Dylan Patel predicts the AI inference market could surpass oil, with OpenAI+Anthropic needing 100GW by 2030. According to SemiAnalysis founder Dylan Patel in a Sequoia Capital podcast interview on July 1, AI inference will become one of the largest global markets, potentially surpassing oil and accounting for several percentage points of global GDP. Patel predicts that by 2030, the combined computing power demand of just OpenAI and Anthropic will exceed 100 gigawatts, and power constraints may persist in the long term as task capabilities expand faster than computing growth. SemiAnalysis team's InferenceX benchmark system shows that inference costs drop by about 60x each year, with efficiency per watt improving about 40x. #Gate完成141只股票股息派发