Why is it said that "the end of AI is electricity"? How can we achieve a win-win development pattern with the automotive industry?

2026-03-10 16:13:37

As the concept of "software-defined vehicles" continues to evolve towards "AI-empowered vehicles", AI has become increasingly deeply integrated with automobiles.

Recently, in the highly anticipated draft of the "14th Five-Year Plan," AI has emerged as one of the significant projects for cultivating new productive forces. The statement that "the ultimate frontier of AI is electricity" has once again brought the energy revolution, represented by new energy vehicles, to the forefront.

Important tool of industrial revolution

AI is a crucial tool for the new round of industrial revolution, and its significance has become increasingly evident in various fields, including the automotive industry.

The latest data shows that during the week of February 9th to 15th, China's AI model surpassed the US model for the first time, with a call volume of 4.12 trillion tokens, compared to the latter's 2.94 trillion tokens during the same period. From the 16th to the 22nd, the call volume of the Chinese model further soared to 5.16 trillion tokens, marking a significant increase of 127% in three weeks. Meanwhile, the call volume of the US model during the same period dropped to 2.7 trillion tokens, further widening the lead of the Chinese model. Among the top five models globally in terms of call volume, four are from Chinese vendors, namely MiniMax M2.5, Dark Side of the Moon Kimi K2.5, Zhìpǔ GLM-5, and DeepSeek V3.2, collectively contributing 85.7% of the total call volume for the top five. Notably, DeepSeek has already established deep collaborations with numerous automotive companies.

Recently, Tesla CEO Elon Musk stated that electricity and heat dissipation are the core limiting factors in the current development of artificial intelligence, and China, with its strong power supply capacity, will have a significant advantage in AI computing power. This statement directly pushed the dependence of AI development on electricity into the global spotlight, and "electricity anxiety" has also begun to appear frequently in people's minds.

How much electricity does AI actually consume? Public information shows that a medium-sized AI data center consumes as much electricity as 100,000 ordinary households combined. Taking GPT-4 as an example, its training process can be described as a power "gamble", consuming as much electricity as a small nuclear power plant generates in a year. And ChatGPT consumes more than 500,000 kWh of electricity daily. This means that its daily electricity consumption can meet the daily needs of thousands of households in a small city.

According to a report released by the International Energy Agency, it is estimated that by 2030, the electricity demand of global data centers will more than double, reaching approximately 945 terawatt-hours, slightly higher than the current total electricity consumption in Japan. And artificial intelligence is precisely the main "user" of this surge in electricity consumption.