A $600 billion market cap loss revealed AI’s commodity future.
DeepSeek’s announcement sent NVIDIA’s stock tumbling, shedding $600 billion in market cap. Wall Street woke up to the fact that AI is becoming cheaper to train, and AI models can train themselves, further lowering the cost. This cost erosion creates an economic reality in which AI becomes a commodity. DeepSeek-V3’s last training run cost just $5.576 million, compared to the $100 million that GPT-4 supposedly cost. That is 18 times smaller. This demonstrated that high-quality AI can be built inexpensively without necessarily requiring the most advanced hardware.
— DeepSeek - New Economic Rules And Regulatory Challenge · Forbes