The next major shift in artificial intelligence will focus on efficiency and cost reduction, according to Kelly, the former privacy chief at Facebook. As the AI industry continues its rapid expansion, the massive infrastructure buildout driven by hyperscalers has raised urgent concerns about sustainability and scalability. According to S&P Global, over $61 billion in infrastructure deals were made in the data center market in 2025 alone, as companies race to construct facilities to support the growing demand for AI compute. OpenAI has committed more than $1.4 trillion in AI-related investments over the coming years, including major partnerships with GPU leader Nvidia, cloud provider Oracle, and infrastructure firm Coreweave. However, this surge has sparked alarm over the strain it places on the global power grid. The energy demands of these data centers are immense. In September, Nvidia and OpenAI announced a project involving at least 10 gigawatts of data center capacity—equivalent to the annual electricity use of 8 million U.S. households. That same 10 gigawatts also matches New York City’s peak summer power demand in 2024, as reported by the New York Independent System Operator. The high cost of building and operating these facilities has also come under scrutiny. A turning point came in December 2024, when DeepSeek released a free, open-source large language model for under $6 million—far less than the hundreds of millions typically spent by U.S. competitors. This milestone has highlighted the potential for more cost-effective AI development. Kelly predicts that this will lead to a new wave of innovation centered on efficiency. He expects a growing number of Chinese AI players to emerge on the global stage, especially after President Donald Trump approved the sale of Nvidia’s H200 chips to China. These developments could accelerate access to foundational AI capabilities in regions previously limited by hardware restrictions. Open-source models from China, he noted, are likely to provide broad access to basic levels of compute, generative AI, and even agentic systems, democratizing AI tools and challenging the dominance of Western tech giants. As the industry matures, the focus is shifting from raw scale to smarter, leaner, and more affordable AI development.