Huang Renxun's latest interview "AI's Blueprint for the Next 10 Years" is worth learning from. I just finished watching Huang Renxun's latest interview, and the core points in the video have benefited me a lot. At present, AI is only the beginning, and the next decade will be the real explosive cycle of AI. Do not overestimate short-term and underestimate long-term changes. He said that currently generative AI is just the beginning, and the next big wave will be the era of AI agents; AI will evolve from simple question answering to autonomous planning, cross software operation, and independent completion of the entire workflow. Then he talked about the underlying business logic of the AI industry. He put forward a core proposition: calculation is income, watts are income, and each token is income. Data centers will evolve into AI factories, with computing power, electricity, and storage becoming core hard currencies; The efficiency of computing power directly determines the profitability level of enterprises. He also talked about the view of AI market foam and AI and employment. He said that there is no need to worry about the AI foam at this stage. Chips, storage, power, and infrastructure manpower are all in short supply, and supply cannot keep up with demand in the long run. It is difficult for the industry to overheat and collapse in the next five years. AI will not eliminate positions, it will eliminate people who do not know how to use AI; A large amount of repetitive mental work is being replaced, and at the same time, a massive number of new tracks are emerging, which will exponentially amplify human productivity. When it comes to the recent competition and model roadmap between China and the United States in AI, he believes that open source models and closed source models will coexist for a long time; Each country needs to master its own AI capabilities, and core data and intelligence cannot be completely outsourced; Recognize the diverse competitive landscape of global AI talent. Regarding the long-term bottleneck of AI, he believes that the biggest short-term constraints are advanced chips and memory storage; The core bottleneck in the medium to long term is power supply, and large-scale AI computing facilities in the future will require supporting energy solutions. In the era of AI, the opportunities for ordinary people and enterprises lie in the gradual iteration of traditional PC models and the popularization of AI terminals; All industries will be restructured by AI. The future winners are participants who understand how to use AI agents to transform their own businesses and master the ability to apply computing power. Do you think Huang Renxun's views on AI are of great reference value?