The days when Big Tech was the unrivalled champion of artificial intelligence are over. A wave of Chinese AI models, led by DeepSeek, has shattered Western assumptions about who leads in this critical technology. DeepSeek's latest model, trained at a fraction of the cost of comparable American systems, matches or exceeds them on key benchmarks. This has sent shockwaves through Silicon Valley — not merely because of the technical achievement, but because it demolishes the premise that export controls on advanced chips could contain China's AI progress. If anything, the constraints appear to have sparked a burst of ingenuity: doing more with less. The geopolitical implications are profound. For years, Washington has bet that restricting access to cutting-edge semiconductors would preserve American AI primacy. DeepSeek has demonstrated that algorithmic innovation can compensate for hardware disadvantage, raising uncomfortable questions about the efficacy of the current export-control regime. Meanwhile, the open-source nature of many Chinese models stands in stark contrast to the proprietary walled gardens of OpenAI and Google, potentially reshaping the economics of the entire AI industry.