Two Months. $10 Million. A Sputnik Moment for AI.

A small Chinese startup, DeepSeek, reportedly built an efficient large AI model (R1) in two months for about $10 million with under 200 engineers, triggering major market volatility and a massive selloff in incumbent AI hardware and cloud names. The article argues MIT research confirms diminishing returns for ever-larger models, suggesting more efficient models could reduce demand for high-end GPUs and reshape winners in the AI ecosystem. That shift creates trading opportunities across software platforms, AI-enabled drug discovery, data providers and on-site energy for data centers. The author highlights recent successful trades (e.g., JD) and promotes a May 28 webinar that applies a 20-factor “Convergence Trigger” system to find volatility-driven setups. Overall, the piece frames this as a bullish reallocation opportunity amid higher sector volatility rather than a structural collapse of incumbents.