AI didn't kill React Native. Shopify just found a different bet.

The catalyst. Shopify announced plans to move its mobile apps from React Native back to native Swift and Kotlin. The decision, detailed in a company blog post, reverses a 2020 bet that had been described as extremely successful just a year prior. The stated reason: AI coding agents have fundamentally changed the cost calculus of building for two platforms natively.

The AI argument. Shopify claims that LLMs now handle enough of the implementation, translation, testing, and review work that the write-once benefit of React Native is no longer a deciding factor. Their internal system Helix uses a checkpoint-and-review loop to port screens from React Native to native. The company rebuilt its Shop app from proof-of-concept to published in 12 weeks using this approach.

The missing Expo factor. The video host argues that Shopify's decision is less generalizable than it appears. Shopify was not using Expo, the modern toolchain that solves most of React Native's development friction, including over-the-air updates, hot reloading, and simplified simulator management. Teams using Expo face a different cost-benefit tradeoff and may not see the same math.

The over-the-air advantage. React Native's ability to push JavaScript bundle updates without App Store approval remains a structural advantage that native code cannot replicate. Shopify's move sacrifices this capability. For teams that ship frequently or have long review cycles, losing OTA updates is a real operational cost that gets little attention in the AI framing.

The talent shift. Meta has significantly reduced investment in the React Native team. Many of its top native engineers have left. Meanwhile, AI agents have narrowed the skill gap between average web developers and expert native developers. The video host argues this talent change, as much as the technology change, is what makes Shopify's bet plausible, but also risky for teams without Shopify's engineering resources.

The bottom line. Shopify's move is a valid experiment, not a general recommendation. It works for organizations with deep pockets, strong AI tooling, and the willingness to rebuild greenfield. Most teams should wait for more data before abandoning React Native. The signal is real but the evidence is thin, and the tradeoffs around OTA updates and dependency management are understated.