[AI Revenue Can't Keep Up with Spending Speed, Google Unveils 'Self-Evolution' New Narrative] According to monitoring by Beating, Google is expected to invest approximately $200 billion this year in AI data centers and devices. DeepMind's Chief Strategy Officer Jasjeet Sekhon admitted that current AI revenue cannot support this level of expenditure, and the industry may face a 'spending first, revenue lagging behind' gap period. The long-term justification Sekhon provided for Google's massive investment is RSI, which involves using AI to develop the next generation of AI, continuously improving training methods and the models themselves. Sekhon stated that RSI has become one of the core logics behind this round of significant investment. However, true RSI has not yet emerged. Current AI can only optimize certain algorithms, kernels, and training processes. Sekhon predicts it may be realized within the next two years. Google is using a yet-to-be-implemented technological goal to justify today's approximately $200 billion investment. [Original Link]