Rather than competing to build generalist models, China's leading AI startups (DeepSeq, Moonshot, ZAI, Minimax) have each carved out a niche like coding, agents, or multimodality. This vertical focus is a necessary survival strategy driven by capital, compute, and talent limitations.

Echoing Don Valentine's VC wisdom that 'scarcity sparks ingenuity,' US restrictions on advanced chips are compelling Chinese firms to become hyper-efficient at optimizing older hardware. This necessity-driven innovation could allow them to build a more resilient and cost-effective AI ecosystem, posing a long-term competitive threat.

Facing compute and capital shortages, Chinese AI labs don't pioneer frontier research. They wait for Western labs to publish breakthroughs, likening it to 'knowing the answer to the homework,' then work backwards to replicate them, focusing resources on efficient post-training.

A critical, under-discussed constraint on Chinese AI progress is the compute bottleneck caused by inference. Their massive user base consumes available GPU capacity serving requests, leaving little compute for the R&D and training needed to innovate and improve their models.

A key strategic difference in the AI race is focus. US tech giants are 'AGI-pilled,' aiming to build a single, god-like general intelligence. In contrast, China's state-driven approach prioritizes deploying narrow AI to boost productivity in manufacturing, agriculture, and healthcare now.

The performance gap between Chinese and American frontier AI models is not due to a lack of talent or different training techniques. Instead, it is primarily constrained by access to massive-scale compute and the capital required to procure it.

An Alibaba tech lead claims the US compute advantage allows for wasteful but effective "rich people innovation" (running many experiments). In contrast, Chinese firms are forced into "poor people innovation," bogged down by operational needs and unable to risk compute on next-gen research.

Faced with limited access to top-tier hardware, Chinese AI companies have been forced to innovate on model architecture to compete. They've developed superior techniques in memory management and multi-token prediction, making their models highly efficient and formidable competitors despite hardware constraints.

Despite the dominance of large AI labs, they face constraints in compute, talent, and focus. Startups can thrive by building highly specialized products for verticals the big players deem too niche. This focused approach allows them to build better interfaces and achieve deeper market penetration where giants won't prioritize competing.