A year after the DeepSeek shock following R1's release, US tech companies still lead in AI chips, model quality, and sales, with no change in investment trends
About the author: Chris Miller is the author of Chip War: The Fight for the World's Most Critical Technology.LinkedIn:Chris MillerLinkedIn:Chris Miller:One year has passed since the Chinese AI lab DeepSeek launched its R1 model, wiping $1 trillion off financial markets in a single day …
Context & Ripple Effects
DeepSeek’s R1 turned an apparent model-efficiency breakthrough into a test of whether lower-cost Chinese innovation could reset the AI competitive order. Earlier coverage found that export controls had also encouraged DeepSeek to innovate without the newest chips, complicating a simple view of controls as a one-way constraint.
The subsequent arc showed both diffusion and adaptation: Chinese companies were building services on DeepSeek, while V3.1 was tailored for next-generation domestic chips. This report assesses whether those developments translated into a change in the US commercial and infrastructure lead.
First-order effects
- The reported result favors US tech companies and their AI-chip ecosystems: their lead in chips, model quality, and sales remains intact despite the R1-driven market repricing.
- DeepSeek remains a consequential technical competitor, but the report indicates that its breakthrough has not, so far, altered overall AI investment behavior or displaced US leaders commercially.
Second-order effects
- Unchanged investment trends preserve demand signals for AI compute and infrastructure, limiting the case that efficient models alone will trigger an immediate capital-spending retreat.
- Chinese AI suppliers have added pressure to build around domestic hardware—V3.1’s optimization for Chinese-made chips is an example—but that adaptation has not yet produced a broader sales or quality reversal in this account.
Third-order effects
- The episode points to a more durable split between model-level innovation and platform-level advantage: a technical shock can move markets without quickly reallocating chip leadership, distribution, and revenue.
- If this pattern persists, AI competition will be shaped less by a single model release than by the ability to pair model advances with sustained compute investment and commercial reach.
The trend: AI is moving toward a stacked competitive structure in which model breakthroughs intensify rivalry but do not automatically overturn infrastructure and distribution advantages.