Does Ruby on Rails still make sense?

The writer argues that Ruby on Rails remains a sensible choice even as LLMs can generate large amounts of code. Frameworks still matter because they provide structure, reusable abstractions, security and performance updates, and predictable behavior that benefits both humans and code-generating agents. Rails reduces the need to reinvent plumbing like ORMs and job queues, makes projects easy to navigate and debug, and offers developer conveniences - instant code reloading and a REPL - that speed iteration. LLMs are trained on these ecosystems, so using established libraries and languages amplifies their effectiveness rather than rendering them obsolete.

A full rewrite into Rust or another language is presented as costly not just in code-generation tokens but in behavioral risk and testing: different runtimes and libraries change subtle behaviors users rely on, and UI changes can provoke complaints. The writer prefers optimizing the existing Rails app, selectively exploring Go for backend services (fast compile, readable) and Svelte for frontend experiments, while warning that moving to a JS-stack implies big breaking changes and UX pitfalls like spinners from API-driven UIs. The conclusion is firm: Rails still makes sense for most business applications; rewrites may be justified for niche performance needs but are not a universal win.