Prominent Developer Ignites AI Code Reading Debate, Delivers Scathing Critique of Apple's iOS Ecosystem

A recent online discussion, sparked by a prominent developer’s tweet, has intensified the debate around the necessity of human review for AI-generated code. The developer argues that most engineers excessively read code and fail to leverage AI for generating sufficient ‘slop’ or disposable code. This ‘cheap code,’ he contends, should be used for extensive testing, verification, and exploratory debugging, particularly for mission-critical systems. He posits that the economics of code generation have fundamentally shifted, making it illogical not to produce vast quantities of auxiliary code to enhance the robustness and quality of production systems, rather than solely focusing on meticulously hand-reviewing every line of shipped code.

Concurrently, the developer delivered a comprehensive and passionate critique of Apple’s iOS ecosystem, alleging it actively stifles innovation. Key grievances include a ‘hellish’ developer experience marked by burdensome provisioning and signing processes, and the perceived ‘garbage’ state of Xcode. He highlights Xcode’s reported legacy as a re-branded NextStep codebase, citing its frequent crashes, non-deterministic builds, poor Git integration, and unreliability of features like Swift UI previews. Furthermore, Apple’s App Store policies are described as ‘hostile,’ citing the mandatory 30% revenue cut, restrictive content bans (e.g., game streaming), forced ‘Sign in with Apple,’ and alleged tacit endorsement of predatory subscription models that disproportionately affect children. The restrictive sandboxing model, which prevents inter-app communication, dynamic execution (JIT), alternative browser engines (WebKit exclusivity), and plugin development, is also condemned for hindering entire categories of software and future AI-driven experiences. The critique extends to Android, which is characterized as a ‘B-tier iOS’ that largely copies Apple’s restrictive behaviors, collectively suppressing a generation’s potential for creative software development.