Cross-Platform Development Without the Old Trade-Offs
Cross-platform work is no longer automatically a compromise. With React Native and Expo, one codebase can serve iOS, Android and web while keeping the experience consistent, and AI-assisted coding can shorten the path from idea to working feature. Charlie Cheever, Expo co-founder and CEO, explains how Expo lets developers build apps and sites that run across platforms with React, why venture-backed teams have gravitated toward React Native, how AI is speeding up development, and why the pace of iteration outweighs the choice of programming language.
Lightning-fast iteration
A small change used to trigger a long rebuild as an app grew. A blank screen with a single line of text builds almost instantly, but by six or seven screens the wait can stretch to 45 seconds or 1 minute 45, and for huge sprawling apps like Uber it can exceed 45 minutes, breaking the flow of testing an idea. Expo and React Native change that rhythm: save a change and see it reflected on a phone or simulator right away. The practical result is shorter experiment cycles, quicker fixes and more frequent releases. Before choosing a stack, check how fast a trivial UI change reaches a device, whether hot feedback works on both iOS and Android, and how the workflow behaves once the project has dozens of screens.
Unified tech stack
Building separately for each platform often produced inconsistent interfaces and required platform-specific hires. Yet users increasingly expect the same product everywhere: the same layout, design language and placement of controls, whether they open the app on iPhone, Android or a browser. A shared React Native codebase makes that consistency easier to maintain and simplifies staffing. Bluesky is given as an example, built for iOS, Android and web by one developer in four months. Use this as a scope check: list the screens and interactions that must match exactly across platforms, then verify that one codebase can deliver them without separate native implementations.
Coding with AI
AI tools such as Claude and GitHub Copilot now handle much of the routine work of writing code and building features. Their strength comes from a feedback loop: generate code, run it, and observe whether it works. That shifts developer effort toward higher-level decisions, including user-experience design and complex problem-solving. A useful division of labour is to let AI draft and test routine implementation while people review product behaviour, edge cases and interface quality. Speed from AI combined with human judgement is the advantage, not automation alone.
Conclusion
The broader lesson is that tooling choices shape how quickly a subscription app business can learn. Fast feedback, one consistent codebase and AI assistance reinforce each other: changes are written sooner, seen sooner and shipped sooner. For a fuller discussion, listen to the Sub Club podcast conversation with Charlie Cheever on how Expo, React Native and AI are changing app development.
Frequently asked questions
What does a unified tech stack change for users and hiring?
A single codebase can keep screens, navigation and design language aligned across iOS, Android and web, so users encounter the same product rather than platform-specific variants. It also reduces the need for separate platform specialists.
Why does iteration speed matter so much?
When a change appears on a phone or simulator immediately after saving, teams can test and refine ideas in short cycles instead of waiting for a rebuild. That feedback speed is presented as more important than the particular programming language.
How do AI coding tools fit into the workflow?
Tools such as Claude and GitHub Copilot can draft code, run it and check whether it worked, reducing routine implementation work. Developers can then spend more time on user-experience choices and harder problems.