deep, standalone AI units — will likely capture more capital, talent and scale. Those that treat AI as a toolkit for incremental efficiency risk being sidelined. Geopolitics and chip access also loom in the background; trade restrictions and export controls make domestic compute and model sovereignty an investment priority for both regulators and boards.

The choice ahead

Investors now ask a blunt question: is AI your product or your PR? Kuaishou’s results show that having a promising model or a headline capex number isn’t enough. If companies want the market to reward them, they must decide whether to pivot fully — with the attendant costs and strategic trade‑offs — or accept a valuation that treats AI as an adjunct. Which path will Chinese tech take? The answer will determine who gets the next era’s spoils — and who is left watching from the sidelines.