The recent release of Moonshot's Kimi 3 AI model has sparked comparisons to the "DeepSeek moment" of early 2025, fueling concerns about the US AI lead. While Jevons Paradox suggests increased efficiency leads to more chip demand, this "Moonshot moment" might differ.

Unlike the past, where Nvidia was the sole proxy for AI investment, the market now includes public entities like SpaceX and significant secondary market activity for private AI firms like OpenAI and Anthropic. This diversification means that while chip demand might remain robust, the profitability of specific model makers could be compressed by competitors like Kimi 3.

This scenario could favor US infrastructure companies such as Meta, Amazon, and Alphabet, which could host cheaper Chinese models. Alternatively, if AI companies' massive capital expenditures don't yield profits due to a fleeting AI edge, future infrastructure growth might be driven by consumer needs rather than a competitive race among AI giants. The pace of demand could slow if the anticipated profits from AI development don't materialize.