Why I Still Choose Ruby on Rails in the Age of AI

On May 17th, a developer published a defence of Ruby on Rails on Dev.to that arrives at exactly the wrong moment to be fashionable and exactly the right moment to be useful. The AI coding boom has sent developers scrambling toward whatever framework feels most "modern," most JavaScript-adjacent, most aligned with whatever a language model can scaffold in thirty seconds. Rails, meanwhile, has been quietly accumulating a different kind of advantage. The argument in the post is structural rather than sentimental: because Rails has been around long enough to saturate the training data of large language models, and because its "convention over configuration" philosophy means there is genuinely one right way to do most things, AI assistants can navigate a Rails codebase with unusual confidence and accuracy. The framework's opinions, long criticised as constraints, turn out to function as a shared grammar between human developer and AI pair-programmer. Magic, in Rails parlance, becomes legibility.

If confirmed as a broader community sentiment rather than a single developer's experience, this matters in ways that cut across the current vibe coding conversation. The dominant assumption in that conversation is that newer, more flexible frameworks give AI more room to work. This argument inverts that entirely — constraint, not flexibility, is what makes AI collaboration coherent. A codebase built on strong conventions gives an LLM fewer ambiguous choices to make badly. Second-order effects follow: teams evaluating frameworks for AI-assisted development may need to weight "training data saturation" and "convention density" as serious criteria, not nostalgia metrics. Rails communities that have been playing defence for several years may find themselves, improbably, playing offence.

Watch for other developers — particularly those working in AI-assisted workflows — publishing similar experience reports, or for any structured comparison between Rails and newer frameworks on AI pair-programming outcomes. A second signal from an independent source would shift this from interesting hypothesis to something worth building strategy around.

NewsHive monitors these sources continuously. All signal titles above link to the original reporting.

Intelligence by NewsHive. Need help navigating what this means for your business? Contact GeekyBee →