Source: Compiled from ifeng Tech, July 21, 2026
According to a report compiled and published by ifeng Tech on July 21, The Wall Street Journal, on the ground at WAIC in Shanghai, observed that the rapid progress of Chinese AI has made US companies feel the pressure intensifying.
Kimi K3: Refuting the "Distillation" Narrative with Strength
The Wall Street Journal noted in its report that Anthropic and OpenAI had previously accused multiple Chinese companies of using American models to train their own, calling it "adversarial distillation." Anthropic executives even claimed that without distillation, American companies might have maintained a one-to-one-and-a-half-year lead over China.
However, the release of Kimi K3 by Moonshot AI on the first day of the conference overturned this narrative. Kimi K3's scores on multiple benchmarks surpassed Anthropic's Claude Opus 4.8—the very model it was theoretically supposed to have distilled. Research suggests that smaller models rarely outperform the larger models they are distilled from. Some researchers, including OpenAI executives, have stated that model distillation can no longer explain the rapid progress of Chinese AI models.
One Quarter of China: AI is Replicating the Smartphone and EV Story
The report quoted tech author Mehran Gul's view that China has already captured dominant market share from the West in three industries: smartphones, electric vehicles, and batteries. "If this has happened in three industries, will it happen again in AI? My answer, based on what has already happened in AI, is 'most likely.'"
China's manufacturing capabilities give it an advantage in pushing AI into the real world. Wang Xiaogang, Chairman of Daxiao Robotics, stated: "Chinese robots are complex systems with many components. You need not just one company, but a complete supply chain and many partners. In China, companies are very good at this."
The report also noted that Chinese AI models are gaining increasing global attention, including in the US market. Many Chinese models are open-source and generally cost less to run than Western models, factors that are driving US companies to turn to Chinese solutions. As previously reported by NPR, under AI cost pressures, many US companies are adopting "high-quality and affordable" Chinese models. One startup founder put it bluntly: "This is a very, very simple business decision."
This WAIC observation piece from The Wall Street Journal represents, to some extent, mainstream anxiety in Silicon Valley. The emergence of Kimi K3 is not just a product success but also logically dismantles the narrative that "Chinese AI can only rely on distillation." When Chinese AI simultaneously possesses both "independent innovation capability" and "cost-performance advantages," the competitive landscape of the global AI industry is undergoing profound changes.